ANALYST, SEPTEMBER 1989, VOL. 114 1029 Copper( 11)-selective Membrane Electrodes Based on o-Xylylene Bis(dithi0carbamates) as Neutral Carriers Satsuo Kamata, Hiroyuki Murata, Yoko Kubo and Ajay Bhale Department of Applied Chemistry, Faculty of Engineering, Kagoshima University, Korimoto, Kagoshima 890, Japan o-Xylylene bis(diisobuty1- and methyloctadecyl-dithiocarbamate) were synthesised and used as neutral carriers in membrane electrodes t o improve the selectivity for Cu2+. The electrodes based on these ionophores, with o-nitrophenyl octyl ether as a plasticising solvent mediator and potassium tetrakis- (p-chloropheny1)borate as an anion excluder, exhibit a linearity range of 10-1-10-6 M and have a Nernstian slope of 28-29 mV per decade at 25°C. The highly selective electrode based on o-xylylene bis(diisobuty1dithiocarbamate) rejected the interference of alkali, alkaline earth and transition metal cations by a factor in the range l O 2 - l O 4 and showed a high selectivity for Cu2+ even in chloride and bromide media.The properties of the electrodes are discussed and also compared with those of o-xylylene bis(diethy1dithiocarbamate) under similar measurement conditions. Keywords: o-Xylylene bis(diisobuty1- and meth yloctadec yl-dithiocarbamate); neutral carrier; copper(//)- selective membrane electrode Ion-selective electrodes based on ionophore ligands are well established for alkali and alkaline earth metal cations.1-5 Electrically neutral, lipophilic ion-complexing agents are known to behave as ionophores. These neutral carriers have the capability to extract ions selectively from aqueous solutions into a hydrophobic membrane phase and to trans- port these ions across barriers by carrier translocation.It is also well known that sulphur ligands co-ordinate with tran- sition metal cations as exclusive donor atoms. In this respect, macrocyclic polythiaethers and dithiocarbamates have attrac- ted widespread attention owing to the unique properties of the compounds. The macrocyclic polythiaethers have the ability to discriminate between closely related metal ions based on the relative fit of the ligand cavity size to the metal ion radius.6 Recently, we have investigated the use of the macrocyclic polythiaether 13,14-benzo-1,4,8,11-tetrathiacyclopenta- decane (BTTCP) as a neutral carrier for a Cull-selective membrane electrode.' The dithiocarbamate derivatives are highly versatile chelating agents for metal cations, but these compounds form complexes with many metals non-selec- tively.In connection with this, we recently prepared o-xyly- lene bis(diethy1dithiocarbamate) (o-XBDEDTC) and showed it to be usable as an ionophore for a Cu"-selective membrane electrode .s In copper-selective membrane electrodes, there are two other types of electrode based on a liquid ion exchanger9710 and a pressed crystal of copper and silver sulphides which is used commercially such as the Orion 94-29 CuIL-selective electrode. The commercial electrode is affected by a consider- able amount of interference from chloride and bromide ions, because of the influence of silver sulphide on the membrane.It is possible to modify the structure of ionophores as neutral carriers to improve the Cuz+ ion selectivity. In line with the above, we synthesised the lipophilic ionophores o-xylylene bis(diisobuty1dithiocarbamate) (0- XBDiBDTC) and o-xyl ylene bis(methyloctadecy1dithiocarba- mate) (o-XBMODTC), which have non-macrocyclic cavities with four donor sulphur atoms and an o-xylene backbone, expected to form a complex selectively with the transition metal cations. We therefore employed these ionophores with o-nitrophenyl octyl ether (NPOE) as a plasticising solvent mediator in the CuQelective membrane electrode and compared these results with those of the o-XBDEDTC ionophore and with a commercial CuS - AgzS solid membrane electrode.Experimental Reagents and Chemicals Poly(viny1 chloride) (PVC) and NPOE were obtained from Fluka (Buchs, Switzerland). Potassium tetrakisb-chloro- pheny1)borate (KTCPB) and sodium tetrakis[3,5-bis(trifluo- romethy1)phenyllborate (NaTFPB) as anion excluders were obtained from Dojin Kagaku (Kumamoto, Japan). All solutions were prepared from analytical-reagent grade salts using distilled, de-ionised water. The chlorides of the metals were used in all instances, except for strontium bromide. Preparation of o-XBDiBDTC and o-XBMODTC (Fig. 1) o-XBDiBDTC Diisobutylamine (0.14 mol) and propan-2-01 (25 ml) were dissolved in water (300 ml) containing sodium hydroxide (0.14 mol). This mixture was stirred at room temperature and carbon disulphide (0.14 mol) was slowly added.The resulting precipitate of sodium N,N'-diisobutyldithiocarbamate was filtered off and crystallised from propan-2-01, yield 87%, m.p. 99-100 "C. The sodium N, N'-diisobutyldithiocarbamate (0.05 mol) was dissolved in ethanol (300 ml) and o-xylylene dibromide (0.025 mol) was added slowly to the solution while refluxing and stirring for 4 h. The final compound, o-XB- DiBDTC, was obtained as white crystals and recrystallised from ethanol, yield 6.07 g, 47%, m.p. 88-90°C. Elemental analysis: calculated (found), H 8.65 (8.69, C 60.87 (60.75), N 5.47 (5.46)%. IR (KBr), cm-1: 720 ( v ~ - ~ - ~ ) . NMR (in CDC13), 6 p.p.m.: 0.89 (d, 24H), 2.28 (m, 4H), 3.56 (d, 8H), 4.47 (s, 4H), 7.08 (m, 4ArH). R \ / R R ' S i-C4H9 i-C4H9 1, o-XBDiBDTC CH3 C18H37 2, 0-XBMODTC C2H5 C2H5 3, 0-XBDEDTC Fig.1. Sensor materials1030 ANALYST, SEPTEMBER 1989, VOL. 114 0-XBMODTC N-Methyloctadecylamine (0.03 M) and sodium hydroxide (0.03 M) were dissolved in a 1 + 1 aqueous ethanolic solution (200 ml). Carbon disulphide (0.03 M) was slowly added to the mixture while stirring at room temperature. Sodium N, N'- methyloctadecyldithiocarbamate was obtained by evapora- tion of the ethanol in solution. The compound was crystallised from distilled ethanol, yield 52%. o-XBMODTC was pre- pared by the reaction of sodium N , N'-methyloctadecyldithio- carbamate with o-xylylene disulphide by the same procedure as for o-XBDiBDTC, yield 51%, m.p. 51-52 "C. Elemental analysis: calculated (found), H 10.80 (10.67), C 70.18 (70.14), N 6.85 (6.85)%. IR (KBr), cm-1: 720 ( Y ~ - ~ - ~ ) .NMR (in CDC13), 6 p.p.m.: 0.85 (m, 6H), 1.15 (d, 68H), 3.34 (m, 6H), 4.60 (s, 4H), 7.25 (m, 4ArH). Membrane Composition and Electrode Fabrication The solvent polymeric membranes and electrodes were prepared as described.11 The membrane compositions are as follows: ionophore, 6.9 or 5.8% mlm of o-XBDiBDTC or o-XBMODTC; plasticiser, 34.3 or 46.6% mlm of NPOE; anion excluder, 1.6% mlm of KTCPB or 0.9% mlm of NaTFPB; PVC, 57.2 or 41.7% mlm. The electrodes were conditioned before use by immersion for 24 h in 10-3 M CuC12 solution. An Orion 94-29 copper-selective electrode was used to compare the properties of the trial electrodes. Electrode System and E.m.f. Measurement All measurements were performed with the following cell assembly: Ag - AgCl 1 10-3 M CuC12 1 sensor membrane I sample solution I reference electrode.A Corning digital research pH meter (Model 112) was used for monitoring the voltage. All the e.m.f.s were measured relative to a saturated calomel electrode (Iwaki Glass, Tokyo, Japan) in solutions stirred with a magnetic stirrer. The performance of the electrode was investigated by measuring the e.m.f.s of copper chloride solutions prepared in the concentration range 10-1- M by serial dilution. The activities of metal ions were based on activity coefficient (y) data calculated from the modified form of the Debye - Huckel equation, which is applicable to any ion: Log y = -0.511~2[</(1 + 1 . 5 6 ) - 0.2 p] where p is the ionic strength and z the valency. All the measurements were carried out at 25 k 0.1 "C.Selectivity Coefficients (kp&) The selectivity coefficients for foreign ions were determined by the mixed solution (fixed interference) method.7 A background concentration of the foreign ions was employed: 10-2 M solution for alkaline earth and transition metal cations and 10-1 M for alkali metal cations. A reference solution of 10-3 M CuC12 was employed. Results and Discussion The calibration graphs for the Cut'-selective membrane electrodes are shown in Fig. 2. Electrode A, which has o-XBDiBDTC as a neutral carrier, exhibited a short-range linearity of 10-2-10-4 M for Cu2+ activity and had a slope of 27 mV per decade. The slope and linearity range of electrode A remained constant despite an increase in the amount of ionophore. Electrode B with KTCPB added to the membrane as the anion excluder produced a straight line in the range 10-1-10-6 M and a Nernstian slope of 28 mV per decade. The o-XBMODTC electrode C produced a linearity range of lo-1-10-5 M with a slope of 26 mV per decade, and electrode t Lu a Fig.2. Calibration graphs for Cu2+ in CuC12 solution. Composition of membrane electrodes: A, o-XBDiBDTC only; B, o-XBDiBDTC and KTCPB; C. o-XBMODTC only; and D, o-XBMODTC and NaTFPB 2 3 4 5 b 7 8 PH Fig. 3. o-XBDiBDTC and KTCPB; and 2, o-XBMODTC and NaTFPB Effect of pH. Composition of membrane electrodes: 1, D, containing NaTFPB as an ion exchanger, exhibited a straight line from 10-1 to 10-6 M with a Nernstian slope of 28 mV per decade. The electrode gave a stable response, probably owing to the high lipophilicity of NaTFPB.12 This reagent is therefore also a very useful anion excluder, as is KTCPB, for the Cur1 membrane electrode.The detection limit is not related to the alkyl chain of the dithiocarbamate groups in the ionophore. Whereas the calibration graphs for the commercial electrode were straight lines between 10-2 and 10-6 M CuCI2, concentrated chloride solutions always pro- duced interference. The pH response profiles of the electrodes were examined using 10-3 M CuC12 solution, adjusted with hydrochloric acid (0.1 M) and sodium hydroxide (0.1 M) as acidic and alkaline media. The electrodes based on o-XBDiBDTC and o-XB- MODTC gave a useful pH range of 3.2-5.5 and of 3.44.1, respectively. Fig. 3 shows the e.m.f. responses of the electrodes at various pH values.The response times of the electrodes were measured by a 10-fold increase in Cu2+ concentration from 10-3 to 10-2 M. The electrodes based on o-XBDiBDTC and o-XBMODTC achieved a steady potential within 9 and 31 s, respectively. TheANALYST, SEPTEMBER 1989, VOL. 114 1031 Table 1. Properties of Cu"-selective membrane electrodes Electrode Commercial electrode" Parameter 1 2 Detectionlimit/M . . . . 1.4 X lop7 3.9 X 3.2 X SlopeImVperdecade ... . . 29 28 29 pHranget . . . . . . . . 3.2-5.5 3.4-6.1 3.4-6.2 Response time/st . . . . 9 31 20 * Orion electrode, Cu2+ 94-29. t Measured with a 10-fold increase in Cu2+ concentration, i.e., $ Measured in 10-3 M CuC12 solution. from 10-3 to 10-2 M . 0 -1 -2 m c - g OY 0 -I -3 -4 -5 Fig. 4. cu2+ -- cu2+ -- C"2+ - - cu2+ -- Na+ K+ Pbz+ Ni2+ Ca2+ Cd2+ Mg2+ Sr2+ M n2+ Zn2+ co2+ 1 2 3 4 Selectivity characteristics for Cu2+ incorporating various metal cations.Electrode construction: 1, o-XBDiBDTC and KTCPB; 2, o-XBMODTC and NaTFPB; 3, o-XBDEDTC; and 4, Orion electrode, Cu2+ 94-29 response time of o-XBDiBDTC is the shortest among the electrodes, probably owing to the rapid rate of complex formation of the ionophore, although we used a different anion excluder. The properties of the electrodes are illustrated by comparing them with those of the commercial electrode in Table 1. The selectivity coefficients of the two ionophores for various cations were evaluated by the mixed solution method and, using similar experimental conditions, they were com- pared with o-XBDEDTC, which also has the same o-xylene backbone with a non-macrocyclic cavity.When the o-xylene backbone of the compounds is kept constant and the alkyl chains of the dithiocarbamate groups are varied, the selectiv- ity patterns for Cu2+ were observed and are illustrated by comparing them with those of the commercial electrode in Fig. 4. It can be seen that the electrode based on the ionophore with isobutyl as an alkyl group has the best selectivity of the three. The electrode based on o-XBDiBDTC is superior to the others, because of its complex-forming ability. The best electrode with o-XI3DiBDTC rejects interference by a factor of 102 for alkali, 103 for alkaline earth and 102-104 for transition metal cations. This electrode shows a large interfer- ence by Pb2+ ion. Also, when comparing the levels of Zn2+ and Mn2+ ions in this electrodc with their levels in the commercial electrode, they are reversed.However, with o-XBDiBDTC, there is no interference from chloride and bromide ions. The selectivity coefficients of strontium halides (log kc?&) for the commercial and trial electrodes were measured by the mixed solution method. The values of log kc::, for the o-XBDiBDTC electrode were -4.2 (chloride) and -3.2 (bromide) when employing 10-1 M strontium halide solutions. On the other hand, the commercial electrode gave values of -3.9 (chloride) using 10-2 M SrCI2 and -0.9 (bromide) using 10-5 M SrBr2 solutions. On the basis of the above results, it is confirmed that the selectivity for a particular ion depends not only on the backbone and donor atoms of the compounds but also on the alkyl chain of the dithiocarbamate group in the ionophore, which affects the non-macrocyclic cavity of the compound.Although the o-XBMODTC ionophore electrode is highly lipophilic, its selectivity is not as good as that of the o-XBDiBDTC electrode. Conclusion Highly sensitive and selective electrodes for Cu2+ based on the ionophores o-XBDiBDTC and o-XBMODTC were devel- oped. The best electrode with o-XBDiBDTC rejected the interference of various cations by a factor in the range 102-104. This high selectivity may be induced by the non-macrocyclic cavity with an appropriate alkyl chain of the dithiocarbamate group. Also, the membrane with a neutral carrier shows no interference from chloride and bromide ions, because there is no Ag2S in the membrane. The authors thank Dr.J. D. R. Thomas of the University of Wales College of Cardiff (UK) for discussions and encourage- ment. This work was supported by a Grant-in-Aid for Scientific Research from the Japanese Ministry of Education, Science and Culture. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. References Petranek, J . , and Ryba, O., Anal. Chim. Acta, 1974, 72, 375. Oehme, M., and Simon, W., Anal. Chim. Acta, 1976, 86, 21. Simon, W., Pretsch, E., Morf, W. E., Ammann, D., Oesch, U., and Dinten, O., Analyst, 1984, 109,207. Nieman, T. A . , and Horvai, G., Anal. Chim. Acta, 1985, 170, 359. Kitazawa, S., Kimura, K., Yano, H., and Shono, T., Analyst, 1985, 110, 295. Jones, T. L., Zimmer, L. L., Diaddaris, L. L., Rorabacher, D. B . , and Ochrymowycz, L. A . , J. Am. Chem. Soc., 1975,97, 7163. Kamata, S . , Yamasaki, K., Higo, M., Bhale, A., and Fuku- naga, Y., Analyst, 1988, 113, 45. Kamata, S., Ogawa, F., and Fukumoto, M., Chem. Lett., 1987, 533. Baiulescu, G. E . , and Cosofret, V. V., Rev. Chim. (Bucharest), 1975, 26, 1015; Chem. Abstr., 1976, 85, 86607t. Baiulescu, G. E . , and Cosofret, V. V., Rev. Chim. (Bucharest), 1976, 27, 158; Chem. Abstr., 1976,85, 136549~. Craggs, A., Moody. G. J . , and Thomas, J. D. R., J. Chem. Educ., 1974,51, 541. Kobayashi, H . , Sonoda, T., and Iwamoto, H., Chem. Lett., 1982, 1189. Paper 8f03255F Received August 9th, 1988 Accepted March 23rd, 1989