3590 © 1900 The Chemical Society of Japan Bull. Chem, Soc. Jon, 68, 3590—3595 (1980) Kinetic Studies of Fast Equilibrium by High-performance Liquid Chromatography I. Ternary Complex Formation of N,N-Disubstituted Dithiocarbamate Chelates of Ni(II) and Cu(II) Masataka Moriyasu* and Yohei Hasuimoro Kobe Women’s College of Pharmacy, Moteyamakite-machi, Higashinado-ku, Kobe 658 (Received May 23, 1980) “The equilibrium of labile ternary complex formation of metal dialkyldthiocarbamate chelates (MAy+ MBs ‘2MAB, K=[MAB]}'/[MA,][MB,]) has been investigated by high-performance liquid chromatography, the separa- tion process of which is 20 fast that kinetically unstable ternary complexes can be eluted with no disproportiona- tion during the course of chromatography. High-performance liquid chromatography makes it possible to halt the progress of considerably fast reactions at the moment when the solution of reaction system is injected into the [Vol. 58, No. 12 column. facilitating the determination of equilibrium constant K “Thus, chromatograms obtained indicate the concentration of each species before chromatography, ‘of Ni(II) and Cu(II) chelates. The equilibrium constants obtained are always 40, suggesting that ternary complex formation is controlled by a statistical factor. Dispropor~ tionation takes place by collision of two ternary metal chelates, the rate constants being of the order I s-! (Ni(II) chelates) and 10® M-t 5"? (Cu(II) chelates). hhave been obtained. Study of fast reactions in solution has drawn the attention of investigators, the flow and relaxation methods being utilized. The conventional method, in which two reactants are mixed and then the change of concentration of reactants or products is measured, was found to be ineffective because of the lack of appropriate means of rapid analysis. High-performance liquid chromatography (HPLC) developed in recent years has been successfully applied to microdetermina- tion of various substances, ‘The merit of rapid separation in HPLC might be hopeful for determining kinetic characteristics of considerably fast reactions in solution. Since the first report on metal acetylacetonates by Huber and Kraak,” only a few reports have appeared ‘on HPLC separation of metal chelates, many of them being on HPLC of metal diethyldithiocarbamates.* We carried out quantitative determination of various metal diethyldithiocarbamates.” This article deals ‘with the applicability of HPLC to kinetical investigation of relatively fast equilibrium of labile ternary complex formation. Experimental Reagents, Commercial sodium salts of diethyldithio- carbamic acid and dibenzyldithiocarbamic acid were used, Sodium salts of other N,N-disubstituted dithiocarbamie acid ‘were prepared by the usual procedure from dialkylamine, carbon disulfide and sodium hydroxide. They were recrystal- lized from either chloroform-methanol or chloroform-hexane, Standard solutions of Ni(II) and Cu(II) were prepared by dissolving NiC-6H,O and CuSO,-3H,O in water. The ‘concentrations of there metal ions were determined by colori- metry as diethyldithiocarbamate chelates. ‘Apparetus, An HPLC apparatus we constructed was used.” The two plunger reciprocal pump (Model KHD-W- 294, Kyowa Seimitsu Co. Ltd.), damper, sample injector (5 ul for variable) and column were combined . A single-beamed spectrometer (Model Spectra 20, Toshiba Beckmann Co. Ltd., 210—700 nm) equipped with a flow cell was used asa detector, Since the detector is not designed for HPLC use, its sensitivity 108 Mt “The activation parameters of ternary complex formation js not so high. The upper limit of rate constants determined by the present method is restricted by the sensitivity of the detector. Silica gel packings (LiChrosorb SI 100, 5) E. Merck Darmstadt and Wako gel LC-SH, 5 yu, Wako Pure Chemicals Co. Lid.) were packed into stainless steel columns (diameter 2 mm, length 25 cm) by a balanced slurry packing technique. A pre-packed silica gel column (Shodex silipak, 5 y, diameter 4mm, length 15 em, Showa Denko Co. Lid.) was used. Absorption spectra of metal chelates were observed with a double-beamed spectrometer (Model 124, Hitachi Co. Ltd). Procedure. n adsorption chromatography, adjustment ‘of column is important as regards the mode of elution and reproducibility, and the water content on silica gel was eon- trolled as follows. Ca. 100.em® of water containing acetone (3%) was pumped through the column. At least 200 em? of water saturated hexane flowed out, the activity of silica gel column being reduced. Reproducible results were obtained when water saturated eluent was used. Before use each organic solvent was saturated with water by shaking the mixture of the solvent and water vigorously, followed by standing for Id. Since itis desirable in some cases to perform ‘chromatography at low temperature, column temperature was kept low by plunging solvent reservoir and column into an ice bath when necessary. ‘Metal chelates were prepared by mixing the corresponding metal ion and sodium dialkyldithiocarbamate. The metal chelates produced were extracted with chloroform under conditions of complete extraction (at pH 4.0 in acetate buffer). In order to avoid any effects of residual free ligands in chloroform, the extract was washed with water four times. ‘The standard solutions of each metal chelate (0.1 to 2mM) thus prepared were used for HPLC after being diluted t0 proper concentration. Measurements of rate constants were carried out as follows, Two dilute solutions of binary com- plexes were mixed under thermostated conditions. After the lapse of a certain time, a portion of the mixture was supplied to HPLC. All the measurements were carried out three times. ‘Theoretical Possible chromatogram patterns of a mixture of woDecember, 1980] different binary complexes, MA, and MB, will be considered first, It is assumed that no decomposition of cach chelate takes place during the course of chromatography. If no interaction exists between the ‘two binary complexes, two peaks corresponding to each complex should appear. However, the chromatogram patterns would be complicated if'a mixed ligand complex is formed in the mixture as follows. MA, + MB, == 2MAB a = (MAB}#[MA,]EMB,] ® Equilibrium is hardly attained when M is a metal fon which forms very inert complexes. In such a case, a ternary complex will be formed when the mixture is left to stand for a long time or brought to elevated temperature. The ternary complex formed can be separated by the conventional method without dispro- portionation. IfM isa metal ion which forms very labile complexes, the teary complex undergoes disproportion into two binary complexes, as soon as it is separated from binary complexes. Studies have been carried out on the equilibrium by mixing two solutions of MA and MB, and observing the resulting remarkable change of physico-chemical properties such as UV absorption. The separation process of HPLLG is so fast that it might compete with the progress of dispropor- tionation in the column. Possible chromatogram Lh at ® { mw DA eld TRA Column, Fig. 1. Possible chromatogram patterns of the system of labile ternary complex formation. patterns are shown in Fig. 1. When the separation process is so fast that disproportionation of the ternary ‘complex during chromatography can be neglected, three peaks corresponding to MA,, MB,, and MAB would appear on the chromatograms (Fig. I(a)), the chromato- grams obtained directly indicating the concentration ‘of each species before chromatography was carried out When disproportionation is much faster than the separation process, peaks of only two binary complexes would appear on the chromatograms (Fig. 1(c)). In case of the intermediate of these two extremes, the peak of a ternary complex would appear accompanied by disproportionation in the column (Fig. 1(b)). In this case disproportionation is retarded with increase in flow rate, bringing about a change of chromatogram ‘Ternary Complex Formation of Labile Dithiocarbamates 3591 patterns. Thus, ifthe peak height of the ternary complex remains unchanged with flow rate, disproportionation is negligible. In HPLC there are some favorable factors to retard disproportionation in the column. (I) Each species is diluted rapidly during the course of chromatography, suppressing disproportionation which is anticipated to ‘oceur by the bimolecular collisional process of two ternary complexes. Actual concentration in the liquid phase will be low since a considerable part of each species is retained on the column, collision of two ternary complexes adsorbed on silica gel surface being unlikely to occur. (2) The rate of disproportionation can be controlled by a suitable choice of initial concentrations ‘of MA, and MB, disproportionation being slow when initial concentrations are low. (3) Control of column temperature is useful, disproportionation being retarded ‘when column temperature is kept low. ‘Thus a combina~ tion of high speed separation, rapid dilution, suitable choice of initial concentrations and low temperature, similar to the case of the rapid quenching method, might halt the progress of the reaction in the column, Ifthese conditions are fulfilled, the equilibrium constant K will be determined by direct measurements of peak heights appearing on chromatograms. Rate constants & and k_ are determined as follows. ‘Two dilute solutions of MA, and MB, are mixed. A ternary complex is produced gradually and the equilib- rium Eq. I will be attained. After the lapse of a certain time a portion of the mixed solution is supplied to HPLC. The progress of the reaction ceases when the solution is injected into the column, ‘Thus, concentra- tions of each species at a definite time after mixing can be determined by HPLC, ‘The rates of formation of ternary complex and decrease of binary complexes are expressed as follows. MAA] _ @IMBAL_ (MAB) — AMAJIMB,] (3a) SIMABI s(Ma,J(MB,] — ¢.[MAB)* (3b) Assuming that the initial concentrations of MA, and MB, are the same (=ay), we have a 2 ket fay alt — = (IK = I)kst — BaghsK + 40,hiK (4a) ® < ios 12 ~ BB = Hog = 712)" = at = WINK 1) — ay Hath (4b) where x and y are concentrations of MA, (or MB,) and MAB, respectively. Rate constants & and k- are determined by means of Eq. 4a or 4b. When K=4.0, we have the following equations by integrating Eqs. 4a and 4b. In (1 — Balog) = hat (5a) = In (1 — ag) = dat (5b) When K-4.0, kand kare determined by means of the following equation.1")3592 ‘Masataka Monivasu and Yohei Hastussoro [Vol. 53, No. 12 = (2)mt) (»(fSBE | ta) | ° m= 40/1] Results and Discussion Determination of Equilibrium Constant. By use of a well deactivated silica gel-column, various metal chelates of N,N-disubstituted dithiocarbamic acid gave chromatograms with no decomposition of chelates Metal chelates of Cu(II), Hg(II), Ni(II), Pa(II), and Co(II) etc. gave linear calibration graphs (detection limit, few ug-few ng). A mixed solution of two different N,N-disubstituted dithiocarbamate chelates of Ni(II) always gave three peaks") (Fig. 2(c)). Peak heights of these three peaks remained unchanged in the flow rate range of the pump 3.00.4 em?Jmin (Figs. 2(c)— 2(f)).. This indicates that the conditions shown in Fig. I(a) are satisfied. When the initial concentrations of two binary complexes were chosen to be the same (=), the equilibrium concentrations of the chelates were both ay/2. ‘Thus the equilibrium concentration of ternary complex was equal to ay. Exactly half of the initial binary complexes was transformed into a ternary complex, the equilibrium constant K being 4.0. Forma- tion of the ternary complex is controlled by a statistical factor. In all the systems of a mixture of two binary complexes K was always equal to 4.0, suggesting that Fig.2. Chromatogram patterns of two different dic thiocarbamate chelates of nickel. Column: Shodex silipak (4mm>%15 em), eluent hexane : ethyl acctate= 100 : 7 (water saturated), flow rate: (a)—(c) 2.5 emY/min; (A) 3.0 em*/min; (¢) 1.5 min; (f) Ob em?/min, detector: 325 nm, sample: (@) 0.4 mM_Nil(CH,CH,CH,),NCSS), in’ CHC): (b) 04 mM Ni[(CH,CH,),NCSS}, in CHCl; (c) 04 mM _Nil(CH,CH,CH,),NCSS]__ “+0. mM Nif(CHy- CH,),NCSS), in CHCl, sample size: 5 ul (A): Nil(CH,CH,CH,).NCSS]., (B): Nif(CH,CH)- NSS], (C): Nif(CH,CH,CH,),NCSS][(CH,CH,),- NSS} a]o oF the the lack of the factor either stabilizes of destabi ternary complex.) Cu(II) chelates showed different chromatogram @ wip le oat ORE OR ete ae ae Fig. 3. jomatogram patterns of two different die thiocarbamate chelates of copper at relatively higher Column: LiChrosorb SI 100 (2 mm>25 em), eluent: hhexane : cyclohexane : isopropyl acetate=50 : 50:4 (water saturated), flow rate: (a) 1.3 em4/min; (b) 0.9 m/min; (c) 0.5 cm?/min; (d) 0.2 em¥/min, detector 440 nm, sample: 1 mM Cul(CH,CH,CH,),NCSS),+ mM Cu{(CH,CH,),NCSS}, in CHCl, sample size 5 (A): Cul(CH,CH,CH,)NCSS},, (B): Cuf(CH,CH,) NCSS},, (C)? Cuf(CH,CH,CH,),NCSS]I(CH,CH,),- CSS}. The broadending of this peale (C) is due to disproportionation of this ternary chelate in the column, @® © @ © | il L all OOF OTL OF FFTE ig. 4. Chromatogram pattems of two different di- thiocarbamate chelates of copper at low concentrations. Column: Shodex silipak (4mm 15 em), eluent: hhexane ‘ethyl acctate=100 : 15. (water saturated), flow rate: (a)—(e) 25 em/min; (d) 3.0¢m4/min; (@) 1.5 emi)min, detector: 280 nm, sample: (a) 4.0% 10-* M Cul(CH,CH,),NCSS},; (b) 4.0% 10- M Cu- [(CH,)NCSS},;" (6). 40% 10 M_ Cuf(CH,CH, NCS}, 4.0% 10-* Gul(CH,),NCSS},; these samples are solutions in hexane : chloroform : ethyl acetat 8: 1:1, sample size: 50 (A): Cu[(CH,CH,),NCSS},, (B): Cuf(CH,),NCSS]., (©): Cu[(CH,CH,).NOSS}[(CH,) NCSS). No disproportionation was observed in this case.December, 1980) ® ®» © @ @ © cl ORE OTE OTE ORT OTTO Fig. 5. Change of chromatogram patterns of the mixture ‘of to different dithiocarbamate chelates of nickel. Column: Shodex silipak (4mm>15 em), eluent hexane : ethyl acetate= 100 :20 (water saturated), low rate:2.5 em#/min, detector: 325 nm, sample: 2.5% 10-5 M_Nif(CH,CH,)_NCSS),-+2.5 10-* M_Nif(CH,).° NCSS}, (a) After mixing 6.5 min; (b) 4.0 min; (e) 7.5 ‘in; (4) 11.0 min; (€) 18.0 min; (f) 60.0 min (eq librium) ; these samples are solutions in hexane: chloro- form: ethyl acetate=3 : 1:1 at 25°C, sample size 50h. (A): Nif(CH,CH,),NCSS},,_(B): Nif(CH,),NCSS],, (C):Ni{(CH,CH,),NCSSI[(CH,}.NCSS}. patterns, When a relatively high aj was chosen (a= 1.010"! M), chromatogram patterns were seriously affected by the flow rate, suggesting the disproportiona~ tion of the ternary complex in the column. On the other hand, when a very low ay was chosen (= 4.0x10-* M), they were unaffected by the flow rate (from 1.5 to 3.0 cm}/min) (Fig. 4). In order to compen- ‘Ternary Complex Formation of Labile Dithiocarbamates 3598 sate the decrease of concentration of the sample, it was necessary to inject a large volume of sample. The composition of the solvent of the injected sample should be similar to that of the eluent, otherwise serious broadening of each peak takes place. About 200 ul of sample volume could be injected into the column without remarkable broadening of each peak. For Cu(II) chelates the chromatogram patterns were also sensitive to column temperature. At ambient tempera- ture, when a, was chosen 4.0% 10-* M, chromatogram petterns were affected slightly by ow rate. However, ‘when column temperature was low, no disproportiona- tion in the column was observed. The values K were also always equal to 4.0 forall the systems of Cu(II) chelates. Determination of Rete Constants. As shown in Fig. 5, a trace peak of the ternary complex appeared on chromatograms immediately after mixing of two solutions. With the lapse of time the peak ofthe ternary complex rose while the peaks of binary complexes fell ‘The rate constants & and k were determined by means of Eq. 5b. The results are summarized in Tables 1 and 2 (for Ni(II) and Gu(It) chelates, respectively). ‘The values of k and k- were sensitive to temperature and composition of solvent of the mixture. Since & and k- "were affected by the water content ofthe mixed solution, experimental conditions were so chosen that each solution of MA, and MB, was saturated with water before mixing." Disproportionation was slow when solvent was not saturated with water. Column tempera- ture and composition of eluent play insignificant role By use of HPLC we can promptly separate various species including kinetically unstable ones and elute them. Varying chromatographic conditions such as composition of eluent, column temperature and flow rate dors not change the values of & and kif dispropor- tionation can be neglected during the course of chromatography. The rate constants k of Ni(II) chelates are of the order 10.10! M-15-1. Since Cu(II) chelates are more labile than Ni(II) chelates, values off of Cu(II) chelates are larger and of the order 10M-!s-. Values of & seems (© be small when molecular weights of two ‘Tapur 1. RATE CONSTANTS OF FORMATION AND DISPROPORTIONATION OF LADILE “TERNARY prriiocaRsaMare.cuELATES oF Ni(IT) NiA, + NiB, = 2NiAB, K = [NiAB}/INIAJINB, Retction sme Gn) (tts) ats) MA; MAB MB, (GH) NCSS T3L0.DxI GSLOAxI TR RRS. ~ (CH,CH,CH,),NCss 2440.2) x10" (6.00.6) x10 3.927 (CH) .NCsS (140.07) x108(2.840.2)x100 3.142. (CH,)gNcSS (9.740.5)x10! Q440.15)x108 44 51S. (GyH,CH,),NCSS (1.940.2)x 10" (4.840.5)x10" 3.9 2B (CH,CH,OCH,CH,)NCSS—_(7.340.5)x101(1-B40.1)x10\ 224 ‘A: (GH,GH),NCSS, a5=2,5% 10M, Chromatographic conditions; column: Shodex slipak (4mm x 15 em), cluent 2.5 em?/min, detector: 325 nm, sample si were mixed at 25°C. hexane: ethyl acetate (water saturated)~ 100: 20 (A), 100 : 8 (B), 100 : 6 (C), 100 : 5 (D), flow rate 50 ul, column temperature: ambient, Equal iB, dissolved in the mixed solvent system, chloroform : ethyl acetate : hexane=:3594 Masataka Mortvasu and Yohei Hasumoro [Vol. 53, No. 12 “Tanus 2. Rare coNsrAvr OF FORMATION AND DPROPORTIONATION OF LABILE ‘TERNARY DITHIOCARBAMATE CHELATES OF Cu(II) Cun, + CuB, Ae 264AB, (CuAB}/[CuA,}(CuB,] =h/A_=4.0 " 7; Recon tne on) rte ort) MA,__MAB_ MB, . (CH,),NCSS_ (3,540.4) x 105 (8.841.0) x 108 15 2.0 28 A (CH,CH,CH,),NCSS (6.40.8) 108 (1.640.2)x10 3.0 QT 13. B (cu Ness (1:24.0.05)x10 (3.040.15)x10@ 3.7 #3 SG (GyH,CH,) NCSS (440.3)x10" (85208)xI0 302518 ‘A: (GH,CH,),NCSS, @,=4.0x 10M, Ghromatographie conditions: column: Shodex silipak (4mm x 15 em), ‘eluent: hexane : ethyl acetate (water saturated)=100 : detector: 270 nm, sample size: 50 ul, column temperatur Equal amounts of CuA, and CuB, dissolved in the mixed solvent system, chloroform : ethyl TI (water saturated), were mixed at 25 °C. into an ice bath). acetate : hexane Fig. 6, Plot of Eq. 5a at different ay Chromatographie conditions: column: Shodex slipak (4mmx 15m), eluent: hexane : ethyl _acetate= 100:7 (water saturated), flow rate: 2.5 em?/min, detector: 325 nm, sample: the mixed solution of Ni- {(CHJCH,CH,).NCSS}, and Nif(CH,CH,)NCSS}, in hexane : chloroform : ethyl acetate subjected to HPLC after mixing adefint r 0x 10M, (B) a 6.25% 10-* M, ‘Tate 3. AcHIVATION PARAMETERS OF TERNARY COMPLEX FORMATION OF N,N-DISUBSTITUTED M[(CH,CH,),NCSS}, + M[(CH,CH,GH,).NCSS), == M[(CH,CH,) .NGSS] {(CH,CH,CH,).NCSS] a He @ @5-G) = (kJ mol) (kJ mol) _ (kJ Kt mol!) Nia” rs 6043 (1.722.) x10 Gum M5 S$ (9.13.0) x 10% a) In chloroform : ethylacetate hexane=1: 1:3 (water saturated). b) In chloroform : ethyl acetate: hexane=I :1 :8 (water saturated). dithiocarbamates are close to each other, Validity of the bimolecular collisional process of disproportiona- tion was confirmed by changing a) (Fig. 6). When ay exceeds this range, the rate of disproportionation does not increase in proportion to 15 (A), 100 :4 (B), 100 :3(C), flow rate: 5 em/min, ‘0°C (Column and solvent reservoir were plunged ‘The activation parameters of disproportionation was determined by measuring & values at various tempera- tures. This was carried out by mixing two solutions of MA, and MB, under thermostated conditions at various temperatures. Results of the system consisting of diethyldithiocarbamic acid, dipropyldithiocarbamic and metal ion are given in Table 3. The activation enthalpy of Cu(II) chelates is smaller than that of Ni(II) chelates, Conclusion Fast bimolecular reactions can be traced by HPLC. Chromatographic conditions should be so chosen that no decomposition of species occurs during the course of chromatography. This can be done to a certain extent by choice of column packings and solvent systems. ‘The upper limit of the rate constants is restricted by the sensitivity of the detector. Absorption of 0.05 units could be extended to a full scale without serious noise with the spectrometer we used. It’seems that rate constants of the order 10'M-!s-! can be determined. Since the sensitivity of a commercial UV detector for HPLC is about ten times larger than that of the present one, fast reactions with rate constants of the order 10° M-s-t might be traced by HPLC if absorption coefficient of the detected species is large (log e>4). ‘The present method will also be useful for investigation of physicochemical properties of kinetically unstable species in solution. References 1) J. F.K, Huber and J. ©. Kraak, Anal. Qh. 44, 1554 agi} 2) P, Helzmann and K. Ballchmiter, J. Chomat 153 (1977). 3). P. G. Uden and I. E, Bigley, Anal. Ghim. Acta, 94, 29 (97, 4) G, Schweds, Chromatographia, 1, 145 (1978) 8) G. Schwedt, Chromatgraphia, 12, 289 (1979), 8) J. W. O'Laughlin and T. P, O'Brien, Anal. Lt, AM, 829 (1978). 7) M. Moriyasu and Y. Hashimoto, Anal. Lett, AM, 598 1978) 137,December, 1980] ‘Ternary Complex Formation of Labile Dithiocarbamates 3595, 8) O, Liska, J. Lehotay, E, Brandstererova, G. Guichon, 12) M. Bodenstein, Z, Pls. Ghem., 13, 56 (1894); 22, 1 and H. Colin, J. Chromatogr, 172, 384 (1979). (1897); 28, 295 (1898). '9) J. Lehotay, E. Brandstererova, O. Liska, and G. 13) -M. Moriyasu and Y. Hashimoto, Qhem. Lett, 1980, 117. Guichon, J. Chromatogr, 172, 379 (1979). 14) |. Liska, G. Guichon and H. Colin, J. Chromatagr., 171 10) R.F. Majors, Anal Chem., 44, 1722 (1972) 145 (1979). 11) RG. Bray, Biochem. J., 81, 189 (1961). 15) H. Sioge, Angew. Chem. Int. Ed. Engl, 14, 397 (1975).