Scientific paper
Isopolyniobotungstate HxNb2W4Ol4-x)- Ions: Analysis of the State of the Ions in Aqueous Solutions, Formation Constants Calculation and Thallium
Salts Synthesis
Svetlana M. Vavilova,1 Maksym A. Kryuchkov,2 Katerina E. Belousova1 and Georgiy M. Rozantsev1
1 Department of Inorganic chemistry, Faculty of Chemistry, Donetsk National University, 24 Universitetskaya Str.,
Donetsk 83001, Ukraine
2 Department of Chemistry, University of Montreal, Montreal H3T 2BI, Canada
* Corresponding author: E-mail: razantsev @dongu.donetsk.ua
Received: 28-08-2009
Abstract
By means of pH-potentiometric titration, the processes of the complexes formation in the system Nb6O198--WO42--H+-H2O with CNb : CW = 2 : 4 was studied at different Nb + W concentrations. Exerimental data, being processed by mathematical modeling, allowed to obtain the distribution diagrams of individual niobium and tungsten isopoly anions, and mixed isopolyniobotungstates in the range of Z = CH+/CNb+W = 0 - 2.0 (background electrolyte is NaCl). Concentrational and thermodynamic formation constants were calculated using quasi-Neuton method (CLINP 2.1 software) and it was shown, that the formation of isopolyniobotungstates (HxNb2W4O(4-x)-, x = 0 - 2) of the 6th row of Periodic table proceeds through intermediate Nb3W3O195- ion formation. Thallium salts Tl3HNb2W4O19 • 10H2O and Tl2H2Nb2W4O19 • 10H2O were isolated and characterized by elemental and EDX spectral analysis, electron microscopy and FTIR-spectroscopy.
Keywords: Polyoxometalates, isopolyniobotungstate, pH-complexonometric titration, ionic equilibrium.
1. Introduction
Isopolyniobotungstate anions (IPNTA) HxNb2W4 Oj49-x)-, x = 0 - 2 have proven potential in prevention of cellular adsorption of viruses and viral bodies. Thus, IPNTA can serve as a viral fusion inhibiting component of complex anti-virus, anti-tumor and HIV-treatment pharmaceuticals.1 Besides, they can form stable coordination compounds with transition metals, thus allowing to use them and their derivatives as the ligands in organome-tallic complexes.2,3
But the question of namely HxNb2W4O((9x)-, x = 0 - 2 ions formation in solution is still actual. The most common procedure, that was developed by Dabbabi and Bo-yer,4 and still used nowadays, assumes the preparation of the mixed Na-K-NBu4 salts with Nb2W6-nO(29+n)-, n = 1 - 4
anions by mixing of sodium tungstate and an excess of potassium perniobate, followed by acidification of the obtained solution and, finally, the desired product was precipitated by tetrabutylammonium bromide. The authors indicate, that this approach leads to the product, being contaminated by the salts with anions possessing other ratio between Nb and W. This is because the pH regions of different IPNTA domination are overlapped at that conditions.
To avoid the formation of such an impurities, firstly, the synthesis must be carried out at the fixed acidity (Z) values. Secondly, only ortho- and hexatungstates of alkali metals should be used, avoiding any use of peroxides or other auxiliary compounds. And thirdly, only precisely stoichiometric ratio between Nb and W will ensure the formation of the target product.
Two latter issues are easy to solve, but for the first issue one needs to know the acidity regions of predominant formation of IPNTA with the desired composition. To define the correct acidity regions, the state of the ions in aqueous solution must be studied. To the best of our knowledge, these studies were not carried out till now, what can be explained by enormous number of consequti-vely-parallel reactions, occurring in such systems, and by imperfectness of mathematical methods at that time, that did not allow to estimate qualitative and quantitative composition of the anions in solution. Modern methods of mathematical modeling (MMM) make it possible to solve such problems accurately by interpretation of the experimental data obtained from investigation of interactions in aqueous solutions of IPNTA.
In the present work we studied the formation of the complexes in the system Nb6O89-- WO4-- H+- H2O (CNb : CW = 2 : 4) by pH-potentiometric titration and the regions of existence for the forming IPNTA were determined. These allowed us to prepare the compounds with Nb : W = 2 : 4, that does not contain the impurities of the salts with other IPNTA. This was achieved by using the solutions of sodium orthotungstate, potassium hexaniobate and hydrochloric acid. Moreover, no hydrogen peroxide has been used.
2. Experimental
2. 1. Solutions Preparation
The initial solutions were prepared from solids or from concentrated solutions by diluting with distilled water, purified from CO2. Potassium hexaniobate solution was prepared by dissolving of freshly prepared salt 4K2O ■ 3Nb2O5 ■ 12H2O. It was prepared by annealing of Nb2O5 with 5-fold excess of KOH, followed by thorough washing with the distilled water and double recrystallization from acetone. Sodium tungstate and sodium chloride solutions were prepared by dissolving of solids in water, and the solution of hydrochloric acid was prepared by diluting of the concentrated (10M) solution. The precise concentrations of initial solutions were determined according to chemical analysis data: the contents of tungsten and niobium were determined gravimetrically (gravimetric forms WO3 and Nb2O5, S< ± 0.5%), hydrochloric acid was standardized by titration of Na2B4O7 ■ 10H2O with methyl red indicator, 8< ± 0.8%.
2. 2. Complexonometric Titration
The studies of the complexes formation in the system Nb6O189-WO4-H+-H2O were carried out by pH-complexonometric titration at 25 ± 0.1°C, using I-500 (Aquilon, Russia) ionometer. Indicator electrode was hydrogen-ion selective glass electrode ESL 63-07 Sr (Belarus) with isopotential point pH, = 7.00 and A, = -25 ± 10
mV, auxiliary electrode EVL-1M3 was silver chloride electrode (Ag/AgCl, sol. KCl, saturated) with the potential of 202 ± 2 mV, according to standart hydrogen electrode. Calibration and preciseness of the readings were controlled by the series of standard buffer solutions, prepared according to Bates.5 In the systems under investigation the overall concentrations of Nb + W (C^+W were 10, 5, 2.5 and 1 mmol/L (mM) and CNb : CW = 2 : 4.
The acidity of the systems during titration Z = CH+/ C0Nb+W (CH0 + is the overall concentration of acid and CN0 b+W is the overall concentration of niobium and tungsten in solution) was controlled by the amount of the acid being added with step AZ = 0.02 within the interval Z = 0 - 2. The ionic strength was created by the background electrolyte (NaCl) and was varied within I = 0.01 - 1.00.
2. 3. Experimental data treatment
For interpretation of experimental data, the mathematical modeling, using CLINP 2.16 software, was utilized. Each model was evaluated for consistency with the experimental data. The models were checked for adequa-teness and excessness, and the main criterium of calculation results coherence with the experimental data was the value of F, that is the sum of squares of deviations between calculated and experimental values of pH along the entire titration curve:
F =
i
k I
(ApH{T= I (pHf'-pH™P)2 /=1 i=l
where k is the number of points in the titration curve. It is worth to notice, that the titration considered to be successful, if any point was within |ApHJ < 0.12.
The result of mathematical modeling was the determination of concentrational formation constants (lgKC) for the anions in solution. Based on the obtained values, thermodynamic formation constants lgK° for individual IPNTA were calculated by Pitzer method.7-9 It is based on the Debai-Huckel equation (Equation 1), that has been expanded by the insertion of ^.-coefficients.8 They allow to consider the influence of different kinds of ions on each other (Equations 2 and 3):


l + b-JÏ b Jl
(1)
(2)
(3)
For equations (1) - (3): Av is Debai-Huckel coefficient for osmotic function A = 0.3921 at T = 298 K); b is
the parameter for Pitzer model; I is the ionic strength of
solution; m. and m, are the molal concentrations of the
j k
background electrolyte (j denotes cation and k denotes anion); NK and NA are the numbers of cation and anion correspondingly; N is the overall number of background ions; zl is the charge of /-ions; Xa and ^^ are the calculated values that include the parameters describing the interaction between ions.7
Calculation of the formation constants allowed to build the distribution diagrams and to determine the regions of predominant formation of the desired IPNTA. These gave a background to create the procedure for thallium salt synthesis. Furthermore, in these regions at the fixed Z and CNb : CW values, the solid phase was precipitated. This phase either was the individual salt with HxNb2W4O(4-x)- anion, or was the mixture of salts with two different x values.
2. 4. Thallium Salts Synthesis
33.7 mL of potassium hexaniobate solution (CNb = 99 mM) was diluted with 955.5 mL of H2O and 11.0mL of sodium orthotungstate solution (CW = 607 mM) were added dropwise at room temperature with stirring to obtain the solution with CNb : CW = 2 : 4. Thus, the set of six solutions was prepared.
Three of the obtained solutions were acidified by the aqueous hydrochloric acid (CHCl = 464 mM) till the desired Z was reached (Step 1 in Table 1) and, 1.5-fold excess of thallium (I) nitrate solution (CT = 330 mM) was added dropwise at vigorous stirring (Step 2 in Table 1).
tion, washed with cold water and dried until constant mass in open air.
The composition of the solid phases were determined by various means of chemical analysis, following the procedure described below. A 500 mg sample of each thallium salt obtained was treated with the 3:1 v/v mixture of HNO3 (70%) and HCl (36%) during 2 hours at stirring. The solid residue, that is Nb2O5 ■ nH2O and WO3 ■ H2O, was filtered, and all the thallium remained in solution. After the solid was annealed for 2 h at 800 °C, the obtained mixture of Nb2O5 and WO3, was heated with (NH4)2SO4 in concentrated sulfuric acid (98% w/w). The obtained melt was then dissolved in 2% aqueous solution of EDTA and pH was brought to 8.0 by aqueous ammonia. The suspension was brought to boil and filtered hot after 30 min. Finally, the isolated Nb2O5 ■ nH2O was annealed at 900 °C until constant mass. Gravimetric form - Nb2O5, 8 < ± 0.5%. The weight of WO3 was found as a difference between the mass of the solid after the first annealing and the mass of Nb2O5.
The contents of thallium in the filtrate were determined by the reverse complexonometric titration of the excess of Trilon B by the standard (25 mM) solution of Zn-Cl2 using xylynol orange as an indicator, 8 < ± 0.8%. The contents of crystallization water were determined by annealing of the sample at 500 °C until constant mass (8 < ± 0.5%).
The anion identification was based on Fourier-transform infra-red (FTIR) and energy-dispersion X-Ray (EDX) spectroscopic data. FTIR spectra of thallium salts were obtained in KBr pellets, using Thermo Nicolet IR
Table 1. Amounts of solutions, used in thallium salts synthesis.
	Reagent Z	1.13	1.31	1.60
Step 1	V HCl' mL	24.3	28.2	34.5
Step 2	Vjp mL	30.3	22.7	15.2
Step 3	V buffer' mL	5.0 (buffer 1)	8.0 (buffer 2)	5.0 (buffer 3)
Another three solutions were acidified in the same way, but then the acetate buffer solutions (Table 2) were added (Step 3 in Table 1), followed by thallium nitrate addition (Step 2 in Table 1). The buffer solutions were prepared using 2M solution of sodium acetate and 2M solution of acetic acid.
The obtained heterogenous mixtures were stirred for 5 h, then the white precipitation was separated by filtra-
Table 2. Acetate buffer solutions preparation.
Buffer	V (CH3COONa), mL	V (CH3COOH), mL	pH
1	5.0	1.0	5.46
2	1.0	7.4	3.59
3	0.1	20.8	2.36
300 spectrometer within the range of 400-4000 cm1. EDX analysis was conducted by means of JSM 6490 LV scanning electron microscope (SEM) instrument, using aluminium stand and the carbon film as a support.
3. Results and Discussion
For the system Nb6O89--WO42--H+-H2O with CNb+W = 10, 5, 2.5, 1 mM and CNb : C^ = 2 : 4, pH-potentiometric titration was carried out at the ionic strengths I lying within the range of 0.01 -f 1.00 and created by NaCl. It turned out that the general titration curves behaviour does not depend on the ionic strength at given CN0 b+W. Therefore, it is enough to have just one typical curve for each concentration CNb+W (Figure 1). Moreover, in case of CNb+W > 5 mM,
the ionic strength needed is one order of magnitude higher than CNb+w.
All the titration curves contain two characteristic pH gaps, that correspond to the protonation with subsequent polycondensation of the initial tungsten and niobium containing anions. It is worth to notice, that, with the decrease of CNb+W, pH gaps become less vivid, especially the one, corresponding to the higher Z in solution. The analysis of Ztheor of different polyanionic forms of tungsten and niobium formation allows to assume, that the first gap (Z < 0.3) may correspond to the protonation of the initial
Nb6Oi89-:
mH+ + Nb6O189 o H Nb6O,(9-m)-
6 19	m 6 19
Figure 1. pH-potentiometric titration curves: a) CNb+w = 10 mM, I = 500; b) CNb+w = 5 mM, I = 140; c) CNb+w = 2.5 mM, I = 50; d) CNb+w = 1 mM, I = 50.
The second gap (Z > 0.6) may correspond either to the polycondensation of the initial wO4- into w6O20(OH)2-, w12O40(OH)10- and Hw7O5-, or to the formation of the mixed IPNTA with the niobium contents being higher than that, determined by the initial feed CNb :
* = 2 - 4 : ^	^
~ VxNbnWb_nO\-

H->0,n< 3
In this case one might expect that the processes of protonation take place consequently and can be easily assigned. whatsoever, the processes of tungstate anions polycondensations are more complicated and, according to the recent data,10 proceed through consequtive-parallel schemes. It is also possible, that consequtive-parallel hypothesis can be applied to IPNTA, especially because in parallel reaction there forms a complex with higher nio-
bium contents than expected, as it was observed in similar tungsten-niobium systems with CNb : CW = 1 : 5.11
Such assumptions inambigously indicate the problem of experimental data interpretation and reveal the impossibility of truthful explanation of gaps in the pH = f(Z) curves, if based only on the theoretical values of Z for both individual and mixed isopoly anions. Indeed, it is impossible to build the schemes of ions interchange and to calculate the formation constants of IPNTA. Thus, to treat the experimental data, the method of mathematical modeling was used. This allows to find the models that adequately describe the complexation processes in the studied systems.
The modeling of these systems was started from assuming the presence of only mixed isopoly anions HxNb2W4O(4-x)-, x = 0 - 2, with the ratio between niobium and tungsten being determined by the initial solutions. As a result, the proposed model (Model 1) contained only three anions (Nb2W4O49-, HNb2W4O39- and H2Nb2W4O29) and could satisfactorily describe (with ApH < 0.12) the experimental data for Z > 1.6 (Figure 2a), and thus was discarded.
Then, individual IPNTA W6O20(OH)2- and HW7O^-were added to Model 1, and the Model 2 was created. This allowed to expand the adequacy region for calculated pH values from Z > 1.6 to Z > 1.2 (Figure 2b). So, Model 2 describes well the processes at high Z, but at low Z < 1.2 the error exceeds the accepted value (ApH < 0.12). If other tungsten anions, such as W7O264, W12O40(OH)1°-, W12O38(OH)2-, W10O342 were used in Model 2, it did not change the region of experimental and calculated pH values coincidence, but negatively affected other model parameters, clearly indicating the excessiveness of the proposed model and the subsequent excluding latter mentioned particles.
When the mixed isopoly anion Nb3W3O59- was added to the Model 2, the Model 3 was obtained. This anion seemed to be the only one to help eliminate uncertainity in the region of the middle Z values. The results (Figure 2c) showed that satisfactory coincidence of experimental and calculated pH values was observed in relatively wide region Z > 0.6. But, at lower Z values this model was still not correspondent to the real process.
Finally, when protonated anions of niobium, HNb6O/9 and H2Nb6O69-, were added to the Model 3 to obtain Model 4, it correctly described the processes of complex formation in the entire studied Z interval, within the error of ApH < 0.12 (Figure 2d).
Aside of the models described, we tried to apply the series of more complicated anions, including other IPN-TA, but all of them proved to be invaluable and/or excessive. Indeed, all subsequent calculations at all the concentrations and ionic strengths used, the Model 4 was chosen as a reference.
From one side, the chemical nature of this model includes well-studied protonation processes of hexaniobo-
pH	a	pH	b
-1-1-1-1-.-1-.-1-.-1-1-1-p-1-1-1-1-1-p--■•—'—1--1--1-i-1-1--1—i-1-r [----1-.-1--1--1-•—
o 0,2 0,4 0,6 0,3 1,0 1,2 1,4 1,0 1,8 2,0	0 0,2 0,4 0,6 0,8 1.0 1,2 1,4 1,6 1.8 2,0
Figure 2. Step-by-step mathematic modeling of the system Nb6O189-WO2 -H+-H2O with CNb+W = 10 mM, experimental (dots) and calculated (solid line) curves: a) - Model 1: ions Nb2W4O49-, HNb2W4O13;- and H2Nb2W4O29-; b) - Model 2: ions Nb2W4O14-, HNb2W4Oj9-, H2Nb2W4O29-, HW7O24-and W6O20(OH)6-; c) - Model 3: ions Nb2W4Oj49, HNb^Oj3-, H2Nb2W4Oj29, Nb3W3O59-, HW7O254- and W6O20(OH)6-; d) - Model 4: ions Nb2W4O49, HNb2W4Of9-, H2Nb2W4O29-, Nb3W3O59-, HW7O254-, W6O20(OH)2-, HNb6O/9- and H2Nb6O16-.
tungstate-ions and polycondensation of orthotungstate-ions, with the known formation constants:1213
Nb6Ol9- + 12WO2- + 20H+ o 3Nb2W4O149- + 10H2O
Z = 1.11
Nb6Ol9- + H+ o HNb6O/9
Z = 0.056 (lgKc = 11.90)
Nb6Ol9- + 12WO2- + 23H+ o 3HNb2W4O39- + 10H2O
139-
Z = 1.28
Nb6Ol9- + 2H+ o H2Nb6O69-
Z = 0.11 (lgKC = 22.90)
Nb6O18- + 12WO2- + 26H+ o 3H2Nb2W4O129- + 10H2O Z = 1.44
6WO4- + 6H+ o W6O20(OH)26- + 2H2O Z = 0.67
(lgKc = 50.41)
Nb6O89- + 6WO4- + 10H+ o 2Nb3W3O159" + 5H2O Z = 0.56
7WO4 + 9H+ o HW7O5 + 4H2O	Z = 0.86.
(lgKc = 70.70)
From the other side, the model contains the equilibrium processes of IPNTA formation, whose constants were unknown till now:
The formation constants for isopolytungstates and isopolyniobates were introduced into the model as fixed values, and the average values of formation constants for the mixed IPNTA were calculated during modeling for all the concentrations at the correspondent ionic strengths (Table 2).
The complete set of concentrational formation constants allowed us to calculate the concentrations of anions and to build the distribution diagrams for the ionic forms (mol % a as a function of acidity Z) at CNb+W = 10, 5, 2.5, 1 mM and varying ionic strengths. Since the diagrams for different ionic strengths do not change for particular concentration, we illustrated the general behaviour at one ionic strength for simplicity (Figure 3).
me the existence of the variety of the protonated forms of both initial Nb6O89- and final Nb2W4O14- (HxNb2W4O(4-x)-, x = 0 - 2) anions, as it follows from mathematical model. Lowering the concentration decreases the degree of protonation of the final anion (HxNb2W4O1(49-x)-, x = 0 - 1) and the protonated forms appear only at higher values of ionic strength. In the diluted systems, where C0[b+W = 1 mM, isopoly forms of niobium and tungsten interact with each
Table 2. Average values of concentrational constants logarithms for the mixed IPNTA at different ionic strengths. Values in brackets denote RMS deviation.
Ionic Strength,	Concentrational constants, lgKC (S)
I, mol/L	Nb2W4Ou4-	HNb2W4O193-	H2Nb2W4Oi92-	Nb3W3Oi95-
0,01	51,96 (0,14)	-	-	-
0,02	51,98 (0,13)	-	-	-
0,03	51,85 (0,12)	-	-	41,16 (0,08)
0,04	51,83 (0,13)	55,93 (0,13)	-	40,63 (0,14)
0,05	51,44 (0,10)	-	-	-
0,06	51,59 (0,11)	55,60 (0,13)	-	40,08 (0,18)
0,07	51,16 (0,13)	55,94 (0,10)	-	40,68 (0,12)
0,08	51,54 (0,11)	-	-	40,92 (0,12)
0,09	-	-	-	40,17 (0,18)
0,10	52,44 (0,12)	56,05 (0,18)	-	40,51 (0,16)
0,12	-	-	-	41,99 (0,17)
0,14	52,60 (0,15)	55,98 (0,16)	-	43,35 (0,16)
0,16	53,05 (0,14)	56,62 (0,14)	-	43,76 (0,14)
0,18	52,57 (0,17)	55,85 (0,17)	-	43,72 (0,15)
0,20	54,52 (0,11)	59,16 (0,11)	-	43,80 (0,15)
0,30	54,81 (0,10)	59,61 (0,10)	-	43,87 (0,14)
0,40	54,60 (0,10)	-	61,63 (0,16)	43,93 (0,12)
0,50	54,69 (0,12)	-	62,04 (0,20)	43,94 (0,15)
0,60	54,76 (0,12)	-	62,59 (0,16)	44,06 (0,14)
0,80	54,33 (0,12)	-	61,39 (0,23)	43,91 (0,14)
1,00	54,29 (0,12)	-	61,50 (0,20)	43,79 (0,15)
According to the distribution diagrams, the region of the first gap correspond to the protonation process of the initial Nb6O89, that totally agrees with Ztheor. In parallel to this process, the polycondensation of the initial WO42- takes place, that leads to W6O20(OH)26- formation. As soon as the tungstate-anions, that possess the same as Nb6Of9 oxygen coordination of the metal, appear in solution, the formation of IPNTA, i.e. Nb3W3O159-, begins. The deficit of tungsten in this form, if compared with the most expected Nb2W4O149- could be possibly explained by insufficient tungsten isopoly anions formation at lower Z. Notably, when reaching Z > 0.8 (Ztheor = 0.86), in parallel with the above mentioned processes, HW7O254- appears in the solution. Along with W6O20(OH)6- and Nb3W3Of9-, it takes part in the formation of the expected Nb2W4O149-. The latter, as the acidity of the system grows, undergoes step-by-step protonation to give HxNb2W4 O(4-x)-, x = 0 - 2.
Even though the processes of complexes formation in the system Nb6Of9--WO42-- H+- H2O at different CNb+W are general, there exist certain peculiarities, characteristic for each concentration. Thus, at CNb+W = 10 mM we presu-
other, giving Nb2W4Of9 without intermediate formation of Nb3W3O59-.
The average values of concentrational constants (Table 2), calculated at different concentrations and the same ionic strength, allowed to calculate thermodynamic formation constants of mixed IPNTA (Table 3). The calculations were carried out by Pitzer method using extrapolation of lgKC dependence to the zero value of ionic strength.
Following the distribution diagrams for the anions in the studied systems at CNb+W = 10 mM, we isolated the solid phases in the regions of predominant formation of HxNb2W4O(9x)-, x = 0 - 2 ions (Table 4).
Table 3. Thermodynamic constants of the mixed IPNTA formation.
Anion	Thermodynamic constants, IgK
NbWPit	54.39 ± 0.24
HNb2W4O193-	60.83 ± 0.68
H2Nb2W4O192-	67.04 ± 0.67
Nb3W3O195-	41.68 ± 0.42
Figure 3. Diagrams of ions distribution in the system Nb6Oj89-WO4 -H+-H2O at different concentrations: a) C°Nb+W = 10, I = 500; b) CNb+W = 5, I = 140; c) CN,+W = 2.5, I = 50; d) CN+W = 1, I = 50. In these charts: 1 - WO;;-, 2 - Nb2W4O49-, 3 - HNb2W4Of9-, 4 - H2Nb2W4O129-, 5 - Nb3W3O59-, 6 - W6O20(OH)2-, 7 - HW7O254-, 8 - HNb6Oj7-, 9 - H2Nb6Oj69-.
In case of H2Nb2W4O129- and HNb2W4O139- there is a substantial overlapping of the regions of the anions existence, and this dictated the use of higher acidity than calculated for H2Nb2W4O29-.
Preliminary investigations revealed that pH of the mother liquor decreases dramatically during cation addition and formation of heterogenous system (Table 4). This leads to the case, when, instead of expected HxNb2 W4O1(49-x)-, there forms the mixture of its salt with
Table 4. Parameters of IPNTA isolation.
Anion	Z exp	^ theor	mol % in solution
Nb2W4O194-	1.13	1.11	71
HNb2W4O193-	1.31	1.28	79
H2Nb2W4O192-	1.60	1.44	42
the compound, containing Hx+1Nb2W4O1(39-x)-. Fortunatly, this was not the case at Z = 1.60, when the only salt with the desired anion H2Nb2W4O129- was formed.
To avoid such a decrease in pH, the acetate buffers were used. Their pH was adjusted in accordance to the Figure 1 for the needed Z. In these buffers the series of thallium salts were obtained and analyzed for the contents of the main constituents: Tl2O, Nb2O5, WO3 ■ H2O (Tables 5 and 6).
The chemical analysis of the synthesized thallium salts allowed to suggest the following molecular formulas: Tl3HNb2W4O19 ■ 10H2O and Tl2H2Nb2W4O19 ■ 10H2O. The presence of the anions HxNb2W4O1(49-x)- was evidenced by FTIR spectroscopy. The following bands at 951(vw=0), 895(vNb=o), 800, 694, 568 and 400 (vM_0_M • 8M_0_M) cm1 clearly indicate on the substructure M6O19, as it was described by Rocchiccioli-Deltcheff et al.14 and Anderson et al.15
Table 5. Chemical analysis data for the salts, being precipitated from solutions without using a buffer solution.
Z = 1.13	Tip	Nb2O5	WO3	H2O
Found, %	33.95	12.61	44.00	9.14
Calculated for	33.89	12.66	44.17	9.00 0.35Tl4Nb2W4O19 + 0.65Tl3HNb2W4O19, %
				
Z = 1.31	TI2O	Nb2Os	wo3	H2O
Found, %	27.58	13.77	47.56	9.79
Calculated for	27.50	13.77	48.03	10.00
0.5Tl3HNb2W4019 + 0.5Tl2HNb2W4019, %				
Z = 1.60	TI2O	Nb2O5	WO3	H2O
Found, %	21.58	14.52	51.64	9.68
Calculated for	23.63	14.78	51.57	10.02
Tl2H2Nb2W40199H20, %				
Table 6. Chemical analysis data for the salts, being precipitated from solutions, using acetate buffer solution.
	Tip	Nb2O5	WO3	H2O
Found, %	35.25	11.97	42.81	8.85
Calculated for				
07Tl4Nb2W4O19+03Tl3HNb2W4O19, %	37.91	11.86	41.38	8.85
	TI2O	Nb2(O5	WO3	H2O
Found, %	31.94	13.42	46.53	9.51
Calculated for				
Tl3HNb2W4O1910H2O, %	31.55	13.16	45.92	9.37
	TI2O	Nb2O5	WO3	H2O
Found, %	22.13	14.34	50.42	11.01
Calculated for				
Tl2H2Nb2W4O1910H2O, %	23.39	14.63	51.06	10.91
SEM images of thallium salt powders (Figure 4) show that there are no zones with different surface morphology and EDX spectral analysis in single spots and different zones with area from 24 x 14.4 to 66.8 x 49.9 jm exhibit no substantial deviation from the ratio Nb : W = 2 : 4. These clearly indicate the formation of monophase
samples of Tl4xHxNb2W4O19, and not the mixture of Tl3 xHxNbW5O19 and Tl5-xHxNb3W3O19. The analogous SEM and EDX results were also obtained for all the thallium salts, described in Tables 5 and 6. The compound Tl2H2Nb2W4O19 ■ 10H2O was chosen to serve as an example.
Figure 4. SEM images of the morphology of Tl2H2Nb2W4O19 ■ 10H2O powder. Compositional spectral analysis was made in single spots and zones indicated (See Table 7).
Scheme 1. Possible anions transformations in the system Nb6Ü198 -WÜ42 -H+-H2Ü.
Table 7. Molar ratio NbiW in different spots and zones of
5. References
Image a	N* Nb		Image b	N* Nb	
spot 1	2	3.89	zone 1	2	4.05
spot 2	2	4.07	zone 2	2	3.95
spot 3	2	3.93	zone 3	2	4.02
spot 4	2	3.92	zone 4	2	3.96
spot 5	2	4.09	zone 5	2	4.03
spot 6	2	4.15	zone 6	2	3.99
spot 7	2	3.89	spot 7	2	3.93
spot 8	2	4.08	spot 8	2	3.98
spot 9	2	3.89	spot 9	2	3.84
spot 10	2	3.85	spot 10	2	3.89
spot 11	2	4.12	spot 11	2	4.00
spot 12	2	3.91	spot 12	2	3.96
spot 13	2	4.07	spot 13	2	3.89
spot 14	2	4.08			
spot 15	2	4.05			
spot 16	2	3.98			
* - the ratio was recalculated for 2 atoms of Nb for simplicity
4. Conclusions
To finalize, mathematical modeling allowed us to adequately interprete the experimental results and to suggest the following sequence of parallel-consequent processes in the system Nb6O189-WO4!--H+-H2O with CNb : CW = 2 : 4 (Scheme 1).
1.	J. T. Rhule, C. L. Hill, D. A. Judd Chem. Rev. 1998, 98, 327358.
2.	C. Besecker, W. G. Klemperer, V. W. Day J. Am. Chem. Soc. 1982, 104, 6158-6159.
3.	V. W. Day, T. A. Eberspacher, W. G. Klemperer, R. P. Planalp, P. W. Schiller, A. Yagasaki, B. Zhong Inorg. Chem. 1993, 32, 1629-1637.
4.	M. Dabbabi, M. Boyer J. Inorg. Nucl. Chem. 1976, 38, 1011-1014.
5.	R. Bates pH Determination. Theory and practice (Book), Khimija, Leningrad 1968, 94-124 (in Russian).
6.	Y. Kholin Quantitative physico-chemical analysis of complex formation in solutions and on the surface of the chemically modified silica: models, mathematical methods and their applications (Book), Folio, Kharkov 2000 (in Russian).
7.	K. S. Pitzer, G. Mayorga J. Phys. Chem. 1973, 77, 23002308.
8.	A. A. Bugaevski, Y. V. Kholin, D. S. Konjaev, A. V. Krasovit-ski Zh. Obshch. Khim. 1998, 68, 753-757 (in Russian).
9.	G. Meinrath Anal. Bioanal. Chem. 2002, 374, 796-805.
10.	J. J. Hastings, O. W. Howarth J. Chem. Soc. Dalton Trans. 1992, 209-215.
11.	G. M. Rozantsev, S. M. Vavilova, K. E. Belousova Rus. J. of Inorg. Chem. 2007, 52, 1478-1485.
12.	G. M. Rozantsev, O. I. Sazonova, Y. V. Kholin Zh. Neorg. Khim. 1998, 43, 1894-1899 (in Russian).
13.	B. Spinner Rev. Chim. Miner. 1968, 5, 839-868.
14.	C. Rocchiccioli-Deltcheff, R. Thouvenot, M. Dabbabi Spec-trochim. Acta 1977, A33, 2, 143-153.
15.	Anderson T. M., Rodriguez M. A., Stewart T. A. et al. Eur. J. Inorg. Chem. 2008, 8286-3294.
Tl2H2Nb2W4Ü19 ■ 10H2Ü powder.
Povzetek
Potenciometrično smo raziskovali proces nastanka kompleksa Nb6O198--WO42--H+-H2O pri razmerju CNb : CW = 2 : 4 ter različnih Nb + W koncentracijah. Eksperimentalne podatke smo analizirali z matematičnim modelom, ki omogoča konstrukcijo porazdelitvenega diagrama posameznih niobijevih in tungstenovih izopoli anionov ter mešanih izopolinio-botungstatov v območju Z = C^+/CNb+W = 0 - 2.0 (ob prisotnosti NaCl). Z uporabo quasi-Neutonove metode (CLINP 2.1 software) smo določili termodinamske tvorbene konstante. Pokazali smo, da te tvorba isopoliniobotungstatov (HxNb2W4O(4-x)-, * = 0 - 2) v 6. periodi periodnega sistema poteka preko itermediatnih ionov Nb3W3O195-. Izolirali smo soli talija Tl3hNb2W4O19 • 10H2O ter Tl2H2Nb2W4O19 • 10H2O ter jih analizirali z elementno analizo in FTIR spektroskopijo.