- Email: [email protected]
Studies on molybdophosphates containing a group 4A metal ion by laser raman spectroscopy
Analytica Chimica Acta, 151 (1983) 29-38 Elsevier Science Publishers B.V., Amsterdam - Printed in The Netherlands
STUDIES ON MOLYBDOPHOSPHATES CONTAINING A GROUP 4A METAL ION BY LASER RAMAN SPECTROSCOPY Interference of Group 4A Metal Ions in the Determination of Phosphorus
KATSUO MURATA* Department
(Jopan)
and SHIGERO IKEDA
of Chemistry,
Faculty of Science,
Osaka University, Toyonaka,
Osaka 560
(Received 7th December 1982)
SUMMARY The interference of group 4A metal ions in the determination of phosphorus (as phosphate) was investigated by means of laser Raman spectroscopy. Raman spectra and elemental analysis showed that 12-molybdophosphoric acid is transformed to ll-molybdometalophosphoric acid (a ternary heteropolymolybdate containing the group 4A element) in the presence of the group 4A metal ions. The 12-molybdophosphoric acid has an intense Raman line at 996 cm-‘, whereas the ll-molybdometalophosphoric acid gives a line shifted by lo-16 cm-’ towards a lower wavenumber. This difference makes it possible to estimate the formation constant of the ternary heteropolymolybdate from 12-molybdophosphoric acid. Ternary heteropolymolybdates containing the group 4A element are stable in aqueous solution, but are less readily extracted than the 12-molybdo compound. This is why the group 4A metal ions interfere in the determination of phosphorus.
The presence of a group 4A element (Ti, Zr, and Hf) always leads to errors in the determination of phosphorus by means of a heteropoly acid [l, 21. Such interfering side-reactions were particularly apparent for solutions containing a group 4A element and 12-molybdophosphoric acid. In previous work [3] , the spectrophotometric determination of titanium(IV), zirconium(IV), and thorium(IV) by means of this reaction was examined. The additive production of a ternary heteropolymolybdate has been proposed for this reaction [4], but the existence of the ternary heteropolymolybdate was not readily distinguished in the aqueous solution because the ultraviolet absorption band of the ternary heteropolymolybdate was similar to that of 12-molybdophosphoric acid. The existence of this complex in the aqueous solution can, however, be established by using laser Raman spectroscopy. The purpose of this study is to clarify the reason for the interference in the determination of phosphorus and to estimate the stabilities of the ternary heteropolymolybdate complexes by laser Raman spectroscopy.
0003-2670/83/$03.00
0 1983 Elsevier Science Publishers B.V.
30 EXPERIMENTAL
Reagents A molybdate solution was prepared from sodium molybdate (NazMo04 * 2Hz0; analytical-reagent grade) and standardized gravimetrically as its 8quinolinolate [5]. Polynuclear molybdate complexes were produced by acidification of this molybdate solution. Solutions of phosphate, zirconium(IV), and hafnium(IV) were prepared from potassium dihydrogenphosphate, zirconium oxychloride octahydrate, and hafnium oxychloride octahydrate (all analytical-reagent grade), respectively. A solution of titanium(IV) was prepared by dissolving titanium metal (99.9%) in dilute sulfuric acid and then oxidizing with a small amount of nitric acid. The group 4A metal ions were stored as 2 M perchloric or hydrochloric acid solutions to avoid hydrolysis. Each solution of Ti(IV), Zr(IV), and Hf(IV) was standardized gravimetrically by ignition of the precipitate with cupferron [6] . A solution of molybdophosphate was prepared by acidifying mixed solutions of molybdate and phosphate. A solution of ternary heteropolymolybdate was prepared by acidifying mixed solution of molybdate, phosphate, and the group 4A metal ion. In this case, the order in the addition of each component may be important because the combination of a group 4A ion and phosphate (or molybdate) may cause the precipitation of the group 4A metal phosphate (or molybdate). In order to prepare a solution of ternary heteropolymolybdate, therefore, the mixed solution of molybdate and phosphate is acidified so that molybdophosphate is produced, and then the group 4A metal ion is added. This order favors rapid formation of the ternary heteropolymolybdate. Otherwise, 50 min may be needed to dissolve the precipitate of the group 4A metal molybdate and then to form ternary heteropolymolybdate. Equipment Raman spectra were measured with a JASCO R750 triple monochromator and JASCO RSOO; a 514.5~nm Ar+ laser was used as the excitation source. Sample solutions of l-ml volume were used for measurements. It was difficult to measure the molybdate solutions below 10m3M, and a molybdate concentration of 0.12 M was used to obtain better Raman spectra at lower wavenumbers. In order to measure quantitatively the intensity of Raman lines, an internal standard was used. Raman lines of molybdate compounds appear at wavenumbers below 1000 cm-‘. The intense and sharp line of nitrate is observed at 1049 cm-‘; sodium nitrate of constant concentration (0.08 M) was therefore used as an internal standard, added to each sample solution. This nitrate addition did not affect the Raman spectra of the molybdate solutions. The Raman lines observed were calibrated against those of indene [7] .
31 RESULTS AND DISCUSSION
The formation mechanism of molybdophosphate complexes in weakly acidic solutions has recently been examined by laser Raman spectroscopy [S, 91; hexamolybdate, Mo60:;, and dodecamolybdate, Mo,,O$;, were found in weakly acidic molybdate solutions (2 > 1.5). It was shown that dodecamolybdate is closely correlated with the formation of 12-molybdophosphoric acid and that 12-molybdophosphoric acid is in equilibrium with 11-molybdophosphoric acid. Further investigations on heteropolymolybdates indicated that the vibrational spectra of ternary heteropolymolybdate are different from that of molybdophosphate. Figure 1 shows polarized Raman spectra of three ternary heteropolymolybdates and 12-molybdophosphoric acid solutions. The intense Raman lines observed at the highest wavenumber (996,986,980 and 980 cm-‘) in every heteropolymolybdate disappear for measurements perpendicular to the excitation. These polarization measurements support the idea that the Raman lines of 12-molybdophosphoric (996 cm-‘), molybdotitanophosphoric (986 cm-‘), molybdozirconophosphoric (980 cm-‘), and molybdohafnophosphoric (980 cm-‘) acid are all vl lines, i.e., symmetric stretching of Mo=O. Thus a shift of the v1 line towards lower wavenumber is observed in the case of ternary heteropolymolybdates, and the intense IQ line of each ternary heteropolymolybdate is readily distinguished from that of 12-molybdophosphoric acid. The intense sharp v1 line is suitable for the evaluation of each species [8], thus the Raman measurements discussed here were mainly concerned with the v1 line observed over 900-1000 cm-‘. Figure 2 shows the influence of titanium(IV) concentration on the Raman spectra of 12-molybdophosphoric acid solution. These spectra clearly indicate the conversion of 12-molybdophosphoric acid to molybdotitanophosphoric acid. The peak intensity at 996 cm-’ decreases while a peak appears at 986 cm-’ and its intensity increases in proportion to the concentration of titanium(IV). The relative intensities of the peaks at 996 cm-’ and 986 cm-’ are plotted against the concentration of titanium(IV) in Fig. 3. It can be seen that the increasing intensity at 986 cm-’ is closely related to the decreasing intensity at 996 cm-‘, i.e., molybdotitanophosphoric acid is produced, while 12-molybdophosphoric acid disappears as the concentration of Ti(IV) increases. Although the intensity at 996 cm-’ indicates only the l2-molybdophosphoric acid, the intensity at 986 cm-’ always contains not only the contribution of molybdotitanophosphoric acid but also that of 12-molybdophosphoric acid. The contribution of 12-molybdophosphoric acid to curve D (Fig. 3) can be estimated from curve B and the calibration curves of Fig. 4, so that the formation curve (C) for molybdotitanophosphoric acid alone can be plotted by subtracting curve D from curve A. Similar features were observed in the case of molybdozirconophosphoric acid and molybdohafnophosphoric acid. The influence of the zirconium(IV) concentration on the Raman spectra of 12-molybdophosphoric acid is shown
32 1
996
+
980
I
LJ ft.,..
Km
__/---
300
/’
___/----
600
Wavenumber
ml
: I’
:
:
:
(4)
y .=
10 ’
9
d 0
200
( cm-’1
Wavenumber
(cm+)
Fig. 1. Polarized Raman spectra of aqueous solutions of binary and ternary heteropoly acids: (1) 12-molybdophosphoric; (2) molybdotitanophosphoric;( 3) molybdozirconophosphoric; (4) molybdohafnophosphoric acid. (-) Raman radiation parallel to excitation;(--) Raman radiation perpendicular to excitation. ([MO] = 1.2 X 10-l M, [P] = 1.0 X lo-’ M, [Me] = 1.0 X 10m2M, pH 1). Fig. 2. Influence of titanium(IV) concentration on the Raman spectra of 12-molybdophosphoric acid solution. Titanium(IV) concentration: (1) 0; (2) 0.2 X lo-” M; (3) 0.4 X lo-* M; (4) 0.6 x lo-’ M; (5) 0.8 x 10-l M;(6) 1.0 x lo-* M. ([P] = 1.0 x 1O-a M, [MO] = 1.2 x 10-l M, [NO;] = 8.0 x lo-’ M, pH 1.0).
in Fig. 5. The relative intensities are plotted against the concentration of Zr(IV) or Hf(IV) in Figs. 6 and 7. The IQ lines of both molybdozircono- and molybdohafnophosphoric acid appear at 980 cm-‘. The addition of Zr(IV) or Hf(IV) to the solution of 12-molybdophosphoric acid produces molybdozirconophosphoric or molybdohafnophosphoric acid. The formation of ternary heteropolymolybdate is considered as follows.
.-
i!
Y
c ._
1.0*
.-Y 3 t
u
0.5 *
Oh
0.5
1.0
IPMo~zO+o’-I (IO-=M 1 Fig. 3. Formation curve of molybdotitanophosphoric acid estimated from the two Raman lines observed. (A) Intensity at 986 cm-’ (observed);(B) intensity at 996 cm-’ (observed); (C) intensity at 986 cm-’ (estimated by A-D); (D) intensity of l2-molybdophosphoric acid at 986 cm-’ (estimated from calibration curve). Fig. 4. Calibration (3) 986 cm-‘.
curves for 12.molybdophosphoric
acid: (1) 996 cm-‘; (2) 980 cm-‘;
Mechanisms It has already been shown [9] that 12-molybdophosphoric acid is in equilibrium with 11-molybdophosphoric acid in solutions at pH l-2. The latter species has a peak at 975 cm-‘, which decreases with increasing acidity, but some 11-molybdophosphoric acid remains at pH 1. Il-Molybdophosphoric acid has a “defect” Keggin structure, missing one of MoOa unit, so that there is a hole surrounded by oxygen atoms [lo]. The anion can then behave as a ligand to take up a metal ion in this hole [lo, 111. Therefore 11-molybdophosphoric acid is more labile than 12-molybdophosphoric acid. When the group 4A metal ion is added to the solution of ll- and 12-molybdophosphoric acid in equilibrium, the metal ion reacts with ll-molybdophosphoric acid to produce ternary heteropolymolybdate; the consumed 11-molybdophosphoric acid is then replaced by conversion of 12-molybdophosphoric acid. Finally, the 12-molybdophosphoric acid is virtually converted to the ternary heteropolymolybdate. The elemental results for the tetraethylammonium salts obtained from a
2.5
.-Y 3 rr”
1.0’ f'Q
0.5
1ooo
950
Wavenumber
C
900 ( cm-’ )
Fig. 5. Influence of zirconium(IV) concentration on the Raman spectra of 12-molybdophosphoric acid solution. Zirconium concentration: (1) 0; (2) 0.2 X lo-’ M; (3) 0.4 X lo-’ M; (4) 0.6 x lo-’ M; (5) 0.8 X 10va M; (6) 1.0 X lo4 M. Other conditions as for Fig. 2. Fig. 6. Formation curve of molybdozirconophosphoric acid estimated from the two Raman lines observed. (A) Intensity at 980 cm-’ (observed); (B) intensity at 996 cm-’ (observed); (C) intensity at 980 cm-’ (estimated by A-D); (D) intensity of 12-molybdophosphoric acid at 980 cm-’ (estimated from calibration curve).
TABLE 1 Elemental results for tetraethylammonium (All results are given as percentages)
C H N Me P MO
ternary heteropolymolybdate
salts
Molybdotitanophosphate
Molybdozirconophosphate
Molybdohafnophosphat
caica
Found
Caleb
caka
Found
Caleb
Cab?
Found
Caleb
13.3 2.9 1.9 2.2 1.4 48.7
14.7 3.2 2.2 2.2 1.3 47.2
12.5 2.7 1.8 2.1 1.3 49.8
13.0 2.8 1.9 4.1 1.4 47.8
11.9 2.7 1.75 4.9 1.35 48.0
12.3 2.6 1.8 3.9 1.3 49.3
12.5 2.7 1.8 7.8 1.35 45.95
12.2 2.7 1.8 7.0 1.3 46.8
11.8 2.5 1.7 7.3 1.3 47.2
aCaicuIated for [(C,H,),N] ,PMo,,MeO,,H,O, where Me = Ti, Zr or Hf. bCalculated for [(C,H,),N] ,MePMo,,O,,H,O, where Me = Ti, Zr or Hf.
36
solution at pH 1 (Table 1) tend to support reaction (a): PMoIIOg + Me02+ + 2H20 + PMollMeO$, + 4H+
(a)
PMo,O& + Me02+ + Hz0 + PMo12042Me* + 2H’
(b)
with formation of 11-molybdometalophosphates, PMollMeO&,, rather than 12-molybdometalophosphates PMo12042MeJ-. Previously [3] , it was wrongly concluded that the ternary heteropolymolybdate was formed via reaction (b); this might be attributed to the fact that the amounts of ternary heteropolymolybdate formed could not be estimated accurately because of the similarity of the near-ultraviolet spectra of binary and ternary heteropolymolybdates. The present Raman measurements, which can estimate only the ternary heteropolymolybdate, and the elemental analysis of ternary heteropolymolybdate salts support reaction (a). Estimation of the formation constant In the solutions relevant to reaction (a), only PMo~~O&, PMo~~O~Y,and PMo,,MeO& species are active in Raman spectroscopy. The Ti(IV), Zr(IV), and Hf(IV) species have no Raman lines in the same region. The simple spectrometric estimation of the formation constant, K, for the reaction A + (a-x)
B
=+C
(b-x)
x
is based on K = x/(a - x)(b - LX),where a and b are the initial concentrations of A and B, respectively, and, x is the equilibrium concentration of C. If d is the absorbance and p is the molar absorptivity, then d = bAta -x)
+ pB(b -x)
+ h$ = @Aa + PBb) + & --PA -h)x
(1)
Combination of Eqn. 1 and the expression for K yields l/K=[(tl--o)/(p,-_A-PB)I
--a--b+
[ab(Pc--OA-P~/@-~O)I
(2)
where do = @Ac + &b). When B does not absorb, pB = 0, and Eqn. (2) is simplified. The resulting plot is called the Rose-Drago plot [ 121. When PA = /3s = 0, Eqn. (2) becomes even simpler: l/K = (d/P,) -a
- b + (ab&/d)
(3)
When the value of l/K is plotted against the arbitrary /3c for several sets of the initial concentration of A and B, the probable value of l/K is obtainable as the intersection point of several straight lines. This procedure was tested for application with the Raman measurements. The actual formation curve (C) of molybdotitanophosphoric acid in Fig. 3 is well suited to the conditions of the simplified Eqn. (3). The sets of l/K and pc for molybdotitanophosphoric acid are plotted in Fig. 8. The sets of l/K and PC for other ternary heteropolymolybdates are given in Tables 2 and 3. The formation constants of the ternary heteropolymolybdates from 12-molybdophosphoric acid were found from these plots to be 5.4 + 0.1 X lo3 for molybdotitanophosphoric acid, 2.8 + 0.3 X lo4 for molybdozirconophosphoric acid and 1 .l + 0 2 X lo4 for molybdohafnophosphoric acid.
36
2.5
5.0
2.0
4.0
h .< E$I ._
1-5
z ._ 2
1.0
B 0.5
C
0.5 CHfl
) 2.3
1.0
2.5
nc
(10” M 1
Fig. 7. Formation curve of molybdohafnophosphoric lines observed. Curves A-D as for Fig. 6.
2.7
(10”)
acid estimated from the two Raman
Fig. 8. Rose-Drag0 plot for the formation of molybdotitanophosphoric acid. Titanium(IV) concentration: (1) 0.1 X lo-’ M; (2) 0.5 X lo-’ M; (3) 0.9 X lo-” M;(4) 1.0 X lo-’ M. Other conditions as for Fig. 2.
TABLE 2 The sets of l/K and pc for molybdozirconophosphoric
PC
l/K
PC
l/K
PC
l/K
PC
l/K
a-1.0~ 2.355 0.66 a = 1.0 2.32 0.90
lo-‘M,b=0.4X 2.360 1.89 X
lo-‘M 2.365 3.21
lOma M, b = 0.9 X 10-l M 2.34 2.36 1.99 3.19
acid
2.370 4.44
2.380 X 10’ 7.10 x 10-5
2.38 4.48
2.40 X lo* 5.98 x lo-’
a=1.0x10-“M,b=0.95x10-aM 2.30 2.32 0.64 1.33
2.34 2.11
2.36 3.10
2.33 X lo= 4.19 x lo+
a=l.Ox 2.24 0.18
2.28 0.97
2.30 1.62
2.32 X 10’ 2.37 X lo-’
10-aM,b=l.OX1O-‘M 2.26 0.52
37 TABLE3 Thesetsofl/Kandpc formolybdohafnophosphoricacid a=l.OXIOmlM,
PC l/K
2.48 0.12
2.49 0.48
a = 1.0 x lo-‘M, PC I/K
PC
l/K
PC
UK
b=0.086~~~-~M 2.50 0.85 1:22
2.53 1.96
2.54 2.34
2.55X 10' 2.71X lo4
2.52 1.76
2.54 2.19
2.56 2.62
2.58x lOa 3.05x 1o-4
2.48 0.82
2.50 1.00
2.52 1.19
2.54x lo* 1.34x lo4
2.42 0.54
2.44 0.67
2.46 0.81
2.48x lo1 0.96x lo+
b = 0.518 x lOwa M
2.44
2.46
2.46
2.50
0.11
0.51
0.92
1.33
a = 1.0 X IO-‘M, 2.40 2.42 0.19 0.33
2.52 1.59
b = 0.864x 10-l M 2.44 2.46 0.48 0.64
a = 1.0 X IO-"M, b = 0.950x IO-“M 2.34 2.36 2.38 2.40 0.13 0.21 0.30 0.42
Assignment of Raman lines The Raman lines were initially assigned for the mononuclear species MoO$- by Busey and Keller [ 131. The vl, v3, and v4 Raman lines are observed at 897, 841, and 318 cm-‘, respectively. The present authors also observed three Raman lines and confirmed that the Raman line at 897 cm-’ is atotally symmetric vibration by polarization measurements. On acidification of the mononuclear molybdate solution, the Raman spectra change drastically. The v3 and v4 lines disappear. The ul line decreases greatly but a new vl line appears at the higher wavenumber of 940 cm-‘; this was also confirmed to be a totally symmetric vibration by polarization measurements [8] . For isopolymolybdates, Griffith and Lesniak [14 ] assigned bands above 900 cm-’ to Mo=O stretching and those near 350 cm-’ to Mo=O bending; 840-750 cm-’ bands were assigned to asymmetric MO-O-MO stretching and those near 200 cm’ to MO-O-MO deformations 1141. The present authors observed that the vl line of Mo=O stretching shifts towards higher wavenumber as condensation proceeds: Mo,O$, at 940 cm-‘; Mo,O$& at 970 cm-l; and Mo60?;, at 980 cm-‘. These intense Raman lines were also confirmed to be totally symmetric vibrations by polarization measurements [S] . Raman lines below 900 cm-’ are very faint in the isopolymolybdate solutions. Only bending of Mo=O is observed at 365 cm-‘. Deformations of MO-O-MO at 220 cm-’ overlaps with strong Rayleigh scattering, and this line is observable only for high concentrations of molybdate (1.2 M). Stretching of MO-G-MO near 600 cm-’ is not obvious in the isopolymolybdate spectra, but binary and ternary heteropolymolybdates show a characteristic broad Raman peak near 620 cm-‘. Although hexamolybdate also has this Raman band near 600 cm-‘, isopolymolybdate solutions containing hexamolybdate do not give the band because of small amounts of MosO:; [ 151. Some Raman lines of 12-molybdophosphoric acid were discussed by Franck and co-workers [16],
who assigned the Raman line at 996 cm -l to the totally symmetric stretching of Mo=O,. In the present work, the conversion of 12-molybdophosphoric acid to the ternary heteropolymolybdate containing the group 4A element was followed by careful observation of the v1 line. The v1 line of the ternary heteropolymolybdate was observed with a shift of 10-16 cm-‘. Lynhamn et al. attempted to estimate force constants for 12-molybdophosphoric acid [17]. The shift towards lower wavenumber in the ternary heteropolymolybdates may reflect a decrease in the force constant (Mo=O) caused by replacement of a molybdenum atom with a group 4A element. Within the isopoly- and heteropoly-molybdates, 12-molybdophosphoric acid, having the best symmetry, provides the u1 line at the highest wavenumber (996 cm-‘). As shown above, the reaction of the group 4A metal ions with 12-molybdophosphoric acid produces ternary heteropolymolybdates containing the group 4A element. These complexes are more stable than 12-molybdophosphoric acid in the aqueous solution, but they are less extractable. Therefore the presence of the group 4A elements causes interference in the determination of phosphorus by extraction of the heteropoly acid. More basic solvents such as cyclohexanone and tributylphosphate, however, can extract the ternary heteropolymolybdates without interference. As previously reported, spectrophotometric determinations of the group 4A metal ions are possible by using the formation of ternary heteropolymolybdates [3] . REFERENCES 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
P. PakaIns, Anal. Chim. Acta, 40 (1968) 1. M. Ishibashi and M. Tabushi, Bunseki Kagaku, 8 (1959) 588. K. Murata, Y. Yokoyama and S. Ikeda, Anal. Chim. Acta, 48 (1969) 349. A. K. Babko and Y. F. Shkaravskii, Zh. Neorg. Khim., 6 (1961) 2091. M. Borrel and R. Paris, Anal. Chim. Acta, 4 (1950) 267. P. J. Elving and E. C. Olson, Anal. Chem., 27 (1955) 1817. P. J. Hendra and E. J. Loader, Chem. Ind., (1968) 718. K. Murata and S. Ikeda, Spectrochim. Acta, Part A, 39 (1983) in press. K. Murata and S. Ikeda, Polyhedron, 2 (1983) in press. C. Rocchiccioli-Deltcheff and R. Thouvenot, J. Chem. Research (S), (1977) 46. F. Umland, A. Janssen, D. Thierig and G. Wiinsch, Theorie und Praktische Anwendung von Komplexbildenern, Akademische Verlagsgesellschaft, Frankfurt, 1971, p. 42-43. N. J. Rose and R. S. Drago, J. Am. Chem. Sot., 81(1959) 6138. R. H. Busey and D. L. Keller, Jr., J. Chem. Phys., 41(1961) 215. W. P. Griffith and P. J. B. Lesniak, J. Chem. Sot., Sect. A, (1969) 1066. K. Murata, E. Yamamoto and S. Ikeda, Bull. Chem. Sot. Jpn., 56 (3) (1983) 941. C. Rocchiccioli-Deltcheff, R. Thouvenot and R. Franck, Spectrochim. Acta, Part A, 32 (1976) 587. L. Lyhamn, S. J. Cyvin, B. N. Cyvin and J. Brunvoll, Z. Naturforsch., Teil A, 31 (1976) 1589.
STUDIES ON MOLYBDOPHOSPHATES CONTAINING A GROUP 4A METAL ION BY LASER RAMAN SPECTROSCOPY Interference of Group 4A Metal Ions in the Determination of Phosphorus
KATSUO MURATA* Department
(Jopan)
and SHIGERO IKEDA
of Chemistry,
Faculty of Science,
Osaka University, Toyonaka,
Osaka 560
(Received 7th December 1982)
SUMMARY The interference of group 4A metal ions in the determination of phosphorus (as phosphate) was investigated by means of laser Raman spectroscopy. Raman spectra and elemental analysis showed that 12-molybdophosphoric acid is transformed to ll-molybdometalophosphoric acid (a ternary heteropolymolybdate containing the group 4A element) in the presence of the group 4A metal ions. The 12-molybdophosphoric acid has an intense Raman line at 996 cm-‘, whereas the ll-molybdometalophosphoric acid gives a line shifted by lo-16 cm-’ towards a lower wavenumber. This difference makes it possible to estimate the formation constant of the ternary heteropolymolybdate from 12-molybdophosphoric acid. Ternary heteropolymolybdates containing the group 4A element are stable in aqueous solution, but are less readily extracted than the 12-molybdo compound. This is why the group 4A metal ions interfere in the determination of phosphorus.
The presence of a group 4A element (Ti, Zr, and Hf) always leads to errors in the determination of phosphorus by means of a heteropoly acid [l, 21. Such interfering side-reactions were particularly apparent for solutions containing a group 4A element and 12-molybdophosphoric acid. In previous work [3] , the spectrophotometric determination of titanium(IV), zirconium(IV), and thorium(IV) by means of this reaction was examined. The additive production of a ternary heteropolymolybdate has been proposed for this reaction [4], but the existence of the ternary heteropolymolybdate was not readily distinguished in the aqueous solution because the ultraviolet absorption band of the ternary heteropolymolybdate was similar to that of 12-molybdophosphoric acid. The existence of this complex in the aqueous solution can, however, be established by using laser Raman spectroscopy. The purpose of this study is to clarify the reason for the interference in the determination of phosphorus and to estimate the stabilities of the ternary heteropolymolybdate complexes by laser Raman spectroscopy.
0003-2670/83/$03.00
0 1983 Elsevier Science Publishers B.V.
30 EXPERIMENTAL
Reagents A molybdate solution was prepared from sodium molybdate (NazMo04 * 2Hz0; analytical-reagent grade) and standardized gravimetrically as its 8quinolinolate [5]. Polynuclear molybdate complexes were produced by acidification of this molybdate solution. Solutions of phosphate, zirconium(IV), and hafnium(IV) were prepared from potassium dihydrogenphosphate, zirconium oxychloride octahydrate, and hafnium oxychloride octahydrate (all analytical-reagent grade), respectively. A solution of titanium(IV) was prepared by dissolving titanium metal (99.9%) in dilute sulfuric acid and then oxidizing with a small amount of nitric acid. The group 4A metal ions were stored as 2 M perchloric or hydrochloric acid solutions to avoid hydrolysis. Each solution of Ti(IV), Zr(IV), and Hf(IV) was standardized gravimetrically by ignition of the precipitate with cupferron [6] . A solution of molybdophosphate was prepared by acidifying mixed solutions of molybdate and phosphate. A solution of ternary heteropolymolybdate was prepared by acidifying mixed solution of molybdate, phosphate, and the group 4A metal ion. In this case, the order in the addition of each component may be important because the combination of a group 4A ion and phosphate (or molybdate) may cause the precipitation of the group 4A metal phosphate (or molybdate). In order to prepare a solution of ternary heteropolymolybdate, therefore, the mixed solution of molybdate and phosphate is acidified so that molybdophosphate is produced, and then the group 4A metal ion is added. This order favors rapid formation of the ternary heteropolymolybdate. Otherwise, 50 min may be needed to dissolve the precipitate of the group 4A metal molybdate and then to form ternary heteropolymolybdate. Equipment Raman spectra were measured with a JASCO R750 triple monochromator and JASCO RSOO; a 514.5~nm Ar+ laser was used as the excitation source. Sample solutions of l-ml volume were used for measurements. It was difficult to measure the molybdate solutions below 10m3M, and a molybdate concentration of 0.12 M was used to obtain better Raman spectra at lower wavenumbers. In order to measure quantitatively the intensity of Raman lines, an internal standard was used. Raman lines of molybdate compounds appear at wavenumbers below 1000 cm-‘. The intense and sharp line of nitrate is observed at 1049 cm-‘; sodium nitrate of constant concentration (0.08 M) was therefore used as an internal standard, added to each sample solution. This nitrate addition did not affect the Raman spectra of the molybdate solutions. The Raman lines observed were calibrated against those of indene [7] .
31 RESULTS AND DISCUSSION
The formation mechanism of molybdophosphate complexes in weakly acidic solutions has recently been examined by laser Raman spectroscopy [S, 91; hexamolybdate, Mo60:;, and dodecamolybdate, Mo,,O$;, were found in weakly acidic molybdate solutions (2 > 1.5). It was shown that dodecamolybdate is closely correlated with the formation of 12-molybdophosphoric acid and that 12-molybdophosphoric acid is in equilibrium with 11-molybdophosphoric acid. Further investigations on heteropolymolybdates indicated that the vibrational spectra of ternary heteropolymolybdate are different from that of molybdophosphate. Figure 1 shows polarized Raman spectra of three ternary heteropolymolybdates and 12-molybdophosphoric acid solutions. The intense Raman lines observed at the highest wavenumber (996,986,980 and 980 cm-‘) in every heteropolymolybdate disappear for measurements perpendicular to the excitation. These polarization measurements support the idea that the Raman lines of 12-molybdophosphoric (996 cm-‘), molybdotitanophosphoric (986 cm-‘), molybdozirconophosphoric (980 cm-‘), and molybdohafnophosphoric (980 cm-‘) acid are all vl lines, i.e., symmetric stretching of Mo=O. Thus a shift of the v1 line towards lower wavenumber is observed in the case of ternary heteropolymolybdates, and the intense IQ line of each ternary heteropolymolybdate is readily distinguished from that of 12-molybdophosphoric acid. The intense sharp v1 line is suitable for the evaluation of each species [8], thus the Raman measurements discussed here were mainly concerned with the v1 line observed over 900-1000 cm-‘. Figure 2 shows the influence of titanium(IV) concentration on the Raman spectra of 12-molybdophosphoric acid solution. These spectra clearly indicate the conversion of 12-molybdophosphoric acid to molybdotitanophosphoric acid. The peak intensity at 996 cm-’ decreases while a peak appears at 986 cm-’ and its intensity increases in proportion to the concentration of titanium(IV). The relative intensities of the peaks at 996 cm-’ and 986 cm-’ are plotted against the concentration of titanium(IV) in Fig. 3. It can be seen that the increasing intensity at 986 cm-’ is closely related to the decreasing intensity at 996 cm-‘, i.e., molybdotitanophosphoric acid is produced, while 12-molybdophosphoric acid disappears as the concentration of Ti(IV) increases. Although the intensity at 996 cm-’ indicates only the l2-molybdophosphoric acid, the intensity at 986 cm-’ always contains not only the contribution of molybdotitanophosphoric acid but also that of 12-molybdophosphoric acid. The contribution of 12-molybdophosphoric acid to curve D (Fig. 3) can be estimated from curve B and the calibration curves of Fig. 4, so that the formation curve (C) for molybdotitanophosphoric acid alone can be plotted by subtracting curve D from curve A. Similar features were observed in the case of molybdozirconophosphoric acid and molybdohafnophosphoric acid. The influence of the zirconium(IV) concentration on the Raman spectra of 12-molybdophosphoric acid is shown
32 1
996
+
980
I
LJ ft.,..
Km
__/---
300
/’
___/----
600
Wavenumber
ml
: I’
:
:
:
(4)
y .=
10 ’
9
d 0
200
( cm-’1
Wavenumber
(cm+)
Fig. 1. Polarized Raman spectra of aqueous solutions of binary and ternary heteropoly acids: (1) 12-molybdophosphoric; (2) molybdotitanophosphoric;( 3) molybdozirconophosphoric; (4) molybdohafnophosphoric acid. (-) Raman radiation parallel to excitation;(--) Raman radiation perpendicular to excitation. ([MO] = 1.2 X 10-l M, [P] = 1.0 X lo-’ M, [Me] = 1.0 X 10m2M, pH 1). Fig. 2. Influence of titanium(IV) concentration on the Raman spectra of 12-molybdophosphoric acid solution. Titanium(IV) concentration: (1) 0; (2) 0.2 X lo-” M; (3) 0.4 X lo-* M; (4) 0.6 x lo-’ M; (5) 0.8 x 10-l M;(6) 1.0 x lo-* M. ([P] = 1.0 x 1O-a M, [MO] = 1.2 x 10-l M, [NO;] = 8.0 x lo-’ M, pH 1.0).
in Fig. 5. The relative intensities are plotted against the concentration of Zr(IV) or Hf(IV) in Figs. 6 and 7. The IQ lines of both molybdozircono- and molybdohafnophosphoric acid appear at 980 cm-‘. The addition of Zr(IV) or Hf(IV) to the solution of 12-molybdophosphoric acid produces molybdozirconophosphoric or molybdohafnophosphoric acid. The formation of ternary heteropolymolybdate is considered as follows.
.-
i!
Y
c ._
1.0*
.-Y 3 t
u
0.5 *
Oh
0.5
1.0
IPMo~zO+o’-I (IO-=M 1 Fig. 3. Formation curve of molybdotitanophosphoric acid estimated from the two Raman lines observed. (A) Intensity at 986 cm-’ (observed);(B) intensity at 996 cm-’ (observed); (C) intensity at 986 cm-’ (estimated by A-D); (D) intensity of l2-molybdophosphoric acid at 986 cm-’ (estimated from calibration curve). Fig. 4. Calibration (3) 986 cm-‘.
curves for 12.molybdophosphoric
acid: (1) 996 cm-‘; (2) 980 cm-‘;
Mechanisms It has already been shown [9] that 12-molybdophosphoric acid is in equilibrium with 11-molybdophosphoric acid in solutions at pH l-2. The latter species has a peak at 975 cm-‘, which decreases with increasing acidity, but some 11-molybdophosphoric acid remains at pH 1. Il-Molybdophosphoric acid has a “defect” Keggin structure, missing one of MoOa unit, so that there is a hole surrounded by oxygen atoms [lo]. The anion can then behave as a ligand to take up a metal ion in this hole [lo, 111. Therefore 11-molybdophosphoric acid is more labile than 12-molybdophosphoric acid. When the group 4A metal ion is added to the solution of ll- and 12-molybdophosphoric acid in equilibrium, the metal ion reacts with ll-molybdophosphoric acid to produce ternary heteropolymolybdate; the consumed 11-molybdophosphoric acid is then replaced by conversion of 12-molybdophosphoric acid. Finally, the 12-molybdophosphoric acid is virtually converted to the ternary heteropolymolybdate. The elemental results for the tetraethylammonium salts obtained from a
2.5
.-Y 3 rr”
1.0’ f'Q
0.5
1ooo
950
Wavenumber
C
900 ( cm-’ )
Fig. 5. Influence of zirconium(IV) concentration on the Raman spectra of 12-molybdophosphoric acid solution. Zirconium concentration: (1) 0; (2) 0.2 X lo-’ M; (3) 0.4 X lo-’ M; (4) 0.6 x lo-’ M; (5) 0.8 X 10va M; (6) 1.0 X lo4 M. Other conditions as for Fig. 2. Fig. 6. Formation curve of molybdozirconophosphoric acid estimated from the two Raman lines observed. (A) Intensity at 980 cm-’ (observed); (B) intensity at 996 cm-’ (observed); (C) intensity at 980 cm-’ (estimated by A-D); (D) intensity of 12-molybdophosphoric acid at 980 cm-’ (estimated from calibration curve).
TABLE 1 Elemental results for tetraethylammonium (All results are given as percentages)
C H N Me P MO
ternary heteropolymolybdate
salts
Molybdotitanophosphate
Molybdozirconophosphate
Molybdohafnophosphat
caica
Found
Caleb
caka
Found
Caleb
Cab?
Found
Caleb
13.3 2.9 1.9 2.2 1.4 48.7
14.7 3.2 2.2 2.2 1.3 47.2
12.5 2.7 1.8 2.1 1.3 49.8
13.0 2.8 1.9 4.1 1.4 47.8
11.9 2.7 1.75 4.9 1.35 48.0
12.3 2.6 1.8 3.9 1.3 49.3
12.5 2.7 1.8 7.8 1.35 45.95
12.2 2.7 1.8 7.0 1.3 46.8
11.8 2.5 1.7 7.3 1.3 47.2
aCaicuIated for [(C,H,),N] ,PMo,,MeO,,H,O, where Me = Ti, Zr or Hf. bCalculated for [(C,H,),N] ,MePMo,,O,,H,O, where Me = Ti, Zr or Hf.
36
solution at pH 1 (Table 1) tend to support reaction (a): PMoIIOg + Me02+ + 2H20 + PMollMeO$, + 4H+
(a)
PMo,O& + Me02+ + Hz0 + PMo12042Me* + 2H’
(b)
with formation of 11-molybdometalophosphates, PMollMeO&,, rather than 12-molybdometalophosphates PMo12042MeJ-. Previously [3] , it was wrongly concluded that the ternary heteropolymolybdate was formed via reaction (b); this might be attributed to the fact that the amounts of ternary heteropolymolybdate formed could not be estimated accurately because of the similarity of the near-ultraviolet spectra of binary and ternary heteropolymolybdates. The present Raman measurements, which can estimate only the ternary heteropolymolybdate, and the elemental analysis of ternary heteropolymolybdate salts support reaction (a). Estimation of the formation constant In the solutions relevant to reaction (a), only PMo~~O&, PMo~~O~Y,and PMo,,MeO& species are active in Raman spectroscopy. The Ti(IV), Zr(IV), and Hf(IV) species have no Raman lines in the same region. The simple spectrometric estimation of the formation constant, K, for the reaction A + (a-x)
B
=+C
(b-x)
x
is based on K = x/(a - x)(b - LX),where a and b are the initial concentrations of A and B, respectively, and, x is the equilibrium concentration of C. If d is the absorbance and p is the molar absorptivity, then d = bAta -x)
+ pB(b -x)
+ h$ = @Aa + PBb) + & --PA -h)x
(1)
Combination of Eqn. 1 and the expression for K yields l/K=[(tl--o)/(p,-_A-PB)I
--a--b+
[ab(Pc--OA-P~/@-~O)I
(2)
where do = @Ac + &b). When B does not absorb, pB = 0, and Eqn. (2) is simplified. The resulting plot is called the Rose-Drago plot [ 121. When PA = /3s = 0, Eqn. (2) becomes even simpler: l/K = (d/P,) -a
- b + (ab&/d)
(3)
When the value of l/K is plotted against the arbitrary /3c for several sets of the initial concentration of A and B, the probable value of l/K is obtainable as the intersection point of several straight lines. This procedure was tested for application with the Raman measurements. The actual formation curve (C) of molybdotitanophosphoric acid in Fig. 3 is well suited to the conditions of the simplified Eqn. (3). The sets of l/K and pc for molybdotitanophosphoric acid are plotted in Fig. 8. The sets of l/K and PC for other ternary heteropolymolybdates are given in Tables 2 and 3. The formation constants of the ternary heteropolymolybdates from 12-molybdophosphoric acid were found from these plots to be 5.4 + 0.1 X lo3 for molybdotitanophosphoric acid, 2.8 + 0.3 X lo4 for molybdozirconophosphoric acid and 1 .l + 0 2 X lo4 for molybdohafnophosphoric acid.
36
2.5
5.0
2.0
4.0
h .< E$I ._
1-5
z ._ 2
1.0
B 0.5
C
0.5 CHfl
) 2.3
1.0
2.5
nc
(10” M 1
Fig. 7. Formation curve of molybdohafnophosphoric lines observed. Curves A-D as for Fig. 6.
2.7
(10”)
acid estimated from the two Raman
Fig. 8. Rose-Drag0 plot for the formation of molybdotitanophosphoric acid. Titanium(IV) concentration: (1) 0.1 X lo-’ M; (2) 0.5 X lo-’ M; (3) 0.9 X lo-” M;(4) 1.0 X lo-’ M. Other conditions as for Fig. 2.
TABLE 2 The sets of l/K and pc for molybdozirconophosphoric
PC
l/K
PC
l/K
PC
l/K
PC
l/K
a-1.0~ 2.355 0.66 a = 1.0 2.32 0.90
lo-‘M,b=0.4X 2.360 1.89 X
lo-‘M 2.365 3.21
lOma M, b = 0.9 X 10-l M 2.34 2.36 1.99 3.19
acid
2.370 4.44
2.380 X 10’ 7.10 x 10-5
2.38 4.48
2.40 X lo* 5.98 x lo-’
a=1.0x10-“M,b=0.95x10-aM 2.30 2.32 0.64 1.33
2.34 2.11
2.36 3.10
2.33 X lo= 4.19 x lo+
a=l.Ox 2.24 0.18
2.28 0.97
2.30 1.62
2.32 X 10’ 2.37 X lo-’
10-aM,b=l.OX1O-‘M 2.26 0.52
37 TABLE3 Thesetsofl/Kandpc formolybdohafnophosphoricacid a=l.OXIOmlM,
PC l/K
2.48 0.12
2.49 0.48
a = 1.0 x lo-‘M, PC I/K
PC
l/K
PC
UK
b=0.086~~~-~M 2.50 0.85 1:22
2.53 1.96
2.54 2.34
2.55X 10' 2.71X lo4
2.52 1.76
2.54 2.19
2.56 2.62
2.58x lOa 3.05x 1o-4
2.48 0.82
2.50 1.00
2.52 1.19
2.54x lo* 1.34x lo4
2.42 0.54
2.44 0.67
2.46 0.81
2.48x lo1 0.96x lo+
b = 0.518 x lOwa M
2.44
2.46
2.46
2.50
0.11
0.51
0.92
1.33
a = 1.0 X IO-‘M, 2.40 2.42 0.19 0.33
2.52 1.59
b = 0.864x 10-l M 2.44 2.46 0.48 0.64
a = 1.0 X IO-"M, b = 0.950x IO-“M 2.34 2.36 2.38 2.40 0.13 0.21 0.30 0.42
Assignment of Raman lines The Raman lines were initially assigned for the mononuclear species MoO$- by Busey and Keller [ 131. The vl, v3, and v4 Raman lines are observed at 897, 841, and 318 cm-‘, respectively. The present authors also observed three Raman lines and confirmed that the Raman line at 897 cm-’ is atotally symmetric vibration by polarization measurements. On acidification of the mononuclear molybdate solution, the Raman spectra change drastically. The v3 and v4 lines disappear. The ul line decreases greatly but a new vl line appears at the higher wavenumber of 940 cm-‘; this was also confirmed to be a totally symmetric vibration by polarization measurements [8] . For isopolymolybdates, Griffith and Lesniak [14 ] assigned bands above 900 cm-’ to Mo=O stretching and those near 350 cm-’ to Mo=O bending; 840-750 cm-’ bands were assigned to asymmetric MO-O-MO stretching and those near 200 cm’ to MO-O-MO deformations 1141. The present authors observed that the vl line of Mo=O stretching shifts towards higher wavenumber as condensation proceeds: Mo,O$, at 940 cm-‘; Mo,O$& at 970 cm-l; and Mo60?;, at 980 cm-‘. These intense Raman lines were also confirmed to be totally symmetric vibrations by polarization measurements [S] . Raman lines below 900 cm-’ are very faint in the isopolymolybdate solutions. Only bending of Mo=O is observed at 365 cm-‘. Deformations of MO-O-MO at 220 cm-’ overlaps with strong Rayleigh scattering, and this line is observable only for high concentrations of molybdate (1.2 M). Stretching of MO-G-MO near 600 cm-’ is not obvious in the isopolymolybdate spectra, but binary and ternary heteropolymolybdates show a characteristic broad Raman peak near 620 cm-‘. Although hexamolybdate also has this Raman band near 600 cm-‘, isopolymolybdate solutions containing hexamolybdate do not give the band because of small amounts of MosO:; [ 151. Some Raman lines of 12-molybdophosphoric acid were discussed by Franck and co-workers [16],
who assigned the Raman line at 996 cm -l to the totally symmetric stretching of Mo=O,. In the present work, the conversion of 12-molybdophosphoric acid to the ternary heteropolymolybdate containing the group 4A element was followed by careful observation of the v1 line. The v1 line of the ternary heteropolymolybdate was observed with a shift of 10-16 cm-‘. Lynhamn et al. attempted to estimate force constants for 12-molybdophosphoric acid [17]. The shift towards lower wavenumber in the ternary heteropolymolybdates may reflect a decrease in the force constant (Mo=O) caused by replacement of a molybdenum atom with a group 4A element. Within the isopoly- and heteropoly-molybdates, 12-molybdophosphoric acid, having the best symmetry, provides the u1 line at the highest wavenumber (996 cm-‘). As shown above, the reaction of the group 4A metal ions with 12-molybdophosphoric acid produces ternary heteropolymolybdates containing the group 4A element. These complexes are more stable than 12-molybdophosphoric acid in the aqueous solution, but they are less extractable. Therefore the presence of the group 4A elements causes interference in the determination of phosphorus by extraction of the heteropoly acid. More basic solvents such as cyclohexanone and tributylphosphate, however, can extract the ternary heteropolymolybdates without interference. As previously reported, spectrophotometric determinations of the group 4A metal ions are possible by using the formation of ternary heteropolymolybdates [3] . REFERENCES 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
P. PakaIns, Anal. Chim. Acta, 40 (1968) 1. M. Ishibashi and M. Tabushi, Bunseki Kagaku, 8 (1959) 588. K. Murata, Y. Yokoyama and S. Ikeda, Anal. Chim. Acta, 48 (1969) 349. A. K. Babko and Y. F. Shkaravskii, Zh. Neorg. Khim., 6 (1961) 2091. M. Borrel and R. Paris, Anal. Chim. Acta, 4 (1950) 267. P. J. Elving and E. C. Olson, Anal. Chem., 27 (1955) 1817. P. J. Hendra and E. J. Loader, Chem. Ind., (1968) 718. K. Murata and S. Ikeda, Spectrochim. Acta, Part A, 39 (1983) in press. K. Murata and S. Ikeda, Polyhedron, 2 (1983) in press. C. Rocchiccioli-Deltcheff and R. Thouvenot, J. Chem. Research (S), (1977) 46. F. Umland, A. Janssen, D. Thierig and G. Wiinsch, Theorie und Praktische Anwendung von Komplexbildenern, Akademische Verlagsgesellschaft, Frankfurt, 1971, p. 42-43. N. J. Rose and R. S. Drago, J. Am. Chem. Sot., 81(1959) 6138. R. H. Busey and D. L. Keller, Jr., J. Chem. Phys., 41(1961) 215. W. P. Griffith and P. J. B. Lesniak, J. Chem. Sot., Sect. A, (1969) 1066. K. Murata, E. Yamamoto and S. Ikeda, Bull. Chem. Sot. Jpn., 56 (3) (1983) 941. C. Rocchiccioli-Deltcheff, R. Thouvenot and R. Franck, Spectrochim. Acta, Part A, 32 (1976) 587. L. Lyhamn, S. J. Cyvin, B. N. Cyvin and J. Brunvoll, Z. Naturforsch., Teil A, 31 (1976) 1589.