Spectrophotometric determination of silicate with Rhodamine B by flow-injection analysis

Analyrica Chimica Acta, 239 (1990) 151-155 Elsevier Science Publishers B.V.. Amsterdam

151

Short Communication

Spectrophotometric determination of silicate with Rhodamine B by flow-injection analysis F. Mas, J.M. Estela and V. Cerda Department

*

of Chemistry, Faculty of Sciences, University of the Balearic Islands, E-07071 Palma de Mallorca (Spain) (Received

5th January

1990)

Abstract A spectrophotometric flow-injection method for the determination of silicate based on the formation of an ion pair between molybdosilicic acid and Rhodamine B is proposed. It allows silicate to be determined over the concentration range 0.17-2.0 mg 1-l at a sampling rate of 40 h-l, is reasonably precise and is highly tolerant to ions that commonly occur in waters. It has been applied with satisfactory results to the determination of silicate in various types of water. Keywords:

Silicate;

Waters

Silicate is an essential element and even a limiting micronutrient for some aquatic microorganisms. It is usually present in ground waters, normally in association with living microorganisms, at concentrations below 1 mg l-i, and in deep waters at levels up to 4 mg 1-l. Its concentration in natural waters typically ranges between 1 and 10 mg l-‘, although it may reach concentrations as high as 50 mg 1-i in pressurized spring waters. The stagnant oligotrophic waters of warm regions contain variable, low concentrations (up to 2 mg 1-l) of silicate as a result of the growth of seasonal algae, and the waters in hot, dry areas may contain up to 30 mg 1-l silicate. As silica is known to have no toxic effects, there is no limit to its tolerance in drinking water, so much so that some recommended spiking water supplies with 10 mg 1-l silica to form rust-proof layers on pipe walls. Dissolved silica has been determined by various methods, most of which are based on the formation of the yellow /3-molybdosilicic acid, a reaction which has been known since 1898 [l]. Orthosilicic acid and its dimer react with molybdate to form cy- or P-molybdosilicic acid, depending on the pH 0003-2670/90/$03.50

0 1990 - Elsevier

Science Publishers

of the medium (2-4 for the former and l-2 for the latter) [2,3]. Both acids can be determined photometrically as such or after reduction to silicomolybdenum blue [4-151. All reference methods for the determination of silica in water are based on the use of this reaction and different reductants [16,17]. Silicon typically occurs in waters as soluble hydrosilicic acid and various colloidal polysilicic acids. These polymers do not react with molybdate, so the determination of total silica requires a prior alkaline digestion. The interference of phosphate is overcome by adding oxalic or tartaric acid, which decomposes molybdophosphoric acid and reacts with the excess of molybdate to form colourless molybdenum oxalate or tartrate complexes. The determinations of silica based on the formation of molybdosilicic acid and silicomolybdenum blue have been automated by using segmented continuous-flow analysis [l&20] and flow-injection analysis in its normal [lo] and reagent-addition modes [21,22]. Basic xanthene dyes, including Rhodamine B, are singly charged cations able to form ion pairs with metal complex anions. Rhodamine B occurs B.V.

152

in different forms in aqueous solutions, depending on the pH. Thus, between pH 0 and 1, the nitrogen of its amino group is protonated, and Rhodamine B occurs as an orange cation (RHz+). Golkowska and co-workers [23-251 studied the formation of the complex ion pair between molybdosilicic acid and Rhodamine B and developed a spectrophotometric method allowing silicate to be determined over the range 15-150 pg 1-l by measuring the absorbance at 590 nm in an aqueous solution or at 555 nm in an ethanol solution after redissolving the complex. There are also reports on the behaviour of the Rhodamine B-silicate system in a sulphuric acid medium [26,27]. In the work reported here, the physico-chemical conditions for the determination of silicate were optimized by using a flow-injection manifold to implement the reaction between molybdosilicic acid and Rhodamine B. Experimental Reagents. Chemicals were obtained from Probus unless indicated otherwise. The following stock solutions were used: 0.2 M molybdate solution, prepared from (NH&M•,0z4. H,O; 1 x 10e3 M solution of Rhodamine B (Merck, pro analysi); 5 M nitric acid; standard phosphate solution (1000 mg l-l), prepared from the corresponding monosodium salt; commercially available 34% sodium silicate solution (Merck); and a 0.5% solution of poly(viny1 alcohol) (PVA), prepared by dissolution of the solid product (Panreac, analytical-reagent grade) in hot water and subsequent filtration. The following working standard solutions were used: 0.024 M molybdate solution in 0.2 M nitric acid; 1 X lop2 M solution of Rhodamine B in 1 M nitric acid containing 0.1% of PVA; and sodium silicate solutions in the range 0.2-2.0 mg 1-l silicate. Apparatus. A Hewlett-Packard Model 8452A diode-array spectrophotometer equipped with an 18-~1 Hellma flow cell was used to monitor the reaction of interest by measuring the absorbance at 590 mn. A Gilson Minipuls 2 eight-channel peristaltic pump was used to pump reactants. All tubing used was of Teflon (0.5 mm i.d.).

F. MbrS ET AL.

Results and discussion Preliminary studies. In a preliminary study we had developed a method for the individual spectrophotometric determination of phosphate by formation of the Rhodamine B-molybdophosphate ion pair. The baseline absorbance was stabilized by using 0.1% PVA, which was included in the Rhodamine B solution in this work. Silicate posed no interference in the determination of phosphate owing to the acidity of the medium used (0.9 M). However, the use of lower acid concentrations allowed the formation of molybdosilicic acid, and thus silicate could be determined. However, at such a decreased pH the reactants have to be isolated to prevent their precipitation, which necessitated the configuration depicted in Fig. 1. Optimization of variables. The influence acidity of the medium was studied by changing the nitric acid concentration in the 0.02 M molybdenum solution over the range 0.04-0.3 M. The signal peaked between 0.18 and 0.21 M nitric acid, i.e., at pH = 1 in the final solution. A nitric acid concentration of 0.2 M was chosen for further experiments. The influence of the molybdate concentration on the analytical signal at a nitric acid concentration of 0.2 M was investigated over the range 0.02-0.33 M. The signal was maximum for a 0.024 M molybdenum concentration, which was equivalent to an [H+]/[Mo] ratio of 8.3. The influence of the reaction tube length (L, in Fig. 1) was also studied over the range l-6 m. A

Fig. 1. Manifold for the determination of silicate. C = Carrier (water); R, = molybdate solution; R, = Rhodamine B-PVA solution; L, = 4 m; L, = 56 cm; L, = 1 m; Ox (dashed line) = optional channel for oxalic acid addition; M = sample; A = point to merge oxalic acid with the heteropoly acid.

153

FIA OF SILICATE

length of about 4 m was found to yield the optimum signal in this respect. The influence of the Rhodamine B concentration on the signal was studied over the range 0.5 x 10p4-1.8 X 10m4 M. The signal peaked between 0.9 x lop4 and 1.2 X lop4 M and a working concentration of 1 X 10e4 M was adopted. This solution was 1 M in HNO, and contained 0.1% PVA. As the analytical signal changed with the nitric acid content in this solution, its influence was studied within the range 0.3-l M. A 1 M concentration of nitric acid in Rhodamine B solution was chosen for subsequent experiments. The reaction of molybdosilicic acid with Rhodamine B was much faster than the formation of the heteropoly acid. A 56-cm reactor length (L2 in Fig. 1) was optimum for the ion-pair formation. The temperature affected the reaction rate, which was maximum between 45 and 50°C. 1nnterferences. Phosphate is a serious interferent in the determination of silicate. This, and the claims by some workers [28-301 that the nature of the acid used affects the reaction kinetics, prompted us to study the potential effect of various acids commonly used in the determination of silicate and of phosphate, using the manifold shown in Fig. 1. The results obtained are given in Table 1. The silicate signal was greatest when using nitric or perchloric acid; however, the latter caused Rhodamine B to precipitate. The effects of addition of oxalic acid on the signals for phosphate and silicate can be seen in Table 2. It was concluded that adding a channel allowing > 0.1 M oxalic acid to merge with the heteropoly acid at point A in Fig. 1, prior to reaction with the

Acid

HNO, Hz.504

HClO, HCl

2

Effect of the oxalic acid concentration W2G041

Absorbance

(W

0.00 0.01 0.02 0.05 0.07

0.07 0.10 0.20 0.50

of the acid used on the analytical Absorbance

Silicate

0.238 0.234 0.190 0.215

0.180 0.160 0.181 0.153 cm T=28”C, = [SiO:-] =l

Asi /‘AP a

Phosphate

lmgl-’

1 mg 1-l

0.216 0.270 0.252 0.234 0.252

0.065 0.081 0.065 0.032 0.009

1 mg 1-l

5 mg 1-l

0.252 0.221 0.222 0.225

0.023 0.014 0.014 0.014

3.3 3.3 3.9 7.4 28.0

11.2 16.4 16.5 16.9

absorbances.

TABLE

3 of interferents

in the determination

of 5 mg

Species

(mg 1-l)

Phosphate

a ~,=2 m, L2=50 [H+l a,,od =l M, [PO:-]

(A)

Rhodamine B, decreases the phosphate response to a very small value, whilst having little effect on the silicate response. To show the effectiveness of the oxalic acid masking, a solution containing 1 pg ml-’ each of silicate and phosphate gave a signal equivalent to 1.3 pg ml-’ silicate in the absence of oxalic acid and 1.1 pg ml-’ in the presence of oxalic acid. The responses for a solution containing 1 pg ml-’ silicate and 5 pg ml-’ phosphate corresponded to 2.5 pg ml-’ silicate in the absence of masking agent, and 1.2 pg ml-’ in its presence. Up to 5 pg ml-’ phosphate can be tolerated without masking. Table 3 gives the concentrations of each species below which no interference was detected ( f 2~)

Tolerated amount

signal a

of the signal

Silicate

a Ratio of peak-height

Tolerated amounts 1-l silicate

TABLE1 Influence

TABLE

[H+]M,=0.24 mg 1-l.

500 100 50 10 5 M,

a Positive sponses.

Tartrate, ClOi a SO:-, Br- a, NO; a, Cl-, K+, Na+, MgZC Co(H), Zn(II), Cr(III), Ca(II), EDTA FAs(V) a, I-, Fe(III), H&,0, interferences;

all other

species

gave

decreased

re-

F. M&i

154

for the determination of 5 mg 1-l silicate. A maximum ratio of interferent to analyte of 20 : 1 (100 mg 1-l of interferent) was examined, except for tartrate and perchlorate. Fluoride was only tolerated at concentrations below 10 mg l-‘, whereas As(V), Cu(II), Pb(II), V(V), iodide, Fe(II1) and oxalic acid were tolerated in ratios up to 1: 1; greater amounts were not tested as they are rarely found in waters. Amounts of EDTA 1000 times greater than that of silicate and salt contents above 0.15 M interfered with the determination. Calibration. Under the optimum working conditions used (L, = 4 mm, L, = 56 cm, 50 o C, 0.024 M MO-0.2 M HNO,, 1 x 10e4 M Rhodamine B-l M HNO,-0.1% PVA, volume injected 220 ~1 and total-flow rate 2.65 ml mm-‘), the calibration graph was linear over the range 0.17-2.0 mg 1-l silicate, with a correlation coefficient of 0.9996 (six, points) and an equation H = 0.17 + 0.2216 [SiOi-] (mg l-l), where H is the peak-height absorbance. The detection limit achieved for a signal-to-noise ratio of 3 was 0.06 mg 1-l silicate, and the sampling rate was 40 h-i. Twelve replicate injections of a standard of 0.5 mg 1-l yielded an average result of 0.56 mg 1-l and a standard deviation of 0.06 mg 1-l. Application to real samples. The proposed method was applied to the determination of silicate in various types of water from different sources. Table 4 compares the results obtained with those provided by the conventional molybdo-

TABLE 4 Determination of silicate in real water samples Sample

Tap 1 Tap 2 Tap 3 Tank 1 Tank 2 Bottled 1 Bottled 2 Bottled 3 Spring Waste

Proposed method Dilution

[SiOf-] (mg I-‘)

10:50 5:50 5:50

10.22 11.82 14.56 3.26 3.30 72.80 41.47 11.82 16.91 28.00

1:50 1:50 5:50 5:50 5:50

Reference method Dilution

_ _ 25:50

5:50

[SiOz- ] (mg I-‘) 11.90 11.95 14.47 3.37 3.41 71.93 39.52 11.87 17.29 30.20

ET AL.

silicic acid method [17]. With the exception of water from tanks, samples had to be diluted to accommodate their analytical signals within the linear range of the method. Conclusions The proposed flow-injection spectrophotometric method for the determination of silicate in water, which relies on the use of Rhodamine B, is more sensitive than the molybdosilicic acid method [lo], the applicability range of which is 2-10 mg 1-l silica. However, it is slower (40 h-’ compared with 60 h-’ in the latter). Thomsen et al. [22] increased the sensitivity (up to 14 or 3 pg 1-l Si, depending on the experimental conditions) by using a reagent-injection system and detecting the molybdenum blue formed at 886 nm with a flow cell of light path 2 cm. Considering the concentration levels at which silica normally occurs in natural waters, the sensitivity, precision and reproducibility of the proposed method are more than adequate. A prior dilution of the sample to be assayed is required in most instances, although this can be avoided by exploiting some non-optimized variable such as a smaller injection volume or a tube length shorter than 4 m for the mixing of silicate and molybdate. On the other hand., the proposed method is subject to the same interferences as others based on the formation of molybdosilicic acid, chiefly fluoride, As(V), Cu(II), Pb(II), Fe(II1) and V(V), which usually occur in waters at concentrations that are tolerated by this method. Thanks are due to the DGICyT (Spanish Council for Research in Science and Technology) for financial support (PA86-0033).

REFERENCES 1 H. Jolles and F.Z. Neurath, Angew. Chem., 11 (1898) 315. 2 J.D.H. Strickland, J. Am. Chem. Sot., 74 (1952) 862, 868, 872. 3 V.W. Truesdale and C.J. Smith, Analyst, 100 (1975) 203. 4 M.L. Isaacs, Bull. Sot. Chim. Biol., 6 (1924) 157. 5 L. Trudell and D.F. Boltz, Anal. Chem., 35 (1963) 2122. 6 A.L. Wilson, Analyst, 90 (1965) 270. 7 L.G. Hrgis, Anal. Chim. Acta, 52 (1970) 1. 8 P. Pakalns, Anal. Chim. Acta, 54 (1971) 281.

FIA OF SILICATE

N. Yoza, T. Tarutani and S. 9 Y. Hirai, T. Yokoyama, Ohashi, Bunseki Kagaku, 30 (1981) 350. 10 Y. Hirai, T. Yokoyama, N. Yoza, T. Tarutani and S. Ohashi, Bull. Chem. Sot. Jpn., 55 (1982), 3477. 11 A. Halasz, K. ‘Polyak and E. Pungor, Tahmta, 18 (1971) 691. 12 A. Golkowsa and L. Pszonicki, Talanta, 20 (1973) 749. 13 M.T. Beshikdash’yan and M.G. Vasil’eva, J. Anal. Chem. USSR (Engl. Trans.), 36 (1981) 747. 14 F.A.J. Armstrong, J. Mar. Biol. Assoc. UK, 30 (1951) 149. 15 K. Kato, Anal. Chim. Acta, 82 (1976) 401. 16 American Public Health Association, Standard Methods for the Examination of Water and Wastewater, 13th edn., APHA, Washington, DC, 1971, p. 306. 17 .American Society for Testing and Materials, Annual Book of ASTM Standards, ASTM, Philadelphia, 1975, pp. D859-868, B398-399.

155 18 K. Grasshoff, Fresenius’ 2. Anal. Chem., 220 (1966) 89. 19 P.G. Brewer and J.P. Riley, Anal. Chim. Acta, 35 (1966) 514. 20 V.W. Truesdale and C.J. Smith, Analyst, 101 (1976) 19. 21 D. Betteridge, Anal. Chem., 50 (1978) 832A. 22 J. Thomsen, K.S. Johnson and R.L. Petty, Anal. Chem., 53 (1983) 2378. 23 A. Golkowska and L. Pszonikci, Talanta, 20 (1973) 749. 24 A. Golkowska, A. Chem. Anal. (Warsaw), 14 (1969) 803. 25 A. Golkowska, Chem. Anal. (Warsaw), 15 (1970) 59. 26 P. Shi, Huaxue Shiji, 5 (1983) 374. 27 C. Liu, J. Guo and Y. Wei, Fenxi Huaxue, 18 (1988) 211. 28 S.R. Crouch and H.V. Malmstadt, Anal. Chem., 39 (1976) 1084. 29 C.C. Kicher and S.R. Crouch, Anal. Chem., 55 (1983) 248. 30 P. Linares, M.D. Luque de Castro and M. Valcarcel, Talanta, 33 (1986) 889.