- Email: [email protected]
Automated Flow-Injection Pseudotitration of Boric Acid
Automated Flow-Injection Pseudotitration of Boric Acid PARASKEVI
I. ANAGNOSTOPOULOU AND MICHAEL A. KOUPPARIS’
Received November 29, 1984, from the Laboratory of Analytical Chemistty, University of Athens, Athens 106 SO, Greece. publication April 17, 1985. -
Abstract 0 An automated flow-injectionpseudotitrimetricdetermination of boric acid, using a flow-injectionphotometric analyzer controlled by a microcomputer, is described. The method is based on the injection of
200 pL of sample in a flowing stream of mannitol-bromothymol blue ”titrant”and measuring the peak width in time units. Equivalent times of 10-70 s are measured with GV values of 0.1-0.4% (n = 5) and the analytical range is 1.5409 mgil00 mL (2.5x 10-4-5 x lo-’ M) of boric acid. The method was evaluated by performing recovery studies in mixtures (mean 99.8%)and assays of commercial preparations which were compared with the official classical titrimetric method.
Flow-injection analysis (FIA) is a rapidly developing new technique for automated performance of many routine chemical ana1yses.l The basis of FIA is the controlled, reproducible, flow-induced dispersion obtainable when a sample plug is injected into a reagent carrier stream as it flows through narrow-bore tubing. The changing concentration profile of the sample solution or the concentration of product may be monitored by downstream detectors. The new technique is flexible, allowing performance of a variety of analytical manipulations (mixing of reagents, dilution, dialysis, extraction, heating or cooling, reduction with solid reactors, ion-exchange, etc.) in an unsegmented continuous-flow mode. The only requirement is that a measurable analytical signal is obtained in 1-2 min. All analytical detection devices provided with a flow cell can be interfaced with an FIA system.2-4 The adaptation of titrimetric analysis in the FIA approach has been shown t o be an attractive new technique and has been studied and used in several application^."^ In this FIA concept the sample of volume V, is forced to a large dispersion in the reagent flow (the titrant), of concentration and flow rate, Q, using a mixing chamber of volume V,. The peak
Accepted for
width, measured in time units, is linearly related to the logarithm of the analyte concentration C::
Ateq
=
(Vm/Q)- In 10 * log (C:/C,“ n) f (Vm/Q) X In 10 log (VJV,) (1)
-
or:
where teq refers to “equivalence” time, Atq refers to the interval of teq, n is the molar ratio of the reacting components, and kl and 122 are experimental constants of the FIA system, the reaction and titrant concentration used. This new technique has been called “flow-injection pseudotitration.” The major advantages are the high sample throughput (more than 60/h in most cases), the low sample and reagent consumption, the use of unstable reagents prepared in situ, and the automation of all the titration steps. In this paper, the adaptation of boric acid titrimetric determination, based on complexation with polyhydroxy compounds, to the FIA pseudotitrimetric concept is described. This procedure minimizes the large consumption of mannitol or glycerol used in the conventional titrimetric procedure and shows the usefulness of FIA in pharmaceutical analyses.
Experimental Section Apparatus-The apparatus used for the flow-injection pseudotitration of boric acid was the automated system described in detail elsewhere8 and shown i n Fig. 1. This system provided for continuous aliquot removal of standard and sample solutions and storage in a 200-pLsample loop. During the injection step, the sample zone was inserted into the continuously pumped carrier reagent (titrant) and mixed thoroughly in a 1.0-mL mixing chamber. This type of mixing
Pump
Figure 1-Schematic diagram of the automated flow-injection pseudotitration system. Key: (FC) flow cell; (W) waste. 886 /Journal of Pharmaceutical Sciences Vol. 74, No. 8, August 1985
0022-3549/85/0800-0886$0 1.00/0 0 1985, American Pharmaceutical Association
caused a large dispersion of the sample, resulting in widened transmittance peaks, the width of which, in time units, was measured precisely by the microcomputer. The triggering level [transmittqnce level (T) to start the teq measurement] was set arbitrarily a t 0.40 T, which corresponds to one-half of the height of the most dilute standard peak. All the operations of the system, including removal of aliquots, sample injection, the measuring of time, and the statistical treatment were controlled or performed by the microcomputer. The pseudotitration peaks were also recorded on a chart recorder. Reagents-Boric acid stock solution, 0.1000 M, was prepared by dissolving 1.546 g of boric acid in 250 mL of deionized, distilled water. Working standard solutions, 2.50 x 10-4-5.00 x 10' M (1.55-309 mg/100 mL), were prepared by appropriate dilutions. The "titrant" consisted of mannitol, 7.5 g/lOO mL, and bromothymol blue indicator, 6.24 mg/100 mL (1.0 x M), in phosphate buffer, 1.0 x M, adjusted to pH 7.0. All the reagents used were of analytical grade. Sample Preparation-Samples were prepared from pharmaceutical preparations (dusting powders, ointments, ophthalmic solutions, and dermatological preparations) by dissolving accurately measured amounts of a sample in water, filtering if necessary, and diluting so that the boric acid concentration was in the range of 2.5 x 10-4-5.0 x lo-' M. Sample solutions containing other acids or bases were roughly neutralized with 0.010 M NaOH or HC1 usingp-nitrophenol indicator. Procedur+The buffered mannitol-bromothymol blue titrant was pumped continuously at a rate of 5.0 mL/min. The photometer was set a t 614 nm, the A,, of the basic form of the bromothymol blue indicator. The absorbance of the indicator in the basic form (at pH 7.0 one-half of the indicator is in basic form), equal to 0.648 (0.225 T), served as the baseline. The operator provided the direction of peaks (increasing transmittance), the injection time (5 s), the number and concentrations of standards, the number of runs per standard and sample, and the number of samples. The standards and samples were loaded on the turntable or manually transferred to the sample probe. The determination was accomplished automatically by measuring the standards provided, calculating the calibration curve, and measuring the unknown samples.
Results and Discussion In order to choose the optimum experimental parameters for better sensitivity, precision, and increased sample throughput, mannitol, sorbitol, and glycerol in various concentrations were tested as titrants. Glycerol was the less sensitive titrant, while mannitol and sorbitol gave equal results. Titrants with mannitol concentrations of 0.27-0.54 M (5-10 g/lOO mL) were tested, and the optimum sensitivity was found with a concentration of 0.41 M (7.5 gA00 mL). The titrant flow rates tested, 3-6 mL/min, showed no effect on the peak height, but teq decreased and sample throughput increased by increasing the flow rate. The 5.0 mL/min flow rate chosen was a compromise between high sample throughput and low titrant consumption. Bromothymol blue was chosen as the indicator, because of its pK, value (7.01, its stability, and the distinction between the basic and acidic color. The indicator concentration used, 1.0 x M, was a compromise between a high absorbance baseline and being dilute enough so that small boric acid concentrations (2.5 x M) were able to change the indicator color. The buffer capacity was kept as low as possible for baseline stability. Typical flow-injection pseudotitration curves of boric acid, using the optimum experimental conditions described, are shown in Fig. 2. Table I shows results for the boric acid working curve with excellent precision, 0.1-0.4% (n = 51, in teq measurements and linearity in a wide analytical range of 1.55-309 mg/100 mL (2.50 x 10-4-5.00 x lo-' M). The teq measured varied from 14 to 70 s. The automated flow injection pseudotitrimetric method was evaluated by performing recovery experiments from mixtures containing 1.OO x M boric acid with various acids and substances which are present in commercial preparations (Table 11). Recovery
varied from 97 to 103% (mean 99.8%)with SD values from 0.3 to 1.0 (in percent recovery, n = 3). The new method was also applied to the determination of boric acid in commercial preparations, and the results were compared with the official titrimetric m e t h ~ dThe . ~ results showed excellent agreement (Table 111).
8 F
t
60s 1
t-
Figure 2- Flow-injectionpseudotitration curves of boric acid. Boric acid concentrations: (a) 1.55;(b) 3.09;(c) 6.18;(d) 30.9;(e) 61.8;(f) 309 mg/lOO mL. Table I-Results
for the Boric Acid Working Curve'
Peak Width, s
Conc. mg/lOO mL 1.55 3.09 6.18 30.9 61.8 309
Range
Average
14.16-14.25 21.65-21.79 28.82-28.94 45.67-45.85 53.10-53.20 70.34-70.57
14.20 21.73 28.85 45.76 53.15 70.43
'Working curve: Aqs) Table 11-Recovery
= 9.602
0.4 0.3 0.2 0.2 0.1 0.1
Results of Boric Acid' Mixtures
Acetic acid 1.0 X Amylose 0.1 Benzoic acid 1.0 X 8-Hydroxyquinoline 2.0 x Lactose 0.1 Propionic acid 1.0 x Salicylic acid 1.0 X Tartaric acid 1.0 X
20.25 20.91 20.90 21.21 20.80 21.19 20.51 20.76
'Boric acid concentration 1.OO all values are 2 SD. Table Ill-Boric Preparations
%b
+ 24.37 log C; r = 0.99998. b n = 5
Additive Average Peak Conc., M Width, s
Additive
cv,
x
cv,yob
Recovery, YO
0.4 0.2 0.1 0.4 0.2 0.5 0.4 0.1
'
97.0 f 0.7 100.5 5 0.4 100.4 t 0.3 101.5 2 0.6 99.4 f 0.4 103.0 2 1.O 96.7 2 0.7 99.1 t 0.3
M. b n = 3. 'Mean = 99.8%;
Acid Assay Results from Commercial
Preparation
Concentration
Ophthalmic solution Liquid soap Nasal ointment Borax solution
0.7 2.5 1.o
1 .o
Assay, % Flow Injectiona 1.028 + 0.001 0.703 t 0.007 2.77 t 0.04 0.984 t 0.004
Officialb 1.020 t 0.006 0.712 t 0.012 2.80 t 0.08 0.986 t 0.020
'Average of five determinations (+ SD). bAverage of two determinations. Journal of Pharmaceutical Sciences / 887 Vol. 74, No. 8, August 1985
In conclusion, the new flow-injection pseudotitrimetric method for the determination of boric acid shows high precision, low reagent consumption, sensitivity, high sample throughput (more than 60 measurementsh), qnd automation of all the determination steps. me flow-in@tion pseudotitrimetric technique is used here to minimize the high consumption of mannit01 Or glycerol. However, in boric acid determinations the general application of this new concept in pharmaceutical analysis is promising, especially in adaptstion of nonaqueous titrations where expensive and toxic solvents are used in large volumes. Experiments in progress for automated nonaqueous acid-base FIA titrations showed excellent results, but some technical problems have arisen concerning the lifetime of the silicone rubber tubing of the peristaltic pump. A displacement delivery System for nonaqueous titrants seems to be the solution to this problem.
888 / Journal of Pharmaceutical Sciences Vol. 74, No. 8,August 1985
References and Notes 1. Ruzicka, Jaromir; Hansen, El0 H. “Flow Injection Analysis”; 2, Wiley: Ranger,New York, M.Anal. 1981.Chem. 1981,53,20A32A. 3. Rocks, Bernard; Riley, Clifford Clin. Chem. 1982,28,409-421. 4. Stewart, Kent K.Anal. Chem. 1983,55,931A-940A. 5. Rwicka, J.; Hansen, E. H.; Mosbaek, H. Anal. Chim. Acta 1977, 92,235-250. 6 . Stewad, Kent K.; Brown, j,F,; (&iden, B. M,Anal. Chim. Acts 1980,114,119-127. 7. Stewart, Kent, K.; Rosenfeld, A. G. J . Automat. Chem. 1981,3, 3032. 8. KoupPafis, M.A.; Ana@ostoPO‘-‘h p.; Malmstadt, H.v. Talanta 1985,32,411. 9. ~ J . SPhamaco . eia xx and National ~~~~l~~ xv”; u,s. Pharmacopeial Zonvention: Rockville, MD, 1980; p 1213.
I. ANAGNOSTOPOULOU AND MICHAEL A. KOUPPARIS’
Received November 29, 1984, from the Laboratory of Analytical Chemistty, University of Athens, Athens 106 SO, Greece. publication April 17, 1985. -
Abstract 0 An automated flow-injectionpseudotitrimetricdetermination of boric acid, using a flow-injectionphotometric analyzer controlled by a microcomputer, is described. The method is based on the injection of
200 pL of sample in a flowing stream of mannitol-bromothymol blue ”titrant”and measuring the peak width in time units. Equivalent times of 10-70 s are measured with GV values of 0.1-0.4% (n = 5) and the analytical range is 1.5409 mgil00 mL (2.5x 10-4-5 x lo-’ M) of boric acid. The method was evaluated by performing recovery studies in mixtures (mean 99.8%)and assays of commercial preparations which were compared with the official classical titrimetric method.
Flow-injection analysis (FIA) is a rapidly developing new technique for automated performance of many routine chemical ana1yses.l The basis of FIA is the controlled, reproducible, flow-induced dispersion obtainable when a sample plug is injected into a reagent carrier stream as it flows through narrow-bore tubing. The changing concentration profile of the sample solution or the concentration of product may be monitored by downstream detectors. The new technique is flexible, allowing performance of a variety of analytical manipulations (mixing of reagents, dilution, dialysis, extraction, heating or cooling, reduction with solid reactors, ion-exchange, etc.) in an unsegmented continuous-flow mode. The only requirement is that a measurable analytical signal is obtained in 1-2 min. All analytical detection devices provided with a flow cell can be interfaced with an FIA system.2-4 The adaptation of titrimetric analysis in the FIA approach has been shown t o be an attractive new technique and has been studied and used in several application^."^ In this FIA concept the sample of volume V, is forced to a large dispersion in the reagent flow (the titrant), of concentration and flow rate, Q, using a mixing chamber of volume V,. The peak
Accepted for
width, measured in time units, is linearly related to the logarithm of the analyte concentration C::
Ateq
=
(Vm/Q)- In 10 * log (C:/C,“ n) f (Vm/Q) X In 10 log (VJV,) (1)
-
or:
where teq refers to “equivalence” time, Atq refers to the interval of teq, n is the molar ratio of the reacting components, and kl and 122 are experimental constants of the FIA system, the reaction and titrant concentration used. This new technique has been called “flow-injection pseudotitration.” The major advantages are the high sample throughput (more than 60/h in most cases), the low sample and reagent consumption, the use of unstable reagents prepared in situ, and the automation of all the titration steps. In this paper, the adaptation of boric acid titrimetric determination, based on complexation with polyhydroxy compounds, to the FIA pseudotitrimetric concept is described. This procedure minimizes the large consumption of mannitol or glycerol used in the conventional titrimetric procedure and shows the usefulness of FIA in pharmaceutical analyses.
Experimental Section Apparatus-The apparatus used for the flow-injection pseudotitration of boric acid was the automated system described in detail elsewhere8 and shown i n Fig. 1. This system provided for continuous aliquot removal of standard and sample solutions and storage in a 200-pLsample loop. During the injection step, the sample zone was inserted into the continuously pumped carrier reagent (titrant) and mixed thoroughly in a 1.0-mL mixing chamber. This type of mixing
Pump
Figure 1-Schematic diagram of the automated flow-injection pseudotitration system. Key: (FC) flow cell; (W) waste. 886 /Journal of Pharmaceutical Sciences Vol. 74, No. 8, August 1985
0022-3549/85/0800-0886$0 1.00/0 0 1985, American Pharmaceutical Association
caused a large dispersion of the sample, resulting in widened transmittance peaks, the width of which, in time units, was measured precisely by the microcomputer. The triggering level [transmittqnce level (T) to start the teq measurement] was set arbitrarily a t 0.40 T, which corresponds to one-half of the height of the most dilute standard peak. All the operations of the system, including removal of aliquots, sample injection, the measuring of time, and the statistical treatment were controlled or performed by the microcomputer. The pseudotitration peaks were also recorded on a chart recorder. Reagents-Boric acid stock solution, 0.1000 M, was prepared by dissolving 1.546 g of boric acid in 250 mL of deionized, distilled water. Working standard solutions, 2.50 x 10-4-5.00 x 10' M (1.55-309 mg/100 mL), were prepared by appropriate dilutions. The "titrant" consisted of mannitol, 7.5 g/lOO mL, and bromothymol blue indicator, 6.24 mg/100 mL (1.0 x M), in phosphate buffer, 1.0 x M, adjusted to pH 7.0. All the reagents used were of analytical grade. Sample Preparation-Samples were prepared from pharmaceutical preparations (dusting powders, ointments, ophthalmic solutions, and dermatological preparations) by dissolving accurately measured amounts of a sample in water, filtering if necessary, and diluting so that the boric acid concentration was in the range of 2.5 x 10-4-5.0 x lo-' M. Sample solutions containing other acids or bases were roughly neutralized with 0.010 M NaOH or HC1 usingp-nitrophenol indicator. Procedur+The buffered mannitol-bromothymol blue titrant was pumped continuously at a rate of 5.0 mL/min. The photometer was set a t 614 nm, the A,, of the basic form of the bromothymol blue indicator. The absorbance of the indicator in the basic form (at pH 7.0 one-half of the indicator is in basic form), equal to 0.648 (0.225 T), served as the baseline. The operator provided the direction of peaks (increasing transmittance), the injection time (5 s), the number and concentrations of standards, the number of runs per standard and sample, and the number of samples. The standards and samples were loaded on the turntable or manually transferred to the sample probe. The determination was accomplished automatically by measuring the standards provided, calculating the calibration curve, and measuring the unknown samples.
Results and Discussion In order to choose the optimum experimental parameters for better sensitivity, precision, and increased sample throughput, mannitol, sorbitol, and glycerol in various concentrations were tested as titrants. Glycerol was the less sensitive titrant, while mannitol and sorbitol gave equal results. Titrants with mannitol concentrations of 0.27-0.54 M (5-10 g/lOO mL) were tested, and the optimum sensitivity was found with a concentration of 0.41 M (7.5 gA00 mL). The titrant flow rates tested, 3-6 mL/min, showed no effect on the peak height, but teq decreased and sample throughput increased by increasing the flow rate. The 5.0 mL/min flow rate chosen was a compromise between high sample throughput and low titrant consumption. Bromothymol blue was chosen as the indicator, because of its pK, value (7.01, its stability, and the distinction between the basic and acidic color. The indicator concentration used, 1.0 x M, was a compromise between a high absorbance baseline and being dilute enough so that small boric acid concentrations (2.5 x M) were able to change the indicator color. The buffer capacity was kept as low as possible for baseline stability. Typical flow-injection pseudotitration curves of boric acid, using the optimum experimental conditions described, are shown in Fig. 2. Table I shows results for the boric acid working curve with excellent precision, 0.1-0.4% (n = 51, in teq measurements and linearity in a wide analytical range of 1.55-309 mg/100 mL (2.50 x 10-4-5.00 x lo-' M). The teq measured varied from 14 to 70 s. The automated flow injection pseudotitrimetric method was evaluated by performing recovery experiments from mixtures containing 1.OO x M boric acid with various acids and substances which are present in commercial preparations (Table 11). Recovery
varied from 97 to 103% (mean 99.8%)with SD values from 0.3 to 1.0 (in percent recovery, n = 3). The new method was also applied to the determination of boric acid in commercial preparations, and the results were compared with the official titrimetric m e t h ~ dThe . ~ results showed excellent agreement (Table 111).
8 F
t
60s 1
t-
Figure 2- Flow-injectionpseudotitration curves of boric acid. Boric acid concentrations: (a) 1.55;(b) 3.09;(c) 6.18;(d) 30.9;(e) 61.8;(f) 309 mg/lOO mL. Table I-Results
for the Boric Acid Working Curve'
Peak Width, s
Conc. mg/lOO mL 1.55 3.09 6.18 30.9 61.8 309
Range
Average
14.16-14.25 21.65-21.79 28.82-28.94 45.67-45.85 53.10-53.20 70.34-70.57
14.20 21.73 28.85 45.76 53.15 70.43
'Working curve: Aqs) Table 11-Recovery
= 9.602
0.4 0.3 0.2 0.2 0.1 0.1
Results of Boric Acid' Mixtures
Acetic acid 1.0 X Amylose 0.1 Benzoic acid 1.0 X 8-Hydroxyquinoline 2.0 x Lactose 0.1 Propionic acid 1.0 x Salicylic acid 1.0 X Tartaric acid 1.0 X
20.25 20.91 20.90 21.21 20.80 21.19 20.51 20.76
'Boric acid concentration 1.OO all values are 2 SD. Table Ill-Boric Preparations
%b
+ 24.37 log C; r = 0.99998. b n = 5
Additive Average Peak Conc., M Width, s
Additive
cv,
x
cv,yob
Recovery, YO
0.4 0.2 0.1 0.4 0.2 0.5 0.4 0.1
'
97.0 f 0.7 100.5 5 0.4 100.4 t 0.3 101.5 2 0.6 99.4 f 0.4 103.0 2 1.O 96.7 2 0.7 99.1 t 0.3
M. b n = 3. 'Mean = 99.8%;
Acid Assay Results from Commercial
Preparation
Concentration
Ophthalmic solution Liquid soap Nasal ointment Borax solution
0.7 2.5 1.o
1 .o
Assay, % Flow Injectiona 1.028 + 0.001 0.703 t 0.007 2.77 t 0.04 0.984 t 0.004
Officialb 1.020 t 0.006 0.712 t 0.012 2.80 t 0.08 0.986 t 0.020
'Average of five determinations (+ SD). bAverage of two determinations. Journal of Pharmaceutical Sciences / 887 Vol. 74, No. 8, August 1985
In conclusion, the new flow-injection pseudotitrimetric method for the determination of boric acid shows high precision, low reagent consumption, sensitivity, high sample throughput (more than 60 measurementsh), qnd automation of all the determination steps. me flow-in@tion pseudotitrimetric technique is used here to minimize the high consumption of mannit01 Or glycerol. However, in boric acid determinations the general application of this new concept in pharmaceutical analysis is promising, especially in adaptstion of nonaqueous titrations where expensive and toxic solvents are used in large volumes. Experiments in progress for automated nonaqueous acid-base FIA titrations showed excellent results, but some technical problems have arisen concerning the lifetime of the silicone rubber tubing of the peristaltic pump. A displacement delivery System for nonaqueous titrants seems to be the solution to this problem.
888 / Journal of Pharmaceutical Sciences Vol. 74, No. 8,August 1985
References and Notes 1. Ruzicka, Jaromir; Hansen, El0 H. “Flow Injection Analysis”; 2, Wiley: Ranger,New York, M.Anal. 1981.Chem. 1981,53,20A32A. 3. Rocks, Bernard; Riley, Clifford Clin. Chem. 1982,28,409-421. 4. Stewart, Kent K.Anal. Chem. 1983,55,931A-940A. 5. Rwicka, J.; Hansen, E. H.; Mosbaek, H. Anal. Chim. Acta 1977, 92,235-250. 6 . Stewad, Kent K.; Brown, j,F,; (&iden, B. M,Anal. Chim. Acts 1980,114,119-127. 7. Stewart, Kent, K.; Rosenfeld, A. G. J . Automat. Chem. 1981,3, 3032. 8. KoupPafis, M.A.; Ana@ostoPO‘-‘h p.; Malmstadt, H.v. Talanta 1985,32,411. 9. ~ J . SPhamaco . eia xx and National ~~~~l~~ xv”; u,s. Pharmacopeial Zonvention: Rockville, MD, 1980; p 1213.