Sequential injection spectrophotometric determination of phenylephrine hydrochloride in pharmaceutical preparations

Talanta 63 (2004) 599–604

Sequential injection spectrophotometric determination of phenylephrine hydrochloride in pharmaceutical preparations Negussie W. Beyene, Jacobus F. Van Staden∗ Department of Chemistry, University of Pretoria, Pretoria 0002, South Africa Received 15 October 2003; received in revised form 26 November 2003; accepted 27 November 2003 Available online 8 February 2004

Abstract A fully automated sequential injection spectrophotometric method for the determination of phenylephrine hydrochloride in pharmaceutical preparations is reported. The method is based on the condensation reaction of the analyte with 4-aminoantipyrine in the presence of potassium ferricyanide. The absorbance of the condensation product was monitored at 503 nm. A linear relationship between the relative peak height and concentration was obtained in the range 0.5–17.5 mg l−1 . The detection limit (as 3σ value) was 0.09 mg l−1 and repeatability was 0.8 and 0.6% at 2.5 and 5 mg l−1 , respectively. Results obtained by this method agreed very well with those obtained by the AOAC official method. © 2003 Elsevier B.V. All rights reserved. Keywords: Flow systems; Sympathomimetic drugs; Nasal decongestants; Ophthalmic solutions

1. Introduction Phenylephrine hydrochloride [R-2-methylamino-1-(3-hydroxyphenyl)ethanol hydrochloride], I, is a non-specific sympathomimetic agent that stimulates alpha-adrenergic receptors, producing pronounced vasoconstriction. It is an active ingredient in nasal decongestants relieving congestion caused by allergies, colds, sinus, or ear infections. It also acts as a mydriatic in ophthalmic solutions relieving eye redness caused by colds, hay fever, wind, dust, sun, smog, smoke or contact lenses. It can also be administered along with local anaesthetics to prolong the effect of anaesthesia [1,2].

Most of the analytical methods available in the literature for the determination of I are spectrophotometric. The colorimetric method recommended by AOAC International is based on oxidation of I by potassium ferricyanide ∗ Corresponding author. Tel.: +27-12-420-2515; fax: +27-12-362-5297. E-mail address: [email protected] (J.F. Van Staden).

0039-9140/$ – see front matter © 2003 Elsevier B.V. All rights reserved. doi:10.1016/j.talanta.2003.11.041

and the reaction of the resulting oxidation product with 4-aminoantipyrine in borate buffer. The final product is antipyrine-phenylephrine that absorbs at 490 nm [3]. This manual procedure is time consuming and prone to errors. An air-segmented continuous flow automated method based on the ferricyanide–aminoantipyrine reactions was also considered as the AOAC official method for the determination in drugs [4]. Mestre et al. [5] reported a flow injection chemiluminometric method for the determination of I based on its oxidation by potassium permanganate in sulphuric acid medium. The disadvantage of the chemiluminometric method is that it involves a very acidic medium and requires the reaction coil to be heated at 80 ◦ C posing experimental inconvenience. Very recently, Rocha et al. [6] described a flow injection spectrophotometric method of I using the same reaction with ferricyanide and 4-aminoantipyrine. Both the AOAC automated method and the one reported by Rocha et al. uses the ferricyanide and 4-aminoantipyrine reagents continuously which is uneconomical in terms of reagent consumption and waste generation. Sequential injection analysis (SIA) is known for its economical use of reagents since it is based on a discontinuous flow that takes up the reagents only when required. Moreover, it avoids multiple flow channels since it uses a single channel with a multi-position selection valve and a single pump that can be manipulated as stopped–reversed–forward mode. Thus, the present work

600

N.W. Beyene, J.F. Van Staden / Talanta 63 (2004) 599–604 Table 1 Device sequence and timing to change pump direction and valve position

exploits the advantages of SIA and deals with sequential injection spectrophotometric determination of I based on the ferricyanide–aminoantipyrine reaction described earlier.

Time (s)

Pump

Valve

0.00

Off

Starting position (position 1, reagent 1)

2. Experimental

1.00

2.1. Apparatus

2.72 4.00

Reverse-draws reagent 1 Off

Fig. 1 illustrates the SIA manifold used. Solutions were driven by a Gilson minipuls 3 peristaltic pump (Villiers-le-Bel, France) using a 1.29 mm i.d. pump tubing (Jobling, Staffordshire, UK). A 10-position micro-actuated selection valve from Valco Instruments (Houston, Texas, USA) was used to select the solutions. The holding coil was made by winding a 5 m, 0.89 mm i.d. Tygon tube on a glass rod. A single wavelength Unicam 8625 UV/Vis spectrophotometer (Unicam Ltd., Cambridge, UK) fitted with a 10 mm flow through cell that has a volume of 80 ␮l (Hellma, Mullheim, Baden, Germany) was employed. Data acquisition and device control was accomplished by using a PC30-B interface board (Eagle Electric, Cape Town, South Africa) in combination with an assembled distribution board (Mintek, Randburg, South Africa). The FlowTEK software package (version 1.3a) from Mintek was used throughout. Responses were evaluated using the relative peak heights in mV.

5.00 6.72 8.00

Reverse-draws sample Off Advance to position 3 (reagent 2)

9.00 10.72 12.00

Reverse-draws reagent 2 Off

13.00 58.00 59.00

Forward Off

Advance to position 4 (to detector)

Return to position 1

200-fold where as the other two were diluted 100-fold in 0.01 mol l−1 NaHCO3 . 2.3. Procedure Wavelength scan between 360 and 750 nm was run using an Agilent 8345 diode array spectrophotometer (Waldbronn, Germany) and maximum absorbance was obtained at 503 nm. The device sequence and corresponding timing are shown in Table 1. To compare results obtained by SIA, phenylephrine hydrochloride in the samples was determined by the reference method recommended by AOAC [3].

2.2. Reagents and samples Deionised water was produced using a Modulab apparatus (Continental Water System, San Antonio, Texas, USA). 4-Aminoantipyrine, (R)-(−)-Phenylephrine hydrochloride and potassium ferricyanide were obtained from Sigma–Aldrich. The carrier was 0.01 mol l−1 NaHCO3 (Sigma–Aldrich). Stock solutions of 4-aminoantipyrine (1%, w/v), potassium ferricyanide (5%, w/v) and phenylephrine hydrochloride (100 mg l−1 ) were prepared in 0.01 mol l−1 NaHCO3 . Samples of Prefrin® Liquifilm® eye drops, ENT Nasal drops, and Eye Gene® eye drops were obtained from a local pharmacy. Before analysis, the nasal drop was diluted

3. Results and discussion 3.1. Selection of carrier solution and aspiration order of reagents and sample The condensation reaction of I with 4-aminoantipyrine occurs at a slightly alkaline medium. To make the medium

1

10

Multiposition Selection valve 9

2

Holding coil

Advance to position 2 (sample)

3

8

4

7

Peristaltic pump 5

6

Unicam UV/VIS Reaction coil

Carrier

Computer Waste

Fig. 1. A schematic diagram of the SIA system used for the determination of phenylephrine hydrochloride.

N.W. Beyene, J.F. Van Staden / Talanta 63 (2004) 599–604 Table 2 Aspiration order of reagents and sample Aspiration order

Relative peak height (mean ± S.D. (R.S.D. %))

AP–sample–ferricyanide AP–ferricyanide–sample Ferricyanide–AP–sample Ferricyanide–sample–AP Sample–AP–ferricyanide Sample–ferricyanide–AP

6.9 1.6 2.3 7.1 2.6 3.3

± ± ± ± ± ±

4.7 3.9 6.0 8.5 1.4 4.5

× × × × × ×

10−2 10−2 10−2 10−2 10−1 10−2

(0.7) (2.4) (2.6) (1.2) (5.2) (1.4)

Experimental conditions were: reaction coil 0.89 mm i.d. and 90 cm in length, aspiration volume 100 ␮l in each case, concentration of standard 25 mg l−1 , flow rate 4.5 ml min−1 , concentration of AP 250 mg l−1 , and concentration of ferricyanide 1000 mg l−1 .

slightly alkaline, the AOAC method employs sodium borate solution where as the FIA method reported by Rocha et al. uses sodium bicarbonate solution [3,6]. A preliminary test that compared the use of borate and bicarbonate solutions as a carrier showed that using the latter gives a better signal. Alternatively, distilled water could be used as a carrier and the alkalinising reagent could be aspirated through one of the channels of the selection valve. However, this option was not tested for fear that it significantly extends the analysis time and in turn reduce the sample throughput. Thus, the 0.01 mol l−1 NaHCO3 solution employed by Rocha et al. was used as a carrier. No attempt was made to optimise the concentration of the carrier solution. The aspiration order of reagents and sample was investigated (Table 2) and best result was observed when the sample zone is sandwiched between the two reagent zones. This is in agreement with the observation of Gubeli et al. [7] on three zone penetrations. Interchanging the order of the two reagents (sandwiching the sample zone) did not show

601

significant difference in the response (a difference of 1.2%). The order aminoantipyrine–sample–potassium ferricyanide was chosen because of the low R.S.D.. 3.2. Optimisation of parameters 3.2.1. Flow rates The effect of the flow rate on the peak height was investigated from 1 to 5 ml min−1 at every 0.5 ml min−1 range (Fig. 2). The sample and reagent volumes aspirated were kept constant by changing the aspiration time in accordance with the flow rate. There were two maximum obtained, one at 1.5 ml min−1 and the other at 3.5 ml min−1 . At 1.5 ml min−1 , the slow flow rate allows better colour development since the reaction is time dependent [3]. But due to the high dispersion, it could not give the highest signal of all. Moreover, the sample throughput at this flow rate is very small. At lower flow rate (1 ml min−1 ), though the reaction time was higher favouring better colour development the effect of dispersion was dominant resulting in a lower peak height. As the flow rate increases, both the reaction time and dispersion are diminishing. Thus, a flow rate of 3.5 ml min−1 was taken for subsequent measurements. At this flow rate the sample throughput was found to be around 60 h−1 . 3.2.2. Reaction coil internal diameter The effect of reaction coil diameter on peak height was studied from 0.51 to 1.60 mm (all lengths were 90 cm) at five different diameters based on availability (Fig. 3). A coil diameter of 0.51 mm gave the highest signal and the shortest peak time in agreement with theory [8]. As the diameter increases the relative peak height decreases and the peak time increases.

8.5

relative peak height

8.0

7.5

7.0

6.5 0

1

2

3

4

5

6

flow rate/ mL min-1 Fig. 2. Effect of flow rate on the relative peak height. Experimental conditions were: reaction coil 0.89 mm i.d. and 90 cm in length, aspiration volume 100 ␮l in each case, concentration of standard 25 mg l−1 , concentration of AP 250 mg l−1 , and concentration of ferricyanide 1000 mg l−1 .

602

N.W. Beyene, J.F. Van Staden / Talanta 63 (2004) 599–604 9.0

relative peak height

8.0

7.0

6.0

5.0

4.0 0.25

0.50

0.75

1.00

1.25

1.50

1.75

reaction coil internal diameter/ mm Fig. 3. Effect of reaction coil diameter on the peak height. Experimental conditions were: flow rate 3.5 ml min−1 , aspiration volume 100 ␮l in each case, concentration of standard 25 mg l−1 , concentration of AP 250 mg l−1 , and concentration of ferricyanide 1000 mg l−1 .

the peak time shifts to a higher value (24 s at 30 cm and 32 s at 180 cm) compromising the sample throughput. Thus, to give adequate time for the reaction to proceed and to have relatively low dispersion and better sample throughput, a coil length of 90 cm was chosen for subsequent measurements.

3.2.3. Reaction coil length In order to have a better sensitivity and higher peak height, a reaction coil dimension that gives minimum dispersion is required. The effect of reaction coil length on peak height was assessed from 30 to 180 cm at every 30 cm range (Fig. 4). In agreement with theory, the highest signal was observed at a length of 30 cm. However, there was no significant difference in peak heights in the coil length range tested because the reaction is time dependent and the loss due to dispersion was overcompensated by the increase in colour intensity. Moreover, as the coil length increases

3.2.4. Concentration of potassium ferricyanide In SIA measurements, it is vital to have sufficient reagent excess to have a better sensitivity. The effect of concentration of potassium ferricyanide was investigated from 50 to

relative peak height

5.25

5.00

4.75

20

40

60

80

100

120

140

160

180

200

reaction coil length/ cm Fig. 4. Effect of reaction coil length on the peak height. Experimental conditions were: flow rate 3.5 ml min−1 , reaction coil diameter 0.51 mm, aspiration volume 100 ␮l in each case, concentration of standard 12.5 mg l−1 , concentration of AP 250 mg l−1 , and concentration of ferricyanide 1000 mg l−1 .

N.W. Beyene, J.F. Van Staden / Talanta 63 (2004) 599–604

603

7

relative peak height

6 5 4 potassium ferricyanide 4-aminantipyrine

3 2 1 0 0

250

500

750

1000

1250

1500

1750

2000

2250

concentration/ mg L-1 Fig. 5. Effect of concentration of reagents. Experimental conditions were: flow rate 3.5 ml min−1 , reaction coil 90 cm and 0.51 mm i.d., aspiration volume 100 ␮l in each case, concentration of standard 12.5 mg l−1 . When studying the concentration of ferricyanide, the concentration of 4-aminoantipyrine was 250 mg l−1 and for the study of the concentration of the latter the former was kept at 1500 mg l−1 .

2000 mg l−1 (Fig. 5). The peak height increases with increasing concentration. However, the increase above 1500 mg l−1 was not significant (a difference of 2.5%) and thus a concentration of 1500 mg l−1 was chosen for subsequent experiments. 3.2.5. Concentration of 4-aminantipyrine The effect of concentration of 4-aminoantipyrine was studied from 25 to 1000 mg l−1 (Fig. 5). The peak height increases with increasing concentration up till 500 mg l−1 and started levelling off. The corresponding increase in peak height when the concentration increases from 250 to 500 mg l−1 was not significant (an increase by 2.4%) and the R.S.D. at 250 is lower than that at 500 and hence 250 mg l−1 was chosen for further investigations. 3.2.6. Aspiration volumes One of the parameters that affect SIA measurements is the reagents and sample volume drawn up. In this work the effect of aspiration volumes of the two reagents and the sample were studied in the range 50–250 ␮l at every 50-␮l interval. When varying the volume of solution in question the other two were kept at 100 ␮l. First, a volume of 4-aminoantipyrine was considered. There was no significant difference in the response when the volume varied as abovementioned (Fig. 6). Again, keeping the others at 100 ␮l, that of ferricyanide varied as aforesaid and there was no significant difference in response (Fig. 6). When the volume of the sample varied from 50 to 250 ␮l, there was sharp increase up till 100 ␮l and then a slight increase till 200 ␮l followed by a decrease. The increase after 100 ␮l was not significant resulting unnecessarily in longer aspiration times that in turn

results in longer analysis time. Therefore, 100 ␮l was chosen as sample volume. 3.3. Figures of merit Using the aforementioned parameters the SIA system was evaluated for its response for different concentrations of phenylephrine hydrochloride. The linear calibration was found from 0.5 to 17.5 mg l−1 with an excellent correlation (relative peak height = 0.454C [mg l−1 ] + 0.289, r 2 = 0.999). The linear range obtained in this work is far better than the one reported by Rocha et. al. [6] as well as Mestre et. al. [5]. At concentrations of 2.5 and 5 mg l−1 , R.S.D.s of 0.8 and 0.6%, respectively, were registered (n = 10 measurements in each case). The detection limit (as 3σ value at a concentration of 0.5 mg l−1 for 10 determinations) [9] was found to be 0.09 mg l−1 . 3.4. Analysis of pharmaceutical samples Liquid eye and nasal drops were analysed by the SIA method as well as the AOAC official method [3]. In the case of the former, the samples were diluted in 0.01 mol l−1 NaHCO3 (see Section 2), where as in the latter case samples were diluted in water (dilution factors were the same in both cases). Table 3 presents the results obtained and the manufacturers’ claimed values. Rocha et. al. reported a difference between their method and AOAC method of 6% where as in our work it is 2%. A t-test was used to determine whether the results obtained by the two methods differ significantly. The calculated t-values in all cases were less than the tabulated value (2.45 for 6 degrees of freedom at 95% confidence level) showing that there is no

604

N.W. Beyene, J.F. Van Staden / Talanta 63 (2004) 599–604 7

relative peak height

6

aminoantipyrine potassium ferricyanide phenylephrine hydrochloride

5

4

3 50

100

150

200

250

300

aspiration volume/ L Fig. 6. Optimisation of aspiration volumes of sample and reagents. Experimental conditions were: flow rate 3.5 ml min−1 , reaction coil 90 cm and 0.51 mm i.d., aspiration volume 100 ␮l in each case, concentration of standard 12.5 mg l−1 , concentration of ferricyanide 1500 mg l−1 , and concentration of 4-aminoantipyrine 250 mg l−1 .

Table 3 Phenylephrine hydrochloride in pharmaceutical samples as determined by the SIA and AOAC methods Sample

SIA, mg l−1 (n = 5)

AOAC, mg l−1 (n = 3)

Difference (%)

Calculated t-value

Claimed value (mg l−1 )

Eye Gene® Nasal drop Prefrin®

446 ± 4 (2.59 ± 0.03) × 103 (1.34 ± 0.01) × 103

455 ± 22 (2.60 ± 0.09) × 103 (1.33 ± 0.04) × 103

2.0 0.7 0.8

0.90 0.41 0.50

430 2.5 × 103 1.2 × 103

significant difference between the values obtained by the two methods. 4. Conclusions The developed SIA method gave a better linear range, detection limit and repeatability as compared to previous reports. Moreover, the reagent consumption is significantly reduced (100 ␮l each) as compared to the FIA technique that requires 1350 ␮l of each reagent for a single run [6]. Results for the determination of phenylephrine hydrochloride in pharmaceutical preparations correlated excellently with those obtained by the AOAC official method.

References [1] J.G. Hardman, L.E. Limbird, A.G. Gilman, Goodman and Gilman’s The Pharmacological Basis of Therapeutics, 10th ed., McGraw-Hill, New York, 2001. [2] D.S. Tatro, A–Z Drug Facts, Facts and Comparisons, St. Louis, 1999. [3] Method 969.49, Official Methods of Analysis, 16th ed., AOAC, Gaithersburg, 1999. [4] Method 971.37, Official Methods of Analysis, 16th ed., AOAC, Gaithersburg, 1999. [5] Y.F. Mestre, L.L. Zamora, J.M. Calatayud, J. AOAC Int. 84 (2001) 13. [6] J.R.C. Rocha, C.X. Galhardo, M.A.E. Natividade, J.C. Masini, J. AOAC Int. 85 (2002) 875. [7] T. Gubeli, G.D. Christian, J. Ruzicka, Anal. Chem. 63 (1991) 2407. [8] J.F. van Staden, A. Botha, S. Afr. J. Chem. 51 (1998) 100. [9] J.C. Miller, J.N. Miller, Statistics for Analytical Chemistry, second ed., John Wiley & Sons, New York, 1988.