A Simple Method for the Ampicillin Determination in Pharmaceuticals and Human Urine

Abstract

The new kinetically-based spectrophotometric method for the determination of microquantities of ampicillin is proposed in the present paper. Ampicillin degradation in strong alkaline medium was applied for the method development. The reaction rate was monitored at 265 nm. A differential variation of the tangent method was used to process the kinetic data. The method is valid over the 3.49–55.84 µg/mL ampicillin concentration interval with relative standard deviation (RSD) range 7.79–3.20%. The calculated detection limit was determined at 2.58 µg/mL based on the 3.3S₀ criterion. The interference effects of some metal ions, anions, amino acids and other molecules were investigated in order to assess the method selectivity. The method was successfully applied to determining the content of ampicillin in commercial pharmaceutical preparations and human urine. The obtained results were in good correlation with the HPLC method results. The newly developed method is simple, inexpensive and efficient for the analysis of a large number of samples at room temperature in a short time.

Ampicillin (α-aminobenzylpenicillin; Amp) is a semisynthetic β-lactam antibiotic widely used against Gram-negative and Gram-positive organisms, as well as for enterococcal infections. It has been in the commercial usage from 1961 in the forms of pharmaceutical preparations such as capsules, pills, oral suspension powder, injection and intravenous infusion. Amp is well absorbed from the gastrointestinal tract producing good serum and urine concentrations. The measurement of Amp levels in human urine was particularly valuable in newborn and children who showed larger excretion half-life values than adults and in patients with renal failure. About 20% is connected to plasma proteins and approximately 30% is actively secreted into the renal tubules and excreted mostly unchanged into the urine during the first six hours after an oral administration, while after parenteral administration about 80% of Amp is excreted unchanged into the urine.¹⁾ Half-life, maximum excretion rate and maximum time of maximum excretion rate of Amp in adult human males during an 8-h period after an oral administration of 500 mg Amp capsule were determined at 1.25 h⁻¹, 44.2±7.9 mg/h and 3.5 h, respectively. Amp concentration in urine over the 10–100 µg/mL range was determined, too.²⁾

Amp is an amphoteric compound and it behaves essentially as an aliphatic amino acid. Being amphoteric, Amp in solution exists mainly in three different forms: cation (XH₂⁺) in acid, zwitterions (HX^±) in neutral and anion (X⁻) in alkaline medium (Chart 1). Amp has two dissociation constants: K₁ (pK₁=2.55) corresponds to the dissociation of cation and K₂ (pK₂=7.14) that corresponds to the dissociation of zwitterion. At pH >7.5 the data indicated that only one type of reaction was responsible for the reaction rate, which is the hydroxyl ion interacting with the anionic form of ampicillin. In the basic solution the hydrolytic rate of ampicillin is almost 1400 times faster than that in the acid solution (average values of k_H⁺ and k_OH⁻ are 1.38 L/mol h and 1945 L/mol h, respectively). Both α-aminobenzylpenicilloic acid (a) and α-aminobenzylpenilloic acid (b) are formed in a basic solution. In the strong basic solution only α-aminobenzylpenilloic acid, a decarboxylated product is formed (Chart 2). UV absorption spectra of ampicillin reaction mixture at different pH values are presented in Fig. 1. Absorption maxima at 257, 262 and 268 nm are decreasing with increasing of pH.³⁾

Chart 1. Dissociation of Amp Zwitterion

Chart 2. Ampicillin Degradation in Basic Solution

Fig. 1. UV Absorption Spectra of Amp Reaction Mixtures at Different pH and t=35°C in: 0.5–1.0 M HCl Solution after Heating for About 10 min

(λ_max 280 nm) (—), 0.1–0.5 M HCl solution after heating for about 30 min (λ_max 320 nm) (- - -), freshly prepared pH 5 buffer solution (• • •), freshly prepared pH 10 buffer solution (-・-・-).³⁾

The anionic form of Amp reacts with Ni²⁺ and produces stable complexes [NiAmp]⁺ and [Ni(Amp)₂]⁰ at pH range 4.5–7.5. The appropriate stability constants (log K_ML) values are 4.25 and 7.54, respectively. The system Ni(II)‒Amp is examined spectrophotometrically (Fig. 2). The spectrum of the [Ni(H₂O)₆]²⁺ ion contains three absorption maxima (390, 660 and 720 nm). The peak at 390 nm partially overlaps with the Amp band. The two absorption maxima of [Ni(Amp)₂]⁰ appear at 630 and 940 nm.⁴⁾ Ions of Ni(II) in water solution accept electronic pairs from H₂O molecules forming green colored aqua complex: Ni²⁺+6H₂O⇔[Ni(H₂O)₆]²⁺. Aqua complexes are unstable, because hydrated Ni(II) ion releases protons forming hydroxo complexes. In alkaline solution, at pH=8.0, nickel ions precipitate in the form of Ni(OH)_2(s). In addition, Ni(II) ion in the excess of alkali hydroxides produces water soluble hydroxo complexes [Ni(OH)]⁺, [Ni(OH)₂]_(aq), [Ni(OH)₃]⁻ and [Ni(OH)₄]²⁻, based on the increasing solubility of Ni(OH)_2(s) in alkaline solutions (pH >9).⁵⁾

Fig. 2. Absorption Spectra at pH=7.0: Ni(II) Aqua Ion (1), NaAmp Solution (2) and [Ni(Amp)₂] (3)⁴⁾

Literature review brings several analytical methods for the determination of ampicillin in pure form or in pharmaceutical formulations as well as in biological samples: spectrophotometry,^6–10) colorimetry,¹¹⁾ kinetic–spectrophotometry,¹²⁾ liquid chromatography,¹³⁾ HPLC,^14–17) spectrofluorimetry,^18,19) Flow Injection Analysis,^20–22) polarography^23,24) and NMR.²⁵⁾ British²⁶⁾ and Yugoslav Pharmacopoeia²⁷⁾ describe liquid chromatography method for Amp determination.

The aim of the present research was to develop a simple, fast, reliable, precise and accurate kinetic method for the quantification of ampicillin in pharmaceutical preparations and human urine. The method was based on the hydrolytic reaction of Amp in strong alkaline solution at room temperature, in the presence of Ni(II) ions as catalysts.

Experimental

Apparatus

Deionized water was obtained using MicroMed high purity water system, TKA Wasseraufbereitungssysteme GmbH. The solutions were refrigerated, but thermo-stated at 25±0.1°C prior to use in Julabo MP 5A Open Bath Circulations. Measurements of pH were carried out using Hanna Instruments pH meter. High precision variable volume micropipettes (Lab Mate⁺) of 50, 500 and 1000 µL were used for pipetting the solutions. The readings of the absorbance were recorded by Agilent 8453 UV/Vis spectrophotometer, with 1 cm optical length cell. The model 1200 Agilent technologies was used for performing HPLC analysis.

Reagents

All solutions were prepared with deionized water. All chemicals were of analytical grade provided by Merck, unless indicated otherwise. The stock solution of Amp (1.00×10⁻³ mol/L) was prepared by dissolving the required, accurately weighed, amount of ampicillin-trihydrate (Dolber, Switzerland, Ph.). The stock solution of Ni(II) (1.00×10⁻¹ mol/L) was prepared by dissolving the required, accurately weighed, amount of nickel-chloride. The stock solutions of KCl (1.5 mol/L) and NaOH (1 mol/L) were obtained by dissolving accurately measured amount of salt and base. All the glassware used was cleaned in the aqueous solution of HCl (1 : 1) and then thoroughly rinsed with tap, distilled and finally with deionized water.

Procedure. General Procedure

The selected volumes of reagents were transferred into the four-compartment vessel in the following order: nickel-chloride, ampicillin, sodium-hydroxide, potassium-chloride and water to the predetermined volume of 10 mL that was held constant during the experiment. After thermo-stating the vessel at 25°C for 5 min, the content was well mixed and transferred into the cuvette. Absorbance (A) was measured at 265 nm spectrophotometrically every 30 s during 5 min, starting 60 s after the reactants were mixed. A differential variant of the tangent method was used for the processing of the kinetic data. The rate of the reaction (ν=tg α) was obtained by measuring the slopes (dA/dt) of the linear part of the kinetic curves (absorbance–time plots). The calibration curve was constructed by plotting the rate of reaction (tg α) vs. concentration (c) of Amp. The amount of the drug was calculated on the basis of experimentally obtained values of reaction rate and its substitution into the linear regression equation (Eq. 2).

Procedure for Capsules

The content of eight capsules was emptied and mixed well. An amount of resulting powder, equivalent to the mass of one capsule (500 mg of ampicillin), was transferred into the 500 mL volumetric flask and dissolved with deionized water. The obtained suspension was filtrated (Whatman No. 42) and the appropriate aliquots of filtrate were directly used for measuring the quantity of ampicillin covering the concentration range listed in Table 1.

Table 1. Accuracy and Precision of Amp Determination in Pharmaceutical Preparations

Concentration of Amp (µg/mL)			RSD (%)	(%)	Recovery (kinetic method) (%)	F-Teste)	t-Teste)
Taken µ	Determined by kinetic method x̄±S.D.	Determined by HPLC method x̄±S.D.
20.94a)	22.09±1.10	20.83±0.59	4.98	+5.49	105.49	3.47	2.26
20.94b)	21.42±1.14	20.75±0.65	5.32	+2.29	102.29	3.08	1.14
20.94c)	21.37±0.63	21.14±0.37	2.95	+2.05	102.05	2.90	1.05
10.47d)	9.66±0.58	10.13±0.25	6.00	−7.74	92.26	5.38	1.65

a) Pentrexyl®, capsules, 500 mg ampicillin, Galenika, a. d., Beograd, Srbija. b) Ampicilin, capsules, 500 mg ampicillin, Panfarma, d. o. o., Beograd, Srbija. c) Pamecil®, powder for injection/infusion, 1 g ampicillin, Medochemie Ltd.; Factory B, Limasol, Cyprus. d) Pentrexyl®, powder for oral suspension, 250 mg ampicillin/5 mL, Galenika, a. d., Beograd, Srbija. e) Theoretical values for F-test (ν₁=4, ν₂=4) and t-test (ν=8) for a confidence level of 95% are 6.39 and 2.306, respectively.

Procedure for Children Syrup

The content of one bottle of oral suspension powder was dissolved with deionized water according to the manufacturer instructions. Ampicillin solution (50 µg/mL) was prepared with further diluting and filtrating (Whatman No. 42) of suspension. Selected volumes of filtrate were directly used for the determination of ampicillin (Table 1).

Procedure for Injection/Infusion Powder

The appropriate amount of previously dissolved powder (0.0372 g/100 mL) was directly used instead of the stock Amp solution. The results of Amp quantification in this powder are presented in Table 1.

Urine Sample Preparation

Human urine, after an overnight fast, without any previous sample pretreatment, was used for the analysis. Morning urine is the most concentrated urine regarding the amount of biomolecules, electrolytes, oligoelements, heavy metals, etc. (the overnight diuresis is about two times smaller than the daily diuresis). The standard addition method was applied for the drug determination. Amp was added to match its normal excretion level in human urine,^1,2) covering the concentration range listed in Table 2.

Table 2. Accuracy and Precision of Amp Determination in Human Urine by Standard Addition Method

Concentration of Amp (µg/mL)			RSD (%)	(%)	Recovery (kinetic method) (%)	F-Test	t-Test
Added µ	Determined by kinetic method x̄±S.D.	Determined by HPLC method x̄±S.D.
10.47	9.02±0.85	9.97±0.43	9.42	−13.85	86.15	3.91	2.24
27.92	26.12±1.23	27.45±0.71	4.71	−6.44	93.56	3.00	2.10

Comparative Method

The procedure for comparative method has been described in the Yugoslav Pharmacopoeia.²⁷⁾ The results are given in Tables 1 and 2.

Results and Discussion

Absorption Spectra

The proposed method was based on the Amp degradation in strong alkaline medium. The degradation rate was apparently very slow at room temperature. The addition of Ni(II) ions to the reaction mixture resulted in the increase of the reaction rate which was measured as an absorbance change with time. In order to regulate the ionic strength of the solution, KCl was added. The changes of the absorption spectrum of this reaction mixture with time are presented in Fig. 3. A very small maximum at λ_max=392 nm appeared on these spectra and it did not change with time. The absorption band whose intensity increased with time occurred in the 250–295 nm wavelength interval. The degradation rate of Amp was measured spectrophotometrically at 265 nm because the biggest changes of absorbance with time appeared at this wavelength, which gave the opportunity for the development of more sensitive kinetic method.

Fig. 3. The Absorption Spectrum of the Amp Degradation after: 1 min (a); 5.8 min (b) and 11.8 min (c)

Initial reactants concentrations in solution: c(Ni²⁺)=1×10⁻⁴ mol/L; c(Amp)=34.9 µg/mL; c(NaOH)=1×10⁻² mol/L; c(KCl)=1.5×10^–2 mol/L; t=25.0±0.1°C.

Kinetic Studies

Preliminary experiments were performed in order to optimize conditions for the determination of the lowest possible Amp concentration. The impact of concentration of each reactant on the reaction rate was studied. The chosen concentration values of the variables were maintained constant throughout the experiment.

The effect of the NaOH concentration change was monitored in the interval of (0.1–5.0)×10⁻² mol/L (Fig. 4). On the basis of the obtained plot, the reaction rate increased from (0.1–3.5)×10⁻² mol/L NaOH. With further increase of the NaOH concentration in the range of (3.5–5.0)×10⁻² mol/L, the reaction rate became constant, because the hydrolytic degradation of Amp was completed under the given conditions. The value of 4.0×10⁻² mol/L was chosen as optimal concentration of the reagent NaOH. The reaction was of the zero order with respect to the NaOH concentration in the interval of (3.5–5.0)×10⁻² mol/L.

Fig. 4. Reaction Rate Dependence on the NaOH Concentration

Initial reactants concentrations in solution: c(Ni²⁺)=1×10⁻⁴ mol/L; c(Amp)=34.90 µg/mL; c(KCl)=1.5×10⁻² mol/L; t=25.0±0.1°C.

The results of the reaction rate tested in relation to the KCl concentration are presented in Fig. 5. It can be noticed that the reaction rate significantly decreased with the increase of this electrolyte concentration in the whole examined interval of (0.15–15)×10⁻² mol/L. The increment of this compound increased the ionic strength of the solution, but decreased the ionic activity (especially of the Ni(II) ions) that led to the decrease of the reaction rate. The reaction was −1 order with respect to the KCl concentration in the whole examined interval. The value of 1.5×10⁻² mol/L was selected as optimal.

Fig. 5. Reaction Rate Dependence on the KCl Concentration

Initial reactants concentrations in solution: c(Ni²⁺)=1×10⁻⁴ mol/L; c(Amp)=34.90 µg/mL; c(NaOH)=4×10⁻² mol/L; t=25.0±0.1°C.

The correlation between the reaction rate and Ni(II) concentration in the interval of (0.1–4.0)×10⁻⁴ mol/L is shown in Fig. 6. The rate increased with the rise of the Ni(II) ions concentration to the value of 2.5×10⁻⁴ mol/L, then the rate remained constant until the end of the observed concentration interval. The reaction was of the zero order with respect to this reactant in the range of (2.5–4.0)×10⁻⁴ mol/L. The selected value for the further measurements was 3.0×10⁻⁴ mol/L.

Fig. 6. Dependence of the Reaction Rate on the Ni(II) Ion Concentration

Initial reactants concentrations in solution: c(KCl)=1.5×10⁻² mol/L; c(Amp)=34.90 µg/mL; c(NaOH)=4.0×10⁻² mol/L; t=25.0±0.1°C.

The following kinetic equation for the reaction rate was deduced on the basis of the previously shown dependences:

  

(1)

where k is a relative reaction rate constant. This equation is valid for these concentration intervals: NaOH=(3.5–5.0)×10⁻² mol/L, KCl=(0.15–3.0)×10⁻² mol/L and Ni(II)=(2.5–4.0)×10⁻⁴ mol/L. But for the constant KCl concentration, the following equation is used:

  

where k′ is a relative reaction rate constant, but with the constant value of KCl concentration, c=1.5×10⁻² mol/L.

Validation of the Proposed Method

Linearity

For the evaluation of linearity, determining Amp was done five times at eleven concentration levels under previously chosen optimal conditions: c(Ni²⁺)=3.0×10⁻⁴ mol/L; c(NaOH)=4.0×10⁻² mol/L; c(KCl)=1.5×10⁻² mol/L; t=25.0±0.1°C. The least squares’ equation (y=bx+a, where b and a are its slope and intercept, respectively) for the calibration graph and correlation coefficient r²⁸⁾ for the determination of Amp in the 3.49–55.84 µg/mL interval were calculated:

  

(2)

where c(Amp) is the ampicillin concentration expressed in µg/mL.

The Limit of Detection (LOD) and the Limit of Quantification (LOQ)

The minimum concentration of Amp which can be determined by this method was calculated at 2.83 µg/mL using the formula LOD=3.3×S₀/b, where S₀ is the residual standard deviation, while b is the slope of the calibration curve. The limit of quantification was obtained at 8.58 µg/mL using equation LOQ=10×S₀/b.^29,30)

Accuracy and Precision

Accuracy and precision were determined for four concentrations from the calibration graph based on five measurements results for every chosen value (Table 3). The first two concentration values that were smaller than LOQ were determined with the highest RSD. Values for relative standard deviations were in the 7.79–3.20% range for the 3.49–55.84 µg/mL concentration interval.

Table 3. Accuracy and Precision of Amp Determination

Amp concentration (µg/mL)		RSDb) (%)	c) (%)
Taken µ	Determined x̄±S.D.a)
3.49	3.08±0.24	7.79	−11.75
6.98	7.20±0.45	6.25	−3.15
27.92	28.42±0.95	3.34	+1.79
55.84	55.03±1.76	3.20	−1.45

a) Mean and standard deviation of five measurements at a confidence level of 95%, b) Relative standard deviation, c) (x̄−μ)/μ, accuracy of the method.

Interferences Studies

The effect of a wide range of common ions and organic molecules on the reaction rate was examined in order to assess the selectivity of the method. The applied criterion for their selection was partly based on compounds present in urine and the ingredients of pharmaceuticals soluble in water. The influence of foreign species, that referred only to urine (CH₃COO⁻, NO₃⁻ as urine preservatives; urea as one of the urine most concentrated component; cations such as Pb²⁺, Sn²⁺, Cd²⁺, Sr²⁺ and Mn²⁺ as toxic metals, SO₄²⁻, SO₃²⁻, C₂O₄²⁻ and HPO₄²⁻ as anions of the possible calculuses) was examined, too. The tolerance limits for the species studied on the determination of 8×10⁻⁵ mol/L (27.92 µg/mL) Amp are presented in Table 4. The maximum level tolerated was taken as that causing an error in the concentration less than twice the standard deviation (2 s criterion) of the concentration for the reaction system without foreign species.³¹⁾ Li⁺, Ca²⁺ and F⁻ when present in 1000-fold excess, CH₃COO⁻, NO₃⁻ and I⁻ ions in 100-fold excess, urea and amino acids (Ala, Met Lys, Arg and Phe) in 10-fold excess, as well as mannitol, glucose, citric acid, nicotinic acid, Asn, Mg²⁺, Sn²⁺, Pb²⁺, Cd²⁺, Sr²⁺, SO₄²⁻ and SO₃²⁻ in 1 : 1 M ratio did not interfere with the method. Fructose, oxalate and Ser interfered with the method when present in the same concentration as Amp. More severe interferences were observed for sorbitol, lactose, Al³⁺ and HPO₄²⁻ (they acted as inhibitors), as well as for thiamine and His (they acted as catalysts). The very powerful interference was noticed for ascorbic acid, citrate, Zn²⁺, Mn²⁺, Cu²⁺, because they changed absorption spectrum of the reaction system.

Table 4. Tolerance Level of Foreign Ions and Molecules on Determining Amp Concentration of 8×10⁻⁵ mol/L (27.92 µg/mL)

Tolerance level (cx/cAmp)	Ion/molecule (x)
103	Li+, Ca2+, F−
102	CH3COO−, NO3−, I−
10	Ala, Met, Lys, Arg, Phe, urea
1	mannitol, glucose,citric acid, nicotinic acid, Asn, Mg2+, Sn2+, Pb2+, Cd2+, Sr2+, SO42−, SO32−
Interfere at 1 : 1	fructose, Ser, C2O42−, sorbitol, lactose Al3+, HPO42−, thiamine, His, ascorbic acid, citrate, Zn2+, Cu2+, Mn2+

The Temperature Effect

In order to evaluate activation parameters of this system, the reaction rate was monitored at four temperatures (19, 22, 25, 28°C) under the chosen optimal conditions. Relative reaction rate constants (Table 5) for these temperatures were calculated using the kinetic equation (Eq. 1). Activation energy (21.05 kJ/mol) was determined using the slope of the Arrhenius curve (Fig. 7) and the following equation:

  

(3)

The other thermodynamic parameters (ΔH*=18.57 kJ/mol, ΔS*=−197.94 J/Kmol and ΔG*=77.56 kJ/mol) were calculated using the following equations:

  

(4)

  

(5)

  

(6)

Fig. 7. Dependence of log k on 1/T

Very low values of Ea* and ΔH* resulted from the reactivity of Amp molecule in the presence of Ni(II) as a catalyst. The value of ΔH* (18.57 kJ/mol) was about five times smaller than the obtained ΔH* (93.36 kJ/mol) by Hou and Poole.³⁾ This proves the catalytic role of Ni(II) ions in the reaction of Amp degradation. As a result of dissociation, the system becomes arranged and for endothermic reactions the value of ΔS* decreases, and ΔG*>0. The change of free energy (ΔG) as an important indicator of spontaneity provides no information about reaction rates. If ΔG has positive, but small numerical value, that does not mean the reaction will not occur, but that the value of reaction rate constant k is smaller than 1. The value of ΔS*<0 means that reaction is slow and also that ions formed in dissociation show strong impact on the orientation of the H₂O dipoles, thereby the system becomes organized and its surrounding (H₂O molecules) less disorganized.^32,33)

Applicability of the Method

The method was applied to the evaluation of the ampicillin quantity in drug formulations, commercial products randomly collected from local pharmacies, as well as in human urine. The results of the proposed method were statistically compared to results obtained by HPLC method²⁷⁾ using a point hypothesis test (Tables 1, 2). The values of F-test and Student’s t-test at 95% confidence level did not exceed the tabulated F- and t-values, confirming no significant difference between the performance of the proposed and HPLC method. This indicated that the constituents of human urine and pharmaceuticals did not interfere with the determination of Amp, because the tolerated amounts of foreign species were much higher than those usually present in analyzed samples.

Table 5. Relative Reaction Rate Constant at Different Temperatures

k (mol·dm−3)1−n s−1	T (K)
0.13	292
0.14	295
0.16	298
0.17	301

n, reaction order

Conclusion

As shown in this study, ampicillin degradation in strong basic solution, in the presence of Ni(II) as a catalyst, was used for the development of an effective method for the kinetic–spectrophotometric determination of this antibiotic. Ions Ni(II) did not form either complexes with this drug (no absorption maxima detected, Fig. 3 related to Fig. 2), or precipitated with OH⁻ ions (all kinetic dependences, A=f(t), were linear), because Ni(II) ions formed stable, water soluble hydroxo complexes. The hypothesis was that in the collision of β-lactame ring of Amp-anion and [Ni(OH)₄]²⁻ (the highest catalytic activity due to the greatest dissociation degree) an activated complex would be formed. Ni(II) complex donated OH⁻ ion and became a more stable form of [Ni(OH)₃]⁻. After receiving OH⁻ ion, Amp degradation was conducted according to the Chart 2.

The literature review has shown that there is only one kinetic–spectrophotometric method¹²⁾ for the determination of ampicillin in pharmaceuticals, while there is no kinetic method for its determination in aqueous biological solution (urine). That method is valid for the 8–40 µg/mL concentration interval, with the detection limit of 0.73 µg/mL, which is far below the smallest concentration of the calibration graph. The proposed method is less time consuming than the method to which it was compared.

The properties of the method presented in this paper are: 3.49–55.84 µg/mL linearity range, 7.79–3.20% RSD range, LOD=2.83 µg/mL and LOQ=8.58 µg/mL. The method was easily and successfully applied for the quantification of Amp in the pharmaceutical formulations and human urine at room temperature.

Acknowledgment

This research was supported by the Grant no. 142015 from the Serbian Ministry of Science and Technological Development. The authors are grateful for the financial support provided by the Ministry.

References

• 1)  Moffat A. C., “Clarke’s Isolation and Identification of Drugs,” 2nd ed., The Pharmaceutical Press, London, 1986.
• 2)  El-Sebai A. I., Abdel-hamid M. E., Abuirjele M. A., Hurani A. M., Anal. Lett., 21, 423–434 (1988).
• 3)  Hou J. H., Poole J. W., J. Pharm. Sci., 58, 447–454 (1969).
• 4)  Alekseev V. G., Shcherbakova E. E., Yakubovich Yu. Ya., Vorob’ev N. N., Larin S. V., Shigina O. Yu., Russ. J. Gen. Chem., 76, 321–324 (2006).
• 5)  Mompean F. J., Illemassene M., Perrone J., “Chemical Thermodynamics, Vol. 6, Chemical Thermodynamics of Nickel,” 1st ed., Elsevier Science, Burlington, 2005.
• 6)  Devani M. B., Patel I. T., Patel T. M., Talanta, 39, 1391–1394 (1992).
• 7)  Smith J. W. G., De Grey G. E., Patel V. J., Analyst, 92, 247–252 (1967).
• 8)  Sastry C. S. P., Rao S. G., Y Naidu P., Srinivas K. R., Talanta, 45, 1227–1234 (1998).
• 9)  Mahgoub H., Aly F. A., J. Pharm. Biomed. Anal., 17, 1273–1278 (1998).
• 10)  Amin A. S., Moustafa M. E., Issa Y. M., Microchem. J., 50, 6–13 (1994).
• 11)  El-Shafie F. S., Gad-Kariem E. A., Al-Rashood K. A., Al-Khamees H. A., El-Obeid H. A., Anal. Lett., 29, 381–393 (1996).
• 12)  Belal F., El-Kerdawy M. M., El-Ashry S. M., El-Wasseef D. R., Il Farmaco, 55, 680–686 (2000).
• 13)  Nelis H. J., Vandenbranden J., Verhaeghe B., De Kruif A., Mattheeuws D., De Leenheer A. P., Antimicrob. Agents Chemother., 36, 1606–1610 (1992).
• 14)  Akhtar M. J., Khan S., Khan M. A., J. Pharm. Biomed. Anal., 11, 375–378 (1993).
• 15)  Ishida M., Kobayashi K., Awata N., Sakamoto F., J. Chromatogr. B Biomed. Sci., 727, 245–248 (1999).
• 16)  Tsou T.-L., Huang Y.-C., Lee C.-W., Lee A.-R., Wang H.-J., Chen S.-H., J. Sep. Sci., 30, 2407–2413 (2007).
• 17)  Luo W., Ang C. Y. W., Thompson H. C. Jr., J. Chromatogr. B Biomed. Sci., 694, 401–407 (1997).
• 18)  Gala B., Gómez-Hens A., Pérez-Bendito D., Talanta, 44, 1883–1889 (1997).
• 19)  Fernandez-Gonzalez A., Badia R., Diaz-Garcia M. E., Anal. Chim. Acta, 484, 223–231 (2003).
• 20)  García M. S., Sánchez-Pedreño C., Albero M. I., Ródenas V., J. Pharm. Biomed. Anal., 12, 1585–1589 (1994).
• 21)  Al-Momani I. F., Anal. Lett., 37, 2099–2110 (2004).
• 22)  Fernández-González A., Badía R., Díaz-García M. E., J. Pharm. Biomed. Anal., 29, 669–679 (2002).
• 23)  Sayed G. El., Amin A. S., Issa Y. M., Anal. Lett., 27, 2515–2521 (1994).
• 24)  Belal F., Rizk M. S., Eid M., J. Pharm. Biomed. Anal., 17, 275–282 (1998).
• 25)  Shin M.-H., Park M.-K., Yu C.-H., Choi J.-K., Arch. Pharm. Res., 4, 9–17 (1981).
• 26)  “British Pharmacopoeia 98/34/EEC,” The Stationery Office, London, 2005.
• 27)  “Jugoslovenska farmakopeja 2000,” 5. Izdanje, knjiga 1, Savezni zavod za zaštitu i unapređenje zdravlja i Savremena administracija, Beograd, 2001. “The Yugoslav Pharmacopoeia 2000 (in Serbian),” 5th ed., book 1, Republic Institute for Health Protection and Development and Modern Administration, Belgrade, 2001.
• 28)  Miller J. N., Analyst, 116, 3–14 (1991).
• 29)  Perez-Bendito D., Silva M., “Kinetic Methods in Analytical Chemistry,” Ellis Horwood, Chichester, 1988.
• 30)  Ermer J., J. Pharm. Biomed. Anal., 24, 755–767 (2001).
• 31)  Kirkbright G. F., Talanta, 13, 1–14 (1966).
• 32)  McKee T., McKee J. R., “Biochemistry,” Wm. C. Brown Publishers, Chicago, 1996.
• 33)  Petranovic N., “Hemijska termodinamika,” Fakultet za fizicku hemiju, Beograd, 1992. (Petranovic N., “Chemical Thermodynamics (in Serbian),” Faculty of Physical Chemistry, Belgrade, 1992.)