No job name

Ind. Eng. Chem. Res. 2006, 45, 6368-6374
Solubility of Antibiotics in Different Solvents. 1. Hydrochloride Forms of
Tetracycline, Moxifloxacin, and Ciprofloxacin
Fa´tima Varanda,† Maria J. Pratas de Melo,† Ana I. Cac¸o,† Ralf Dohrn,‡ Foteini A. Makrydaki,‡,§
Epaminondas Voutsas,§ Dimitrios Tassios,§ and Isabel M. Marrucho*,†
Department of Chemistry, UniVersity of AVeiro, CICECO, 3810-293.15 AVeiro, Portugal, ThermophysicalProperties, Bayer Technology SerVices GmbH, B310, 51368 LeVerkusen, Germany, and Department ofChemical Engineering, National Technical UniVersity of Athens, TTPL, Zografou Campus, GR-15780,Athens, Greece The solubilities of tetracycline hydrochloride, moxifloxacin hydrochloride, and ciprofloxacin hydrochloridewere measured in several solvents, such as water, ethanol, 2-propanol, and acetone, in the temperature rangeof 293.15-323.15 K for ciprofloxacin.HCl and moxifloxacin.HCl and 288.15-310.15 K for tetracycline. Allthe antibiotics have the same solubility order; that is, they are more soluble in water than in ethanol, andmore soluble in ethanol than in 2-propanol and acetone. The solubility in water is ∼3 orders of magnitudehigher than that in acetone. The modeling of the experimental solid-liquid equilibria (SLE) data, using theNRTL and UNIQUAC models, proves that these models can correlate the solubility of studied antibioticssatisfactorily in the temperature range for which experimental data are available, with the UNIQUAC modelgenerally being slightly superior to the NRTL model, when only two adjustable parameters are used for eachbinary system.
1. Introduction
the resistance.4 Second-generation quinolones, the so-calledfluoroquinolones, have a broader spectrum of activity and, so The production of pharmaceutical and medium-sized bio- far, do not present any resistance problems. Apart from chemicals customarily involves liquid solvents for reaction, tetracycline hydrochloride, for which some solubility data exists separation, and formulation. Most synthetic pharmaceuticals are on the Internet, two other antibiotics were studied: medium-sized molecules that contain 10-50 non-hydrogen hydrochloride forms of ciprofloxacin (Cipro, Bayer) and moxi- atoms. The molecules are typically composed of several floxacin (Avelox, Avalox, Bayer). The structures of the studied interlinked aromatic cores and multiple substituents that contain antibiotics are presented in Figure 1. Quinolones have two types heteroatoms such as N, P, O, S, and F or Cl. Because of the of ring structures: a naphthyridine nucleus with N atoms at presence of the aromatic delocalized π-electrons and the positions 1 and 8, and a nucleus with only one N atom in electronegative heteroatoms, the molecules are highly polariz- position 1, which is referred to as the quinoline nucleus. They able and, thus, are liable to a variety of specific interactionswith polar solvents, e.g., protonation, hydrogen bonding, specific also contain the keto oxygen at the C4 position and a carboxylic solvation, etc. Furthermore, these complex molecular systems side chain at the C3 position. Moreover, moxifloxacin and may exhibit charge-transfer complexes. All of them are con- ciprofloxacin have a piperazinyl group at the C7 position. The formally flexible, which may affect their reactivity and solva- presence of both the carboxyl and the amine groups makes the tion.1 The procedure of solvent selection is a thermodynamic acid-base behavior of these drugs be influenced by the problem that is solved based on the phase equilibrium theory, physicochemical properties of the solvent.5 The reported pKa experience, and empirical descriptions of experimental results.
values for ciprofloxacin are 6.09 and 8.62 for the protonated Experience shows that >30% of the efforts of industrial property amino group,6 which are within the intervals found for other modelers and experimentalists involve solvent selection.2 piperazinyl fluoroquinolones (5.7-6.3 and 7.6-8.3).7 On the Since the discovery of the antibiotic action of tetracycline in other hand, tetracycline must be considered as a three-protic 1947, a large number of derivates has been synthesized and acid. In water, the first deprotonation step occurs at pka1 ≈ 3.3, successfully applied. Recently, the use of many of them has the second occurs at pka2 ≈ 7.7, and the third one is observed been reduced, because of the fact that numerous bacteria developed efficient resistance mechanisms.3 The microbial Although there is a substantial amount of characterization resistance to anti-infective agents is a growing problem that work, from the molecular point of view, for other piperazinyl challenges researchers in regard to the development of new fluoroquinolones,8 only a limited amount of solubility data is synthetic antimicrobial agents. Quinolones are an example of a available8 and no data at all were observed for ciprofloxacin class of antimicrobial agents that were introduced in the mid and moxifloxacin. The solubility in body fluids has a strong 1980s to overcome this problem. The first agents in this class influence on the bioavailability of the active ingredient. The are active against many gram-positive and gram-negative addition of the HCl group is related to the enhancement of their aerobes; however, their use is again limited by development of solubility in aqueous systems, such as body fluids, because ofthe presence of charges. Therefore, many antibiotics are used * To whom correspondence should be addressed. E-mail address: in the hydrochloride form, e.g., ciprofloxacin hydrochloride and moxifloxacin hydrochloride are used for oral administration in † Department of Chemistry, University of Aveiro.
‡ Thermophysical Properties, Bayer Technology Services GmbH.
§ Department of Chemical Engineering, National Technical Univer- Solubility denotes the solute concentration in a solution that is in thermodynamic equilibrium with the solute in the solid Ind. Eng. Chem. Res., Vol. 45, No. 18, 2006 Figure 1. Structures of antibiotics molecules: (a) tetracycline.HCl, (b) ciprofloxacin.HCl, and (c) moxifloxacin.HCl.
state. At phase equilibrium, for any species i, the fugacities f Table 1. Wavelength of the Maximum Absorbance of the Antibiotic
must be the same in all phases. With adequate correlative or Solutions
predictive models for the fugacities, thermodynamics can be a powerful tool in the modeling of antibiotic solubilities. Activity coefficient models, such as NRTL, UNIQUAC, or UNIFAC, are frequently used to model solid-liquid equilibria (SLE).9-13 Several authors have used empirical equations to correlate a Except with acetone, which is 350.0 nm.
antibiotic solubilities.14-18 Recently, group-contribution methodshave been used to predict solubilities of larger organic com- Table 2. Mole Fraction, Solubility, and Standard Deviation of
Tetracycline.HCl in Different Solvents at Temperatures of
pounds.19,20 The development of new predictive models and the 288.15-310.15 K
improvement of the classic models for the phase equilibrium behavior of complex multifunctional molecules such as antibiot- ics has been difficult to achieve, because of the lack of reliable experimental data.21,22 In this work, activity coefficient models (NRTL and UNIQUAC) are used to correlate the experimental In the modeling of SLE, information regarding pure- component thermophysical properties (including the melting mp) and the enthalpy of fusion (∆Hfus), has an important role. Except for the T ciprofloxacin hydrochloride monohydrate, no experimental data were observed in the literature for the antibiotics that have been investigated. Therefore, different methods to predict these 2. Experimental Section
2.1. Materials. Tetracycline.HCl was purchased from Fluka
(Sigma-Aldrich, 98% purity). Ciprofloxacin.HCl and moxi- floxacin.HCl, each with high purity (>99.8%), were generously provided by Bayer HealthCare AG. Ethanol (absolute puriss.
p.a.) and acetone for gas chromatography (GC), with a purity the equilibrium was attained, the excess solid was allowed to of 99%, were supplied by Riedel de Hæn, 2-propanol (ACS settle and the composition of the liquid phase was determined reagent grade, for ultraviolet (UV) spectroscopy, with a purity via UV spectroscopy (Shimadzu model UV-160A UV-visible of 99.8% was supplied by Fluka, and ethyl acetate (99.8% recording spectrophotometer) with quartz cells at the corre- purity) was supplied by Lab-Scan. All of these compounds are sponding wavelength of the maximum of absorbance, according analytical reagents and were used without further purification.
to Table 1. In separate experiments, it was confirmed that 6 h Deionized and double-distilled water was used. Sodium chloride, of contact was the time necessary to achieve equilibrium. The which was used for ACS-ISO analysis (NaCl), was supplied antibiotics stability throughout the duration of experiment was by Panreac. Physiological serum was supplied by Fresenius confirmed through the analysis of their UV spectra before and after the equilibrium. Whenever necessary, the samples were 2.2. Experimental Procedure. The solubility of ciproflox-
diluted volumetrically with the respective solvent to obtain acin.HCl and moxifloxacin.HCl in water, ethanol, 2-propanol, absorbances in the linear calibration range for each system. All and acetone at temperatures of 293.15, 303.15, 313.15, and the experiment results were an average of at least three agreeing 323.15 K and that of tetracycline.HCl in the same solvents at 288.15, 293.15, and 303.15 K were measured. For the water +tetracycline.HCl system, additional data points were taken at 3. Experimental Results and Discussion
Sealed Erlenmeyer flasks that contained an excess of drug Tables 2-4 list the experimental results for the solubilities powder in the presence of a fixed volume of each of the pure of the three studied antibiotics in the above-mentioned solvents.
solvents were equilibrated for 6 h at each temperature, in a The solubility (S), which is reported in terms of milligrams of temperature-controlled bath. For higher temperatures (313.15 antibiotic per milliliter of solution), is an average of at least and 323.15 K), a thermostatic shaking bath (Julabo Shake Temp three agreeing independent experiments. The corresponding SW23) was used. For lower temperatures (288.15, 293.15, and standard deviation (sd) for each mean value is also reported.
303.15 K), an air bath that had been specifically built for this The solubility results obtained for ciprofloxacin.HCl in acetone purpose (with temperature control of (0.2 K) was used. After are below the accuracy limit of the experimental method used Ind. Eng. Chem. Res., Vol. 45, No. 18, 2006 Table 3. Mole Fraction, Solubility, and Standard Deviation of
floxacin.HCl are 1 order of magnitude more soluble in non- Moxifloxacin.HCl in Different Solvents at Temperatures of
aqueous solvents than ciprofloxacin.HCl. According to Stezow- 293.15-323.15 K
zki,23 the zwitterion is the most important form of tetracycline in aqueous solutions in the pH range of this work (6.2). In organic solvents, the non-ionized form generally is present, and the tetracycline.HCl presents extensive intramolecular hydrogen bonding and, thus, reduced polarity.
4. Modeling
4.1. Models. The phase equilibrium equation for a solid
solute, designated by subscript 2, that partly dissolves in a liquid solvent, at a temperature T and a pressure P can be written as f L(T, P, {xL}) ) f S(T,P, {xS}) where x denotes the mole fraction, the superscripts L and S denote the liquid and the solid phases, respectively. Assuming that the solid phase consists only of pure solid (x ) fugacity of component 2 in the solid phase is equal to the f L(T, P, x ) ) f S(T, P) Table 4. Mole Fraction, Solubility, and Standard Deviation of
The fugacity of component 2 in the liquid phase can be Ciprofloxacin.HCl in Different Solvents at Temperatures of
293.15-323.15 K
f L(T, P, x ) ) x γ (T, P, x )f L(T, P) where γ2 is the activity coefficient of the solute in the liquid phase. The Gibbs free energy of fusion, ∆fusG(T), is related to where R is the gas constant. ∆fusG(T) is computed by separately calculating the enthalpy and entropy of fusion (∆fusH(T) and ∆fusS(T), respectively), supposing that the melting of a solid (at T < Tmp) is performed in a three-step constant-pressure process (heating from T to Tmp, melting, subcooling to T).
Assuming that the difference of the isobaric heat capacity between the solid and the liquid state (∆CP) is independent oftemperature from T up to the melting temperature, the basic (0.02 mg/mL solution) and, thus, were discarded. The solubility equation for the solubility of a solid in a liquid can be written of tetracycline.HCl was only studied at temperatures up to 310.15 K, because it decomposes at higher temperatures.
Through analysis of the the solubility results obtained for the three antibiotics, it is possible to see that the solubility of a compound is determined by both the properties of both the solvent and the pure solid in equilibrium with the solvent. All the anitbiotics have the same solubility order; that is, they are A more detailed derivation can be found in the literature.24,25 more soluble in water than in ethanol, and they are more soluble The two terms in the main brackets of eq 5 are not of equal in ethanol than in 2-propanol and acetone. As expected, for all importance. The first term is the dominant one, whereas the solvent + antibiotic systems, the solubility increases as the second term is small, especially if T and T temperature increases. For a given antibiotic at a given tem- Unfortunately, that is not the case, because antibiotics have a perature, the solubility in water is ∼3 orders of magnitude greater than that in acetone. The solubility of the studied antibiotics in water is due to the presence of the hydrochloride neglected. Finally, the simplified equation that was used for group, which, in water, becomes Cl-, which leads to the formation of ionic species and, thus, promotes an enhancementin the solubility. Note that the solubility of tetracycline.HCl in water shows the strongest temperature dependence of the three antibiotics that have been studied. Tetracycline.HCl and moxi- Ind. Eng. Chem. Res., Vol. 45, No. 18, 2006 Two models were used to calculate the activity coefficients The values for the structural parameters are listed in Table 6 of the solute: NRTL and UNIQUAC. Because these are well- (presented later in this work). The values for the solvents have known activity coefficient models,26,27 only the essential equa- been taken from the Dortmund Databank.29 For the antibiotics, tions are given here. If NRTL is used, the activity coefficient the Bondi group contribution method30 was used to calculate 4.2. Property Data Used for the Modeling. To correlate
the experimental data using eq 6 (the models described in the ln γ ) x [τ ( G12 )2 + previous section), pure-component solute properties such as T and ∆fusH are needed. The Tmp value for tetracycline.HCl andciprofloxacin.HCl were taken from the literature.31,32 For moxifloxacin.HCl, because it decomposes before melting, no experimental data were found. Thus, the value used in the calculations was estimated from the Tmp value of ciprofloxacin- .HCl in combination with the group-contribution method of Marrero and Gani,33 by adding the contributions of the additional For a large number of binary systems, the nonrandomness factor groups that are present in the moxifloxacin molecule. The values R12 varies over a range of 0.20-0.47.25 We set R12 to a value for the melting points of the antibiotics are given in Table 5.
of 0.25 for all calculations, because it is the standard value for No experimental ∆fusH data were available for any of the the VTPLAN process simulator of Bayer.28 For the interaction antibiotics. Following Gupta and Heidemann,14 we calculated parameters (τij), the following temperature dependence was used: ∆fusH values, using the identity with a constant value for the entropy of fusion of ∆fusS ) 56.51 J mol-1 K-1. Yalkowski15 reported that the ∆ ij and bij are adjustable parameters. Except for the water (1) + tetracycline.HCl (2) system, the a drugs and rigid molecules of intermediate size can be estimated zero for all systems that have been investigated.
using this value, confirming earlier observations.16 The calcu- If UNIQUAC is used to calculate the activity coefficient of lated melting enthalpies are reported in Table 5, as well as the the solid in the solvent, the following equation is used: properties of the pure solvents, taken from the DIPPR database.34 To calculate the mole fractions from the solubility data, the the antibiotic solutions were measured using the gravimetric method. The data points could be regressed within experimental q ln(θ + θ τ ) + θ q ( τ12 - where z is the coordination number; Φ* and θ are the segment and the area fraction, respectively; r and q are the pure-component parameters of surface and volume, respectively, and For the temperature-dependent solvent densities F1, data were taken from the DIPPR database.34 The apparent liquid density of the solute (antibiotic) F2 was used as a regression parameter.
F2 was assumed to be independent of temperature within the investigated temperature range, which lies well below the melting temperature of the antibiotics. Values for F2 are given where bij were fitted to the experimental SLE data. For almostall binary systems investigated, aij were set to zero, because 5. Modeling Results
two binary parameters (b12 and b21) were sufficient to obtain a The NRTL and UNIQUAC model parameters were obtained good representation of the experimental data. Adjusted values by fitting the experimental solubility data. The objective for a12 and a21 were used only for the water (1) + tetracycline- function, F, which was used to adjust the model parameters, is .HCl (2) system, because the solubility of this system shows a temperature dependence that deviates significantly from thebehavior of the other binary systems.
l ) z(r - q ) - (r - 1) where i denotes a data point, NP is the number of data points,and and Φ* and θ are calculated with the equations {- [∆ H(T )/(RT)][1 - (T/T )]} (18) Comparisons of UNIQUAC correlations with the experimen- tal data are shown graphically in Figures 2-4. The solubitities cover several orders of magnitude. The representation of the Ind. Eng. Chem. Res., Vol. 45, No. 18, 2006 Table 5. Thermophysical Properties of the Antibiotics and Solvents Used in the Modeling
experimental data with both models is very good. The temper- used, both for UNIQUAC and NRTL. All other curves were ature dependence of the solubility of tetracycline.HCl in water calculated with two adjustable parameters for each binary deviates from its temperature dependence in the other solvents.
To model this behavior, four adjustable parameters had to be The optimized parameters for the two models used to describe the solubility of tetracycline.HCl, moxifloxacin.HCl, and cipro-floxacin.HCl, along with the average absolute deviation (AAD)and the relative absolute error (RAE) of the models are presentedin Tables 6-8, respectively. Overall, the difference betweenthe modeling results of UNQUAC and NRTL is small. For themoxifloxacin.HCl systems, the two models show almost identi-cal results. For the tetracycline.HCl systems, the UNIQUACmodel with two parameters is better than the NRTL model withtwo parameters (with R set to 0.25) to describe the temperaturedependence of the solubilities. This difference between themodels disappears when four parameters are used.
The use of the models for extrapolation toward higher and lower temperatures is demonstrated in Figure 5, which showsthe temperature versus mole fraction diagram of water +ciprofloxacin.HCl. The left side of the diagram represents theequilibrium between the aqueous solution and pure solid water.
Because of the high molecular weight of ciprofloxacin.HCl, thefreezing point depression is small (0.02 K at the eutectic point).
For this part of the diagram, the NRTL and UNIQUAC modelslead to similar results. The right side of the diagram representsthe equilibrium between the aqueous solution and pure solidciprofloxacin.HCl. In our calculations, we did not account forthe existence of ciprofloxacinhydrates and peritectic points, Figure 2. Experimental solubility data for tetracycline.HCl in (b) water,
because insufficient property data were known for ciproflox- ([) ethanol, (O) acetone, and (]) 2-propanol; the lines represent modeling acinhydrates. The NRTL model gives a smooth curve from the Figure 3. Experimental solubility data for moxifloxacin.HCl in (b) water,
Figure 4. Experimental solubility data for ciprofloxacin.HCl in (b) water,
([) ethanol, (O) acetone, and (]) 2-propanol; the lines represent modeling ([) ethanol, and (]) 2-propanol; the lines represent modeling using Ind. Eng. Chem. Res., Vol. 45, No. 18, 2006 Table 6. Optimized Parameters for UNIQUAC and NRTL Models and Respective Average Absolute Deviations and Relative Absolute Errors
for the Solubility of Tetracycline.HCl in Several Solvents
Table 7. Optimized Parameters for UNIQUAC and NRTL Models
Table 8. Optimized Parameters for UNIQUAC and NRTL Models
and Respective Average Absolute Deviations and Relative Absolute
and Respective Average Absolute Deviations and Relative Absolute
Errors for the Solubility of Moxifloxacin.HCl in Several Solvents
Errors for the Solubility of Ciprofloxacin.HCl in Several Solvents
with care. When four binary interaction parameters are used, eutectic point to the melting point of ciprofloxacin.HCl. The such as in the case of water + tetracycline.HCl, extrapolation UNIQUAC model gives a better presentation of the experimental to higher temperatures is not recommended.
data, which show a strong increase of antibiotic solubility in The modeling results should be, of course, viewed taking into the aqueous solution with increasing temperature. At higher account that a predicted Tmp value was used for moxifloxacin- temperatures, the UNIQUAC model leads to a curved behavior, .HCl, as well as predicted ∆fusH values were used for all indicating the existence of a peritectic point. It can be concluded antibiotics. Furthermore, ∆CP was assumed to be zero for all that when model parameters are fitted to experimental data antibiotics, because of the lack of experimental data, although within a very limited temperature and composition range, the temperatures considered were far away from the Tmp values extrapolation to the entire composition range shall be performed 6. Conclusions
New experimental data of solubility are obtained for tetra- cycline.HCl, moxifloxacin.HCl, and ciprofloxacin.HCl in severalsolvents: water, ethanol, 2-propanol, and acetone. The spec-trophotometric method is a good tool to determine the solubilityof those antibiotics, but it is limited for some solvents, such asacetone for ciprofloxacin.HCl, for which the solubility is <0.02mg/mL solution. The solubility is dependent on the solventintrinsic properties and solute-solvent interactions. All theantibiotics have the same solubility order; that is, they are moresoluble in water than in ethanol, and more soluble in ethanolthan in 2-propanol and acetone. As expected, for all solvent +antibiotic combinations, the solubility increases as the temper-ature increases. For a given antibiotic at a given temperature,the solubility in water is ∼3 orders of magnitude higher thanthat in acetone.
The modeling of the solid-equilibria (SLE) data, using NRTL and UNIQUAC, proves that these models can correlate wellthe solubility of antibiotics for the temperature range for whichexperimental data are available, with the UNIQUAC modelbeing, generally, slightly superior to the NRTL model, when Figure 5. Temperature versus mole fraction diagram for ciprofloxacin.HCl
only two adjustable parameters are used for each binary system.
+ water; symbols (O) represent experimental data, the solid lines represent Because the model parameters were fitted to experimental data modeling and extrapolations using NRTL, and the dashed lines refer toUNIQUAC.
within a limited temperature and composition range, extrapola- Ind. Eng. Chem. Res., Vol. 45, No. 18, 2006 tion to the entire composition range shall be performed with Phase Extraction And Liquid Chromatography with Fluorimetric Detection.
care. Finally, it should be noted that the use of more-accurate J. Chromatogr. A 2004, 1041, 27.
(8) Romero, S.; Bustamante, P.; Escalera, B.; Murab, P.; Cirri, M.
values of the pure-components properties could improve the Influence of Solvent Composition on the Solid Phase at Equilibrium with Saturated Solutions of Quinolones in Different Solvent Mixtures. J. Pharm.
Biomed. Anal. 2004, 35, 715.
List of Symbols
(9) Gmehling, J. G.; Anderson, T. F.; Prausnitz, J. M. Solid-Liquid Equilibria Using UNIFAC. Ind. Eng. Chem. Fundam. 1978, 17, 269.
(10) Atik, Z.; Ahlers, J.; Lohnmann, J.; Gmehling, J. Binary Solid Liquid Equilibria of N,N-Dimethylacetamide with 1,2-Dichloroethane, Dichlo- romethane, and 1-Propanol. J. Chem. Eng. Data 2001, 46, 1190.
(11) Teodorescu, M.; Wilken, M.; Wittig, R.; Gmehling, J. Azeotropic difference between the heat capacity in the liquid state and solid-liquid equilibria data for several binary organic systems and in the solid state at constant pressure containing one acetal compound. Fluid Phase Equilib. 2003, 204, 267.
F ) objective function, as described in eq 17 (12) Domanska, U.; Łachwa, J. (Solid + Liquid) phase equilibria of binary mixtures containing N-methyl-2-pyrrolidinone and alcohol at atmo- fusG ) Gibbs free energy of fusion spheric pressure. J. Chem. Thermodynam. 2003, 35, 1215.
(13) Domanska, U.; Bogel-Łukasik, E. Solid-liquid equilibria for systems containing 1-butyl-3-methylimidazolium chloride. Fluid Phase Equilib. 2004, 218, 123.
(14) Gupta, R. B.; Heidemann, R. A. Solubility Models for Amino Acids and Antibiotics. AlChE J. 1990, 36, 333.
(15) Chen, C.-C.; Zhu, Y.; Evans, L. B. Phase Partitioning of Biomol- S ) solubility (mg of solute per mL of solution) ecules of Amino Acids. Biotechnol. Prog. 1989, 5, 111.
(16) Zhu, Y.; Evans, L. B.; Chen, C.-C. Representation of phase equilibrium behaviour of antibiotics. Biotechnol. Prog. 1990, 6, 266.
(17) Gupta, R. B.; Kumar, R.; Betageri, G. V. Phase Behavior of Mixtures containing Antibiotics. Chloroamphenicol Partitioning. Ind. Eng. xS ) mole fraction of the solute in the solid phase Chem. Res. 1997, 36, 3954.
mole fraction of the solute in the liquid phase (18) Yu, X.; Zipp, G. L.; Davidson, G. W. R., III. The Effect of Temperature and pH on the Solubility of Quinolone Compounds: Estimation
of Heat of Fusion; Pharm. Res. 1994, 11, 522.
(19) Gracin, S.; Brinck, T.; Rasmuson, A. C. Prediction of Solubility of Organic Compounds in Solvents by UNIFAC. Ind. Eng. Chem. Res. 2002,
activity coefficient of the solute in solution (20) Gracin, S.; Rasmuson, A. C. Solubility of Phenylacetic Acid, p-Hydroxyphenylacetic Acid, p-Hydroxybenzoic Acid, and Ibuprofen in Pure Solvents. J. Chem. Eng. Data 2002, 47, 1379.
temperature-dependent binary interaction parameter (21) Yalkowski, S. H. Estimation of Entropies of Fusion of Organic Compounds. Ind. Eng. Chem. Fundam. 1979, 18, 108.
(22) Tsonopoulos, C.; Prausnitz, J. M. Activity Coefficients of Aromatic Solutes in Dilute Aqueous Solutions. Ind. Eng. Chem. Fundam. 1971, 10,
593.
(23) Stezowski, J. J. Chemical-Structural Properties of Tetracycline Derivates. I Molecular Structure and Conformation of the Free Base Derivatives. J. Am. Chem. Soc. 1976, 98, 6012.
(24) Sandler, S. I. Chemical and Engineering Thermodynamics, 3rd (25) Prausnitz, J. M.; Lichtenthaler, R. N.; de Azevedo, E. G. Molecular Thermodynamics of Fluid-Phase Equilibria, 3rd Edition; Prentice Hall:Upper Saddle River, NJ, 1999.
(26) Renon, H.; Prausnitz, J. M. Local Compositions in Excess Thermodynamic Functions for Liquid Mixtures. AIChE J. 1968, 14, 135.
(27) Abrams, D. S.; Prausnitz, J. M. Statistical Thermodynamics of Literature Cited
Liquid Mixtures. A New Expression for the Excess Gibbs Energy of Partly
of Completely Miscible Systems. AIChE J. 1975, 21, 116.
(1) Kola´ø, P.; Shen, J.-W.; Tsuboi, A.; Ishikawa, T. Solvent Selection (28) VTPLAN process simulator, www.bayertechnology.com.
for Pharmaceuticals. Fluid Phase Equilib. 2002, 194-197, 771.
(29) Dortmund Databank, available via the Internet at http://www.ddb- (2) Abildskov, J.; O’Connell, J. P. Predicting the Solubility of Complex Chemicals I. Solutes in Different Solvents. Ind. Eng. Chem. Res. 2003, 42,
(30) Poling, B. E.; Prausnitz, J. M.; O’Connell, J. P. The Properties of Gases and Liquids, 5th Edition; McGraw-Hill: New York, 2001.
(3) Schneider, S.; Brehm, G.; Schmitt, M.; Leypold, C.; Matousek, P.; (31) Tetracycline hydrochloride datasheet, Fluka, https://www.sigmaald- Towrie, M. Picosecond Time-Resolved Fluorescence of Tetracycline and rich.com/catalog/search/ProductDetail/Fluka/87130.
Its Complexes with Mg++ or Ca++. Central Laser Facility Annual Report (32) O’Neil, M. J., Smith, A., Budavari, S., Heckelman, P. E., Gallipeau, 2002/2002, 103-105, Lasers for Science Facility ProgrammesChemistry.
J. A. R., D’Arecca, M. A., Obenchain, J. R., Eds. The Merck Index: An (4) Owens, R. C. New Product Bulletin: Avelox. American Pharma- Encyclopedia of Chemicals, Drugs and Biologicals, 13th Edition; Merck ceutical Association: Washington, DC, 2000.
(5) Park, H.; Kim, T. H.; Bark, K. Physicochemical Properties Of (33) Marrero, J.; Gani, R. Group contribution based estimation of pure Quinolone Antibiotics in Various Environments. Eur. J. Med. Chem. 2002,
components properties. Fluid Phase Equilib. 2001, 183-184, 183.
(34) DIPPR 801 Database, available via the Internet at http://dippr.by- (6) Barbosa, J.; Barro´n, D.; Jime´nez-Lozano, E.; Sanz-Nebot, V.
Comparison Between Capillary Electrophoresis, Liquid Chromatography,Potentiometric and Spectrophotometric Techniques for Evaluation of pKa ReceiVed for reView January 13, 2006 Values of Zwitterionic Drugs in Acetonitrile-Water Mixtures. Anal. Chim. ReVised manuscript receiVed June 20, 2006 Acta 2001, 437, 309.
(7) Prat, M. D.; Benito, J.; Compan˜o´, R.; Herna´ndez-Arteseros, J. A.; Granados, M. Determination of Quinolones in Water Samples by Solid-

Source: http://path.web.ua.pt/file/A56%20Varanda%20Melo%20Caco%20Dohrn%20Makrydaki%20Voutsas%20Tassios%20Marrucho%20Antibiotics%20Part%20I.pdf