Professional Documents
Culture Documents
PBA - 8591 Arrhenius Kinetic of Ampicillin and Cephalexin
PBA - 8591 Arrhenius Kinetic of Ampicillin and Cephalexin
Journal: PBA Please e-mail or fax your responses and any corrections to:
E-mail: corrections.esil@elsevier.thomsondigital.com
Dear Author,
Please check your proof carefully and mark all corrections at the appropriate place in the proof (e.g., by using on-screen
annotation in the PDF file) or compile them in a separate list. Note: if you opt to annotate the file with software other than
Adobe Reader then please also highlight the appropriate place in the PDF file. To ensure fast publication of your paper please
return your corrections within 48 hours.
Any queries or remarks that have arisen during the processing of your manuscript are listed below and highlighted by flags in
the proof. Click on the ‘Q’ link to go to the location in the proof.
1 Short communication
11 a r t i c l e i n f o a b s t r a c t
12
13 Article history: A limited number of researches with application of the Arrhenius equation have been reported to drugs
14 Received 8 December 2011 and biopharmaceuticals in biological fluids at frozen temperatures. This paper describes stability studies
15 Received in revised form 6 April 2012 of ampicillin and cephalexin in aqueous solution and human plasma applying the Arrhenius law for
16 Accepted 9 April 2012
determination of adequate temperature and time of storage of these drugs using appropriate statistical
Available online xxx
analysis. Stability studies of the beta-lactams in human plasma were conducted at temperatures of 20 ◦ C,
17
2 ◦ C, −20 ◦ C and also during four cycles of freeze-thawing. Chromatographic separation was achieved
18 Keywords:
using a Shimpak C18 column, acetonitrile as organic modifier and detection at 215 nm. LC-UV–MS/MS
19 -Lactams
20 Q3 Human plasma was used to demonstrate the conversion of ampicillin into two diastereomeric forms of ampicilloic acid.
21 Arrhenius equation Stability studies demonstrated degradation greater than 10% for ampicillin in human plasma at 20 ◦ C, 2 ◦ C
22 Stability prediction and −20 ◦ C after 15 h, 2.7 days, 11 days and for cephalexin at the same temperatures after 14 h, 3.4 days
23 LC-DAD and 19 days, respectively, and after the fourth cycle of freezing–thawing. The Arrhenius plot showed
24 LC-UV–MS/MS good prediction for the ideal temperature and time of storage for ampicillin (52 days) and cephalexin
(151 days) at a temperature of −40 ◦ C, but statistical analysis (least squares method) must be applied to
avoid incorrect extrapolations and estimated values out uncertainty limits.
© 2012 Elsevier B.V. All rights reserved.
vention [9]. 41
26 Penicillins and cephalosporins antibiotics have been the most Stability studies have a great relevance in biological systems 42
27 widely used antimicrobial drugs at long of the years and they are to evaluate the chemical integrity of drugs and biopharmaceuti- 43
28 still considered as one of the most important groups of antibiotics cals and the determination of the kinetic of degradation of analytes 44
29 [1]. Despite their extensive and long standing use in therapeutic, including monitoring of degradation product using LC-tandem MS 45
30 there has been incomplete description of the stability of drugs, [10]. Stability studies and classical Arrhenius method have been 46
31 in specially the beta-lactams antibiotics in biological fluids. Other scarcely used in biopharmaceutical analysis to assess the chem- 47
32 therapeutic classes of drugs are naturally unstable in biofluids and ical stability, potency, purity, quality and shelf life of drugs and 48
33 can be including: acetylsalicylic acid [2], thiolic derivatives [3], biotechnological products [11] under isothermal or non-isothermal 49
34 Zopiclone [4], antimalarial drugs [5] and benzodiazepines [6]. conditions [12], mainly at the frozen temperatures. Statistical anal- 50
35 Stability study in biofluids is an important prerequisite for ysis was applied to evaluate the data uncertainty using linear or 51
36 validation of a bioanalytical method [7] and has practical implica- non-linear methods [13]. 52
37 tions in the drug determination during pharmacokinetic evaluation Application of the short-term stability studies can be an alterna- 53
38 [8] and can avoid error in bioanalytical laboratory in the con- tive to reduce the cost of the bioanalytical validation program and 54
39 text to assurance a more effective patient management, including to evaluate the behavior of drugs in biological matrices. This article 55
ity studies. The purpose of this study was to apply the Arrhenius 58
∗ Corresponding author. Tel.: +55 021 82 3214 1155; fax: +55 021 82 3214 1155.
equation in the stability study of ampicillin and cephalexin in order 59
E-mail addresses: ticianogn@yahoo.com.br (T.G. do Nascimento),
to determine adequate temperature and time of storage of these 60
eoliveira@gmail.com (E. de Jesus Oliveira), ruiomacedo@yahoo.com.br
(R.O. Macêdo).
drugs using an appropriate statistical analysis. 61
0731-7085/$ – see front matter © 2012 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.jpba.2012.04.010
Please cite this article in press as: T.G. do Nascimento, et al., Short-term stability studies of ampicillin and cephalexin in aque-
ous solution and human plasma: Application of least squares method in Arrhenius equation, J. Pharm. Biomed. Anal. (2012),
http://dx.doi.org/10.1016/j.jpba.2012.04.010
ARTICLE IN PRESS
G Model
PBA 8591 1–6
2 T.G. do Nascimento et al. / Journal of Pharmaceutical and Biomedical Analysis xxx (2012) xxx–xxx
62 2. Experimental CQ2 and CQ3 , were transferred into cryovials and stored at 20 ◦ C, 119
2 ◦ C and −20 ◦ C during 15 days, 30 days and 103 days, respectively, 120
63 2.1. Reagents and chemicals to promote sample aging. Non-spiked fresh plasma samples were 121
also stored in the same conditions and at the moment of analysis 122
64 Ampicillin USP (from Ferjinsá, Mexico) and cephalexin phar- were spiked with analytes in order to obtain in all the samples the 123
65 macopeial standard (from Brazilian Pharmacopeia) were used same nominal concentration of the aged samples and calibration 124
66 throughout the study. Perchloric acid, phosphoric acid, potas- curve samples. Stability of beta-lactams was evaluated by compar- 125
67 sium dihydrogen phosphate and HPLC-grade acetonitrile were ing the responses of ampicillin and cephalexin in stored samples 126
68 purchased from Merck, Darmstadt, Germany. Water was purified and with those freshly prepared sample at the same nominal con- 127
69 with a reverse osmosis system connected to an ion exchange unit centration level. An upper and lower limit of 10% of the original 128
70 (Gehaka, Brazil). All other reagents were analytical grade. concentration was established as acceptance criteria for stability. 129
74 autosampler SIL-10AD, and a controller module SCL-10A (all from calculate the order of reaction and degradation rate constant (k) in 133
75 Shimadzu) coupled to a personal computer running the software each temperature studied 20 ◦ C, 2 ◦ C and −20 ◦ C using Arrhenius 134
76 Shimadzu Class VP for data acquisition. The mobile phase consisted equation as time function (Eqs. (1)–(3)) [15]. In each temperature, 135
77 of acetonitrile: dihydrogen phosphate (pH 3.5; 0.1 M) (12.5: 87.5, the mean of the degradation rate constant (k) (from triplicate exper- 136
78 v/v). The pH of the phosphate buffer was adjusted with 4% phos- iments) was obtained and did not presented variation major than 137
79 phoric acid. Separation was achieved at 25 ◦ C using a Shimpak C18 15% of relative standard deviation (RSD). 138
91 Column (250 mm × 4.6 mm; i.d 5 m) fitted with Shimpak secu- final time, t0 = initial time; tx = final time, k0 = zero-order degra- 144
92 rity guard C18 column (4.0 mm × 3.0 mm; i.d. 5 m) kept at 40 ◦ C dation rate constant, k1 = first-order degradation rate constant, 145
93 in a column oven. A triple quadrupole mass spectrometer (Quat- k2 = second-order degradation rate constant, A = preexponential 146
94 tro LC, Micromass) was used for determination of ampicillin, its factor, Ea = energy of activation, R = gas constant, T = absolute tem- 147
95 related degradation products and cephalexin. It was operated in perature in Kelvin. 148
96 positive electrospray ionization mode with a capillary voltage of The degradation rate constant (k) at sub-zero temperature was 149
97 2.70 kV, sampling cone voltage of 21 V, extraction cone voltage of estimate by extrapolation of the best linear regression with least 150
98 3 V, desolvation gas temperature of 250 ◦ C (N2 ) and source block squares method for the temperature of −30 ◦ C and −40 ◦ C, which 151
99 temperature of 110 ◦ C. Full scan acquisition was made in the range were obtained according to graphical analysis using Arrhenius 152
100 of 40–400 m/z with the multiplier voltage at 650 V. equation (4) and the time of storage was also calculated [15]. It was 153
101 2.3. Sample preparation estimating the standard error (SYX ) in the Arrhenius plot (ln k × 1/T) 155
using the least squares method. The 95% confidence interval and 156
102 2.3.1. Human plasma samples 95% prediction interval were used to obtain the upper and lower 157
103 Human plasma (250 mL) collected from healthy volunteers and limits of variations [16]. Analysis of variance (ANOVA-one way) and 158
104 stored in bags containing heparin as anticoagulant was acquired P-value were used to check if the experimental values followed the 159
105 from the Hemotherapy Center of Paraíba (João Pessoa, Paraíba, Arrhenius plot. 160
106 Brazil) after approval from the research ethics committee. The
107 method for quantification of ampicillin and cephalexin during the 3. Results and discussion 161
111 tively transferred to a 10 mL volumetric flask to obtain a (5 mg/mL) of two new chromatographic peaks at 5.5 and 6.1 min became 164
112 stock solution in ultrapure water. Successive dilutions were made apparent in the period after 15 days of storage. These peaks were 165
113 to prepare quality control samples at concentrations of 1.0, 2.5, 5.0, denominated AMP A1 and AMP A2. Additional experiments were 166
114 10.0 and 50.0 g/mL. The stock solutions were stored in cryovials done to determine the origin of the degradation products. Two 167
115 of 2.0 mL. stock solutions containing only one analyte were submitted to 168
116 2.4. Stability studies in stock solution and human plasma appeared in the ampicillin stock solution (10 g/mL) but not in 170
117 Aliquots of 1.0 mL from the quality control samples (5.0, 10.0 (AMP A1 + AMP A2) were formed uniquely from ampicillin after 172
118 and 50.0 g/mL) in stock solution or human plasma, named CQ1 , 10 days of storage at 20 ◦ C (Figs. 1 and 3A and B). 173
Please cite this article in press as: T.G. do Nascimento, et al., Short-term stability studies of ampicillin and cephalexin in aque-
ous solution and human plasma: Application of least squares method in Arrhenius equation, J. Pharm. Biomed. Anal. (2012),
http://dx.doi.org/10.1016/j.jpba.2012.04.010
ARTICLE IN PRESS
G Model
PBA 8591 1–6
T.G. do Nascimento et al. / Journal of Pharmaceutical and Biomedical Analysis xxx (2012) xxx–xxx 3
(AMP A2) in the extracted ion chromatograms and the mass spec- 194
tral presented this evidence and has confirmed the formation of 195
two epimeric forms, 1 (5S, 6R) and 2 (5R, 6R), of the ampicilloic acid 196
ampicilloic acid 1 (5S, 6R) and 2 (5R, 6R) in acid aqueous medium 198
stock solution. The opening of the -lactamic ring by acid hydrolysis 200
loic acid diastereoisomers (Fig. 3). Zhu et al. [17] reported several 202
ampicilloic acid, in which the ampicilloic acid (5S, 6R) exhibited 204
a lower retention time than its (5R, 6R) diastereoisomer and our 205
Fig. 1. (A) Chromatogram of drug-free plasma (a) and spiked plasma (b) with (1) acidic or alkaline conditions (pH 1 or pH 8) and in this conditions 209
ampicillin (10 g/mL) and (2) cephalexin (10 g/mL). (B) Chromatogram of stock cephalexin is 180 times more stable than ampicillin [19]. 210
solution (1) of ampicillin (10 g/mL) and (2) cephalexin (10 g/mL) after 15 days
of storage at 20 ◦ C with detection of two degradation products (AMP A1 and AMP
A2). (C) Chromatogram of separated stock solution of ampicillin (10 g/mL) and (2)
cephalexin (10 g/mL) after 10 days of storage at 20 ◦ C.
Please cite this article in press as: T.G. do Nascimento, et al., Short-term stability studies of ampicillin and cephalexin in aque-
ous solution and human plasma: Application of least squares method in Arrhenius equation, J. Pharm. Biomed. Anal. (2012),
http://dx.doi.org/10.1016/j.jpba.2012.04.010
ARTICLE IN PRESS
G Model
PBA 8591 1–6
4 T.G. do Nascimento et al. / Journal of Pharmaceutical and Biomedical Analysis xxx (2012) xxx–xxx
Fig. 4. Quantitative loss mass of ampicillin (A) with simultaneous conversion in two degradation products AMP A1 and AMP A2 in stock solution; (B) quantitative loss mass
of cephalexin in stock solution. Stability studies of ampicillin and cephalexin in human plasma (C) storage at room temperature, (D) storage at 2 ◦ C, (E) storage at −20 ◦ C and
(F) after freeze–thaw cycles. Estimative of 95% confidence interval (solid line) and 95% prediction interval (dashed line) for the extrapolation up to the temperature of −40 ◦ C
of the degradation rate constant (k1 ) of ampicillin (G) and cephalexin (H) using Arrhenius plot.
211 3.2. Stability studies in human plasma which was observed in the fourth cycle of freezing and thawing 230
Please cite this article in press as: T.G. do Nascimento, et al., Short-term stability studies of ampicillin and cephalexin in aque-
ous solution and human plasma: Application of least squares method in Arrhenius equation, J. Pharm. Biomed. Anal. (2012),
http://dx.doi.org/10.1016/j.jpba.2012.04.010
ARTICLE IN PRESS
G Model
PBA 8591 1–6
T.G. do Nascimento et al. / Journal of Pharmaceutical and Biomedical Analysis xxx (2012) xxx–xxx 5
Table 1
Q5 Determination of order of reaction, degradation rate constants and estimative of storage time at sub-zero temperature for ampicillin and cephalexin applying the least
Q4
squares method in Arrhenius equation.
k0 r2 k1 r2 k2 r2
Ampicillin
20 8.77 ± 0.87 0.983 0.1157 ± 0.015 0.990 0.00157 ± 0.00027 0.995
2 2.88 ± 0.50 0.966 0.0381 ± 0.0090 0.962 0.000523 ± 0.0016 0.953
−20 0.81 ± 0.18 0.991 0.0097 ± 0.0013 0.987 0.000109 ± 0.000032 0.982
Cephalexin
20 8.79 ± 1.39 0.989 0.1237 ± 0.0024 0.993 0.00171 ± 0.00005 0.993
2 2.32 ± 0.14 0.987 0.0312 ± 0.0025 0.974 0.000432 ± 0.00005 0.956
−20 0.42 ± 0.12 0.910 0.0054 ± 0.0002 0.899 0.000061 ± 0.000006 0.878
Statistical parameters
Ampicillin Cephalexin
Linear regression y = 13.387 − 4568.1x ± 213 Linear regression y = 17.609 − 5781.3x ± 177
r2 0.9978 r2 0.9991
SYX 0.08169 SYX 0.06788
ln k (−40 ◦ C) −6.219 ± 0.1412 ln k (−40 ◦ C) −7.204 ± 0.1173
95%CI of ln k(−40 ◦ C) −8.013 to −4.425 95%CI of ln k(−40 ◦ C) −8.695 to −5.713
F 469.7 F 1066
P < 0.0500 0.0297 P < 0.0500 0.0195
Ampicillin Cephalexin
Each value is expressed as mean ± SD of three determinations of the QC samples. k0 = (C0 − Cx )/(tx − t0 ); k1 = ln[Cx /C0 ]/(tx − t0 ); k2 = [(1/Cx ) − (1/C0 )]/(tx − t0 ).
a
k2 = second-order degradation rate constant; t2(90) = 0.001/k2 .
b
k1 = first-order degradation rate constant; t1(90) = 0.105/k1 .
248 correlation and residual analysis, which was calculated to deter- has reduced the hyperbolic amplitude on the regression analysis 275
249 mine the experimental error of the best linear curve for ampicillin [16]. 276
250 (6.6%) and cephalexin (5.5%). In some cases, linear methods have Ampicillin and cephalexin could be preserved for a more long 277
251 some advantages over the non-linear methods because they are time of storage at sub-zero temperature (−30 ◦ C and −40 ◦ C), what 278
252 robust to deviations from homoscedasticity [13]. Fig. 4G and H would be sufficiently to preserved ampicillin for periods of between 279
253 showed the best regression curve with application of the estimative 20 and 50 days, while cephalexin would be preserved for 52 and 280
254 of confidence interval for the extrapolation calculi of the (k1 ) val- 141 days, respectively. The extrapoled degradation rate constants 281
255 ues of ampicillin and cephalexin at sub-zero temperature (−40 ◦ C). (k1 ) values of ampicillin and cephalexin presented relative standard 282
256 Estimative of confidence interval of degradation rate constant (k1 ) deviation (RSD) of 2.3% and 1.6%, respectively, as demonstrated in 283
257 was extrapolated for temperatures of −50 ◦ C, but there was no suc- Table 1 and thus adequate in the previsibility of the extrapolated 284
258 cess in the calculi [16]. The analysis of variance (F) and P value time of storage. 285
261 and were dependent on the storage temperature [15] (Table 1).
262 The best regression curve presented in Fig. 4G and H has demon- Ampicillin and cephalexin were presented instability in human 287
263 strated good estimative of standard error (SYX ) for ampicillin (8.2%) plasma after short periods of storage at temperature of −20 ◦ C. 288
264 and cephalexin (6.8%). Ampicillin presented greater variation in Based on this special class of drugs very unstable in biological flu- 289
265 (SYX ) and thus resulting in major amplitude in the hyperbolic effect ids it was considered the propose to determinate the adequate time 290
266 of the confidence bands in relation to cephalexin. The planning and temperature of storage in short-term stability studies during 291
267 of this stability studies only in three different temperatures was the step of method validation. 292
268 not sufficient to minimize the amplitude of the hyperbolic effect of The beta-lactams, ampicillin and cephalexin, when submitted to 293
269 confidence bands. The minimum sample size of (k1 ) and the esti- natural processes of degradation in aqueous medium or biological 294
270 mative of standard error (SYX ) contributed for the large confidence conditions has attendant to the classical Arrhenius method. The 295
271 interval of the ln k (−40 ◦ C) for ampicillin (28.8%) and cephalexin Arrhenius equation was useful to evaluate the degradation kinetic 296
272 (20.7%) (Table 1). Substitution of the mean values of rate constant of beta-lactams and estimate the ideal time and temperature of 297
273 (k1 ) in each temperature by individual values of rate constants storage, but statistical tools must be employed to avoid incorrect 298
274 (k1 ) obtained for each quality control samples (CQ1, CQ2 and CQ3) extrapolations and estimated values out uncertainty limits. 299
Please cite this article in press as: T.G. do Nascimento, et al., Short-term stability studies of ampicillin and cephalexin in aque-
ous solution and human plasma: Application of least squares method in Arrhenius equation, J. Pharm. Biomed. Anal. (2012),
http://dx.doi.org/10.1016/j.jpba.2012.04.010
ARTICLE IN PRESS
G Model
PBA 8591 1–6
6 T.G. do Nascimento et al. / Journal of Pharmaceutical and Biomedical Analysis xxx (2012) xxx–xxx
300 Acknowledgements M. Laurentie, J.C. Nivet, The SFSTP guide on the validation of chromatographic 328
methods for drug bioanalysis: from the Washington Conference to the labora- 329
tory, Anal. Chim. Acta 391 (1999) 135–148. 330
301 The authors wish to thank CNPq/CAPES/FINEP/ANVISA-DF for [9] P. Bonini, M. Plebani, F. Ceriotti, F. Rubboli, Errors in Laboratory Medicine, Clin. 331
302 financial support. Chem. 48 (2002) 691–698. 332
[10] C. Prakash, C.L. Shaffer, A. Nedderman, Analytical strategies for identifying drug 333
metabolites, Mass Spectrom. Rev. 26 (2007) 340–369. 334
303 References [11] H. Tokumitsu, Y. Fukumori, Y. Tsuda, Y. Okada, F. Kasuya, K. Igarashi, M. Fukui, 335
Degradation of a Novel Tripeptide, tert-Butoxycarbonyl-Tyr-Leu-Val-CH2Cl, 336
304 [1] R. Fernandez-Torres, M.O. Consentino, M.A.B. Lopez, M.C. Mochon, Simulta- with inhibitory effect on human leukocyte elastase in aqueous and in biological 337
305 neous determination of 11 antibiotics and their main metabolites from four fluids, Chem. Pharm. Bull. 45 (1997) 1845–1850. 338
306 different groups by reversed-phase high-performance liquid chromatography- [12] A. Oliva, J.B. Fariña, M. LIabrés, An improved methodology for data analysis in 339
307 diode array-fluorescence (HPLC-DAD-FLD) in human urine samples, Talanta 81 accelerated stability studies of peptide drug: practical considerations, Talanta 340
308 (2010) 871–880. (2012), http://dx.doi.org/10.1016/j.talanta.2012.03.011. 341
309 [2] D. Dadgar, J.W. Hooper, D. Lessard, K. Selinger, A new more reliable method for [13] A.G. Asuero, J.M. Bueno, Fitting straight lines with replicated observations by 342
310 the determination of the long term frozen stability of drug ins biological fluids, linear regression. IV. Transforming data, Crit. Rev. Anal. Chem. 41 (2011) 36–69. 343
311 in: 8th International AAPS Conference, Orlando, Florida, 1993. [14] T.G. Nascimento, C.F.S. Aragão, F.D. Medeiros, E.J. Oliveira, R.O. Macêdo, Val- 344
312 [3] M. Jemal, D.J. Hawthorne, Quantitative determination of BMS186716, a thiol idation of a method for determination of ampicillin in human plasma using 345
313 compound, in dog plasma by high-performance liquid chromatography- LC-DAD, J. Chromatogr. Sci. 47 (2009) 749–755. 346
314 positive ion electrospray mass spectrometry after formation of the methyl [15] J.T. Carstensen, Drug Stability: Principles and Practice, vol. 43, Marcel Dekker, 347
315 acrylate adduct, J. Chromatogr. B 693 (1997) 109–116. New York, 1990. 348
316 [4] G.H. Nilsson, F.C. Kugelberg, R. Kronstrand, J. Ahlner, Stability tests of zopiclone [16] D.M. Levine, M.L. Berenson, D. Stephan, Estatística: Teoria e aplicações usando 349
317 in whole blood, Forensic Sci. Int. 200 (2010) 130–135. o Microsoft® Excel em português, LTC–Livro s Técnicos e Científicos S.A, Rio de 350
318 [5] O.M.S. Minzi, M. Rais, J.O. Svensson, L.L. Gustafsson, O. Ericsson, High- Janeiro, 2000. 351
319 performance liquid chromatographic method for determination of amodi- [17] Y. Zhu, E. Roets, Z. Ni, M.L. Moreno, E. Porqueras, J. Hoogmartens, Evaluation of 352
320 aquine, chloroquine and their monodesethyl metabolites in biological samples, liquid chromatography methods for the separation of ampicillin and its related 353
321 J. Chromatogr. B 783 (2003) 473–480. substances, J. Pharm. Biomed. Anal. 14 (1996) 631–639. 354
322 [6] A.E. Mahjoub, C. Staub, Stability of benzodiazepines inwhole blood samples [18] E. Verdon, R. Fuselier, D. Hurtaud-Pessel, P. Couëdor, N. Cadieu, M. Laurentie, 355
323 stored at varying temperatures, J. Pharm. Biomed. Anal. 23 (2000) 1057–1063. Stability of penicillin antibiotic residues in meat during storage ampicillin, J. 356
324 [7] D. Dadgar, P.E. Burnett, M.G. Choc, K. Gallicano, J.W. Hooper, Application issues Chromatogr. A 882 (2000) 135–143. 357
325 in bioanalitical method validation, sample analysis and data reporting, J. Pharm. [19] T. Yamana, A. Tsuji, Comparative stability of cephalosporins in aqueous 358
326 Biomed. Anal. 13 (1995) 89–97. solution: kinetics and mechanisms of degradation, J. Pharm. Sci. 65 (1976) 359
327 [8] Ph. Hubert, P. Chiap, J. Crommen, B. Boulanger, E. Chapuzet, N. Mercier, S. 1563–1574. 360
Bervoas-Martin, P. Chevalier, D. Grandjean, P. Lagorce, M. Lallier, M.C. Laparra,
Please cite this article in press as: T.G. do Nascimento, et al., Short-term stability studies of ampicillin and cephalexin in aque-
ous solution and human plasma: Application of least squares method in Arrhenius equation, J. Pharm. Biomed. Anal. (2012),
http://dx.doi.org/10.1016/j.jpba.2012.04.010