CHAPTER 1 INTRODUCTION

Transcription

1 CHAPTER 1 INTRODUCTION 1.1 Background In the last few years, special concern has arisen in the food safety field on animal muscle tissues due to the potential risk of antibiotic resistance. Quinolones are the important antibacterial agents used both in veterinary and human medicine today. Quinolones are used in veterinary medicine for (1) prophylaxis, (2) therapy, (3) infection control and (4) growth promotion in healthy animals. They are active against both Gram-negative and Gram-positive organisms through the inhibiting DNA gyrase of the bacterial cell. However, quinolones widely used in animal medicine has led to residues in edible tissue. Their residues need to be controlled because quinolones can contribute to adverse effects on human health. To reduce the risk in a health problem for consumers the European Union (EU), the Joint FAO/WHO Expert committees on food additives (JECFA) have set Maximum Residue Limits (MRLs) for quinolones in animal tissues (Regulation EC No. 37/2010). For this reason, analytical methods determining antibiotic residues at low concentration in animal muscle tissues are needed. Several analytical methodologies have been developed to determine quinolones in difference animal muscle tissues sample. Liquid chromatographic (LC) methods based on UV and fluorescence detection for quinolones analysis have been reported in shrimp tissue (Karbiwnyk, Carr, Turnipseed, Andersen & Miller, 2007), chicken breasts (Stubbings & Bigwood, 2009), chicken tissue (Qiao & Sun, 2010), fish muscle (Alegre, Vicente, Romero & Broch, 2010), porcine tissue (Hu Yu, Mu & Hu, 2012), salmon tissue (Evaggelopoulou & Samanidou, 2013). Liquid chromatography-mass spectrometry

2 2 (LC-MS) for determination of quinolones has been reported in swine liver, bovine kidney, chicken muscle & liver, fish muscle (Yu, Tao, Chen, Pan, Liu, Wang, Huang, Dai, Peng, Wang & Yuan, 2012). Although, LC-MS can provide higher sensitivity but it is strongly affected by matrix and more expensive. Conversely, LC methods based on UV detector could not only reach low detection limits but also decrease the analytical expense. The present work focuses on high-performance liquid chromatography technique using monolith column for determination of six quinolone antibiotics. In addition, the sample preparation techniques for cleaning and extraction are studied. 1.2 Objective 1. To optimize the high-performance liquid chromatography (HPLC) technique with chromolith column for fast screening of quinolones. 2. To study the sample preparation methods before HPLC analysis. 3. To validate and apply the proposed method for determination of quinolones in chicken muscles. 1.3 Scope of study 1. The optimized parameters include mobile phase ph, percentage of acetonitrile in mobile phase, and types of buffer. The analytes are marbofloxacin (MAR), ciprofloxacin (CIP), lomefloxacin (LOM), enrofloxacin (CIP), sarafloxacin (SAR), and difloxacin (DIF).

3 3 2. Two methods of sample preparation are investigated and compared. These includes QuEChERS (Quick, Easy, Cheap, Effective, Rugged and Safe), and solid phase extraction (SPE). 3. The proposed method under the optimized conditions is validated in terms of limit of detection, limit of quantification, calibration curve, accuracy, precision, extract recovery. Finally, the proposed method is applied to determine quinolones in chicken muscles collected from local markets.

4 4 CHAPTER 2 LITERATURE REVIEWS 2.1 Quinolones Quinolones were derived by quinine. The general quinolone structure consists of a 1-substituted-1,4-dihydro-4-oxopyridine-3-carboxylic moiety altogether with an aromatic or heteroaromatic ring. The quinolones structure was affected to gyrase inhibition of the bacterial cell. Figure 2-1 shows the structure of quinolone (Macgowan & Andersson, 2003). Figure 2-1 Structure of quinolone (Monique & Andersson, 2003) In this work, marbofloxacin, ciprofloxacin, lomefloxacin, enrofloxacin, sarafloxacin and difloxacin were studied because the groups of quinolones are frequently used in human and veterinary medicine. The chemical structure of target quinolones is shown in Figure 2-2.

5 5 1. Marbofloxacin (9-fluoro-2,3-dihydro-3-methyl-10-(4-methyl-1-pipera zinyl)-7-oxo-7h-pyrido[3.2.1-ij] [4.2.1]benzoadiazine-6-carboxylic acid), the reaction of marbofloxacin leads to killing of bacteria by inhibition of DNA-gyrase. Marbofloxacin was developed exclusively for the treatment of skin and soft tissue infections and urinary tract infections (Hunter, Koch, Coke, Caepenter, & Isaza, 2007). Antimicrobial properties of marbofloxacin may be advantageous for use in chicken. Marbofloxacin differs from other quinolones on account of its oxadiazine ring, which may provide this molecule some benefit, such as longer half-life, higher bioavailability (Voermans, Soest, Duijkeren, & Ensimk, 2006). 2. Ciprofloxacin (1-cyclopropyl-6-fluoro-1,4-dihydro-4-oxo-7-(1-pipera zinyl)-3-quinoline carboxylic acid), ciprofloxacin differs from other quinolones on position of fluorine atom at the 6-position, a piperazine moiety at the 7-position and a cyclopropyl ring at the 1-position. Ciprofloxacin is effective against most Gram negative organisms. Ciprofloxacin is well absorbed after oral administration and is widely distributed throughout the body. Ciprofloxacin is used to treat infections of the skin, lungs, airways, bones, and joints in chicken. Ciprofloxacin is also frequently used to treat urinary tract infections caused by bacteria such as E.coli. Ciprofloxacin is effective in treating infectious diarrheas caused by E. coli (Poudel, & Thapa, 2013). 3. Lomefloxacin (1-Ethyl-6,8-difluoro-7-(3-methylpiperazin-1-yl)-4-oxo quinoline-3-carboxylic acid hydrochloride), lomefloxacin is used to treat respiratory tract infections in chicken. Lomefloxacin is also used in meningitis, osteomyelitis, urinary tract infections, sexually transmitted diseases, bacteraemia, gastrointestinal infections and treatment of tuberculosis (Amran, Hossain, Baki, Amjad, Sultana, & Hossain, 2011).

6 6 4. Enrofloxacin (1-cyclopropyl-7-(4-ethyl-1-piperazinyl)-6-fluoro-1,4- dihy dro-4-oxo-3-quinoline carboxylic acid (hydrochloride), enrofloxacin is used effectively in the treatment of respiratory tract, urinary tract, skin, soft tissues, and bone. Enrofloxacin is a fluoroquinolone that was developed exclusively for veterinary use in cattle, pigs, dogs, cats and forthe treatment of respiratory and intestinal diseases in chicken ( Khargharia, Barua, Mohan, & Bhattacharya, 2005). 5. Sarafloxacin (6-Fluoro-1-(4-fluorophenyl)-4-oxo-7-piperazin-1-ylquino line-3-carboxylic acid hydrochloride), the reaction of sarafloxacin leads to inhibiting the structure and function of DNA gyrase, a bacterial which is an essential enzyme for DNA replication and transcription. Sarafloxacin is used in the drinking water of chickens to treat bacterial infections caused by E. coli (Elghany, & Madian, 2011). 6. Difloxacin (1-pfluorophenyl-6-fluoro-1,4-dihydro-4-oxo-7-(4-methyl- 1-piperazinyl)-3-quinolone-carboxylic acid hydrochloride), difloxacin is used to treat bacterial infections in the urinary tract, the respiratory tract, and the skin in chicken (Hoven, Wagenaar, & Walker, 2000). O O O O F COOH F COOH F OH N N N N N N N O N HN H 3 C CH HN F 3 marbofloxacin ciprofloxacin lomefloxacin O O O F COOH F COOH F COOH N N N N N N HN N H 3 C H N 3 C F F enrofloxacin sarafloxacin difloxacin Figure 2-2 Chemical structures of the target quinolones

7 7 Several quinolones are used for therapy of animals in many countries. Nevertheless, the usage of these quinolones differs greatly as regards animal species, indications and geographic spread. The datas are summarized in Table 2-1. Quinolones are used for treatment of animal disease in many countries of the world. The use of quinolones in animal production was noted. It would effect in residues in edible tissue, which was potentially harmful to humans health (resistance in humans). So, the European Union (EU) has established maximum residue limits (MRLs) for quinolone residues in animal tissues (Table 2-2). Table 2-1 Quinolones licensed for use in food animals by region of the world (WHO Meeting Geneva, & Switzerland, 1998) Region Livestock Poultry Pet animals Fish Europe enrofloxacin, enrofloxacin, enrofloxacin, sarafloxacin, flumequine, difloxacin, difloxacin, (oxolinic acid) marbofloxacin, flumequine, marbofloxacin danofloxacin oxolinic acid USA none enrofloxacin, enrofloxacin, none sarafloxacin difloxacin, orbifloxacin Japan enrofloxacin, enrofloxacin, enrofloxacin, oxolinic acid danofloxacin, danofloxacin, orbifloxacin orbifloxacin, ofloxacin, difloxacin, vebufloxacin, oxolinic acid oxolinic acid Asia enrofloxacin, enrofloxacin, enrofloxacin oxolinic acid danofloxacin, ciprofloxacin, enrofloxacin, ciprofloxacin danofloxacin, flumequine ofloxacin, flumequine, norfloxacin

8 8 Table 2-2 The MRL concentration of each quinolone (European commission, 2010) Pharmacologically animal species MRL (µg kg -1 ) Target tissue Active substance Enrofloxacin+ Bovine 100 muscle Ciprofloxacin Ovine 100 muscle Porcine 100 muscle Poultry 100 muscle Sarafloxacin Salmon 30 muscle Poultry 100 Liver Difloxacin Porcine 400 muscle Poultry 300 muscle Bovine 400 muscle Turkey 300 muscle Marbofloxacin Bovine 150 muscle Porcine 150 muscle Lomefloxacin All food product No MRLs required No applicable 2.2 High performance liquid chromatography High performance liquid chromatography (HPLC) is a chromatographic technique based on the separation of molecules because of differences in their structure. The analytes in the sample will have different interactions with the stationary support in the column. So, Different compounds can be separated from each other. Figure 2-3 shows the schematic of liquid chromatography. The main types of liquid chromatography are normal-phase liquid chromatography, reversed phase

9 9 liquid chromatography, ion-exchange liquid chromatography and size-exclusion chromatography. Reversed phase liquid chromatography is a popular type of liquid chromatography because it utilizes an aqueous mobile phase that is compatible with most biological samples. This type uses more than 70% of all liquid chromatography. Reversed phase has a non-polar stationary phase and polar mobile phase. The most popular column packing material is octadecylsilyl silica (C18). Silica is covalently modified by C18 functional group (Mansoor, 2002). Figure 2-3 Diagram of liquid chromatography system 2.3 The theory of chromatography The plate theory The related terms are commonly used as quantitative measures for the efficiency of chromatographic column. The height equivalent to a theoretical plate (HETP, H) and plate count or number of theoretical plates (N) are related by the equation (2-1) where L is the length of the column. N = L H (2-1)

10 10 The efficiency of chromatographic columns increases as the plate count term N becomes greater and the height equivalent to a theoretical plate term H becomes lower Van Deemter s equation The peak broadening is directly related to the column efficiency. The equation was developed by Dutch chemical engineers in 1956 and used to explain chromatography efficiency. Van Deemter s equation is well known. H = A + B + Cu (2-2) u Here, H is the plate height or the height equivalent to a theoretical plate (HETP) and u is the linear velocity of the mobile phase in the column. A is multiple path effects (eddy diffusion). B is the longitudinal diffusion coefficient. C is mass transfer between the mobile and stationary phases. The eddy diffusion term A The peak broadening is due to the multitude of pathways. Molecule can find its special way through a packed column. Therefore, there are different the length of pathways on each molecule. The variable of residence time was used in the column for molecules of the same species which leads to a broadening zone. The longitudinal diffusion term B Diffusion of high concentration zone is a process in which species migration from a high concentration portion to low concentration region. The rate of migration is proportional to the concentration and the diffusion coefficient (D M ) of the molecule. The longitudinal diffusion is inversely proportion to the linear velocity of the mobile phase (u) because the molecule is in the column for a smaller period when the flow rate is high. Therefore, diffusion of molecule from the center of the zone to

11 11 the either side has less time to happen. The longitudinal diffusion is a source of zone broadening in GC because gaseous molecules of analytes diffuse at high rate. While the phenomenon of longitudinal diffusion in LC is smaller than because diffusion rate of liquid lower than gas mobile phase. The mass transfer term C Term C is related to the mass transfer of the solute between the mobile phase and stationary phase. The diffusion of solute between the mobile phase and stationary phase is long time process. Therefore, it will be carried along out of equilibrium. The mass transfer is proportion to the linear velocity of the mobile phase (u). The higher the velocity of mobile phase is source of band broadening for term C because non equilibrium of the mass transfer of the solute between the mobile phase and stationary phase The resolution The resolution R s gives a quantitative measure of the potentiality of the column to separate two analytes. The resolution factor can be expressed by Equation (2-3). R = 2 T R 2 T R(1) W 1 +W 2 (2-3) where T R(1) is the retention time of a analyte 1. T R(2) is the retention time of a analyte 2. W 1 is the magnitude of the base peak of analyte 1. W 2 is the magnitude of the base peak of analyte Monolith column Monolithic column is the most interesting in column technology of liquid chromatography. Monolith column is not filled with silica particles like conventional packed HPLC columns, but consist of a single rod of high purity monolith silica.

12 12 Monolith silica has a bimodal pore structure. First, Macropores dramatically reduce the column back-pressure and allow the use of faster flow rates. Second, Mesopores structure provides the very large active surface area for high efficiency separations. Figure 2-4 shows the picture of the typical porous structure of monolith columns. Figure 2-4 The mesoporous structure and the macropores or throughpores of monolith column (Chrom, 2001) 2.4 Benefits of monolith column The most extensively used monolithic silica columns include macropores with a size of about 2 µm and mesopores of about 13 nm in size. The total porosity of a column is about 85% which 15 20% higher collated to a typical column packed with 5 µm particles. The result of the high external porosity typical of monolithic silica columns is their greatly low column back pressure. Figure 2-5 shows the

13 13 column back pressure at different flow rates between a monolithic silica column and conventional columns packed with 3.5 and 5 µm particles (Cabrera, 2004, pp ). Obviously, the back pressure is lowest for the monolithic columns. So, monolith column can be operated at high flow rates leading to fast separations. Figure 2-5 Comparison of back pressure between the monolithic and conventional column at different flow rates (Cabrera, 2004). 2.5 The QuEChERS method The QuEChERS (Quick, Easy, Cheap, Effective, Rugged, and Safe) method for pesticide residue analysis in biological matrices was introduced in 2003 (Anastassiades, & Lehotay, 2003). The QuEChERS multiresidue procedure provides advantages over traditional methods. These include fast, inexpensive which requires few steps, and easy and reduces time taken for the extraction and clean up processes. The QuEChERS method provides high recovery for LC and GC, and high

14 14 reproducibility more than many typical sample preparations (Wang, Macneil, & Kay, 2012). In principle, the QuEChERS is based on two steps, extraction step and dispersive solid phase extraction (dspe). The extraction step involves salting out extraction (sodium chloride and magnesium sulphate) and equilibrium between aqueous and organic layer. The dispersive solid phase extraction involves clean up by using sorbents to remove interference. 2.6 The type of dspe for QuEChERS method Silica sorbent Silica sorbent is an unbonded sorbent (Figure 2-6) used to extract polar compound from non-polar compound. Retention of the analytes onto these dspe sorbent is due to the attractive forces by using hydrogen bonding interactions. Retention of an analyte under normal phase conditions is due to interaction between polar functional groups of the compound and polar groups on the sorbent surface. These include hydrogen bonding, pi-pi interactions, and dipole-dipole interactions. A compound adsorbed by these mechanisms is eluted by passing a more polar solvent that disrupts the binding mechanism. Si OH O Si OH O Si OH Figure 2-6 Structure of silica sorbent

15 Aminopropyl (NH 2 ) sorbent Aminopropyl sorbent can be used for normal phase separation. The aminopropyl sorbent material has an aliphatic aminopropyl group bonded to the silica surface (Figure 2-7). These sorbent can be used as a polar sorbent to extract polar compound from non-polar compound by using hydrogen bonding retention mechanism. Si O Si O Si OH O Si O Si (CH 2 ) 3 NH 2 (CH 2 ) 3 NH 2 Figure 2-7 Structure of aminopropyl sorbent Florisil sorbent Florisil sorbent is a selective synthetic magnesia-silica sorbent (Figure 2-8). These sorbent can be used typically for sample cleanup of organic extracts. Florisil sorbent is highly polar material with strongly adsorbs polar compounds from nonpolar matrices. Figure 2-8 Structure of florisil sorbent O Si O O Mg 2

16 Solid phase extraction (SPE) Solid phase extraction (SPE) is liquid-solid extraction. It was advanced commercially by the Waters Co. SPE is sold as Sep-Pak (Meloan, 1999). The packing material is held in a place within polypropylene column by porous frits and the column ends in a Luer tip to facilitate connection to a vacuum manifold, to a needle or to a collection vessel (Figure 2-9). Analytical laboratories use solid phase extraction to concentrate and purify samples for analysis. Solid phase extraction can be used to separate analytes of interest from a wide variety of matrices, including urine, blood, water, beverages, soil, and animal tissue. Figure 2-9 A Solid phase extraction column (Christie, 1992) The extraction protocol is easy. There are typically four steps in a reversephase SPE procedure 1) Conditioning 2) Loading 3) Washing 4) Elution.Figure 2-10 shows the step of extraction with SPE method.

17 17 Figure 2-10 SPE clean up procedure (Ferenc & Biziuk, 2005) 1) Conditioning The cartridge was conditioned to wet, activate the packing materials and remove any residue materials. 2) Loading The sample was added to the cartridge and passed the sorbent materials. The target of analyte and impurities were retained on the sorbents. 3) Washing The impurities were removed by weak or dilute solution except for the target analyte. 4) Elution The analyte can be eluted by washing the cartridge with a non-polar solvent, which disrupts the interaction of the analyte and the stationary phase.

18 The type of SPE in this work Oasis HLB Oasis HLB is a reversed phase sorbent for SPE. The Oasis HLB is a macroporous polymer consisting of two monomers, the lipophilic m-divinylbenzene and the hydrophilic N-vinylpyrrolidone (Figure 2-11). The advantage of a method using the polymeric sorbent instead of silica sorbent needs less sorbent because higher capacity. The Oasis HLB sorbent provides superior reversed phase capacity with a hydrophilic N-vinylpyrrolidone for improved retention of polar analytes. O N N O Figure 2-11 Structure of Oasis HLB sorbent Oasis MAX Oasis MAX is a strong mixed mode anion exchange. The Oasis MAX sorbents are derivatized with positively charged functional groups that interact and retain negatively charged anions. Anion exchange sorbents of Oasis MAX contain quaternary ammonium groups that have a permanent positive charge (Figure 2-12).

19 19 Oasis MAX is used to isolate strong anionic (very low pka, <1) or weak anionic (moderately low pka, >2) compounds. Oasis MAX is used to extract anions compound when recovery or elution of the anion compound is not desired. Anion compound can be isolated and eluted from Oasis MAX because they can be either displaced by an alternative anion or eluted with an acidic solution at a ph that neutralizes the weak anion. O N N C4 H 9 R Figure 2-12 Structure of Oasis MAX sorbent Oasis tc18 Oasis tc18 is strongly hydrophobic silica based reversed phase sorbent (Figure 2-13). Oasis tc 18 has overall similar retention capacity as C 18 sorbent but the trifunctional bonding on solid support is estimated to give tc 18 a better hydrophobic stability than C 18. The analyte of interest is typically moderately-to nonpolar. The hydrophilic silanol groups at the surface of the raw silica packing have been chemically modified with hydrophobic alkyl or aryl functional groups by reaction with the corresponding silanes. Retention of the analytes onto these SPE sorbent is due to the attractive forces between the carbon-hydrogen bonds in the analyte and the functional groups on the silica surface. These nonpolar-nonpolar attractive forces are

20 20 commonly called van der Waals forces, or dispersion forces. To elute an adsorbed compound from hydrophobic silica based reversed phase sorbent, use a nonpolar solvent to disrupt the forces that bind the analytes to the sorbent. HO O Si O Si OH O OH Si O Si O Si OH OH HO Si OH O Si O Si O Si O OH Si OH OH OH OH O Si O Si Figure 2-13 Structure of Oasis tc18 sorbent (Faria, Collins, & Jardim, 2009) 2.9 Literature reviews Delepine, Pessel & Sanders (1998, pp ) had reported their study on determination of six quinolones (enrofloxacin, ciprofloxacin, marbofloxacin, danofloxacin, sarafloxacin and difloxacin) in pig muscle by liquid chromatography. The analytes were extracted from muscle with phosphate buffer (ph 7.4). The extract was purified with C 18 solid phase extraction cartridge. Sample were analysed by LC. The column used was RP 18 column (125 x 4 mm id, particle size 5µm, LiChrospher, Merck). The mobile phase consisted of solution A (1% formic acid in 0.05 M ammonium acetate, ph 3) and solution B (acetonitrile). Analytes were detected by MS. The method is specific and reproducible and the confirmation of the six quinolones at the 7.5 µg/kg level in pig muscle.

21 21 Gigosos, Revesado, Cadahia, Fente, Vazquez & Franco (2000, pp 31-36) had reported their study on determination of five quinolones (marbofloxacin, ciprofloxacin, enrofloxacin, difloxacin and norfloxacin) in animal tissues and egg by HPLC with photodiode-array detection. The analytes were extracted from tissues with 1 M HCl. The extract was purified with C 18 cartridges. Sample were analysed by LC. The chromatographic separation was achieved on a C 18 Hypersil 5 µm BDS (250 x 4.6 mm) column. Detection was performed with a Model 996 diode array dectector. The mobile phase consisted of a mixture of 0.1 M ortho-phosphoric acid, ph 3.5- acetonitrile (85:15, v/v). The flow rate was 1 ml/min. The injection volume was 20 µl. The detection limit for each quinolone was:enrofloxacin and ciprofloxacin, 1 ng; norfloxacin and difloxacin, 2 ng; marbofloxacin, 4 ng. The intra-day RSD were lower than 7.9% and lower than 8.6% for inter-day. Maraschiello, Cusido, Abellan & Vilageliu (2001, pp ) had reported their study on determination of the fluoroquinolone ofloxacin in chicken kidney, liver, muscle and tissues by HPLC with UV detection. The analytes were extracted from tissues with 0.15 M HCl. The extract was purified with SPE cartridge coated with 60 mg of polymeric sorbent (Oasis; Waters, USA). The separation was accomplished on a LiChrospher 100C 18 HPLC (125 x 4 mm, particle size 5µm) column. The mobile phase consisted of a mixture of water-acn-tea (83:14:0.45, v/v, ph 2.30). The flow rate was 1 ml/min. The analysis of the tissue samples was performed in 12 min. The limits of quantification were obtained by 50µg/kg for muscle and tissues and 100µg/kg for liver and kidney tissues. The recovery values ranged from 80 to 100%. The limits of detection were established by 60µg/kg for liver and kidney tissues and 25µg/kg for muscle and fat plus skin tissues.

22 22 Johnston, Mackay & Croft (2002, pp ) had reported their study on determination of eight quinolones (oxolinic acid, flumequine, piromidic acid, enrofloxacin, ciprofloxacin, danofloxacin, sarafloxacin and orbifloxacin) in fish tissue and seafood by HPLC with MS detection. The analytes were extracted from muscle with acetonitrile. The extract was purified with ENVI Chrom P cartridges and AG MP-1 resin cartridges. The chromatographic separation was achieved on a Zorbax Extend C 18 (150 x 2.1 mm, particle size 5 µm) column. The mobile phase for gradient elution consisted of solvent A (2% formic acid) and solvent B (acetonitrile) and solvent C (Milli-Q water). The flow rate was 0.2 ml/min. The injection volume was 10 µl. All analytes eluted in less than 12 min. The limit of quantification was 5µg/kg for all analytes (10µg/kg for ciprofloxacin). The limit of detection was 1-3 µg/kg. Ramos, Aranda, Garcia, Reuvers & Hooghuis (2003, pp ) had reported their study on determination of the five quinolones (oxolinic acid, enrofloxacin, ciprofloxacin, flumequine and sarafloxacin) in fish and pork muscle by LC with fluorescence detection. The analytes were extracted from muscle with 0.05 M phosphate buffer at ph 7.4. The extract was purified with Discovery DSC-18 cartridges. The separation was accomplished on a Symmetry C 18 (150 x 4.6 mm, particle size 5µm) column. The mobile phase for ciprofloxacin, enrofloxacin and sarafloxacin consisted of a mixture of acetonitrile-0.02 M phosphate buffer ph 3.0 (18:82, v/v). The mobile phase for oxolinic acid and flumequine consisted of a mixture of acetonitrile-0.02 M phosphate buffer ph 3.0 (34:66, v/v). The flow rate was 1 ml/min. Detection limit was lower than 5 ng/g, except for sarafloxacin which had a detection limit higher than 10 ng/g. The standard curves showed a good linear correlation coefficient, r 2 higher than for all quinolones.

23 23 Hermo, Barron, & Barbosa (2005, pp 77-82) had reported their study on determination of quinolones in pig muscle by LC with UV detection. The analytes were extracted from pig muscle with m-phosphoric acid 0.3% and acetonitrile (75:25, v/v). The extract was purified with ENV+Isolute (200 mg, 3 ml) SPE cartridges. The chromatographic separation was achieved on a Zorbax Eclipse XDB-C 8 (150 x 4.6 mm) column. The mobile phase consisted of 10 mm citric acid and acetonitrile (88:12, v/v) at ph 4.5. The flow rate was 1.5 ml/min. All of the quinolones were eluted within 28 min. The recoveries of seven quinolones from pig muscle were % with RSD below 12%. The limits of quantification for each quinolones in pig muscle were between 17 and 39 µg/kg. Hoof, Wasch, Okerman, Reybroeck, Poelmans, Noppe & Brabander (2005, pp ) had reported their study on determination of eight quinolones in bovine muscle, milk and aquacultured products by LC with MS detection. The analytes were extracted from muscle with ultrapure water. The extract was purified with Isolute 500 mg C 18 SPE cartridge. The separation was accomplished on a Symmetry C 18 (150 x 2.1 mm, particle size 5 µm) column. The mobile phase consisted of a mixture of methanol with 0.1% trifluoroacetic acid (A) and water with 0.1% trifluoroacetic acid (B). The flow rate was 0.3 ml/min. The recoveries of eight quinolones from bovine muscle were % Bailac, Barron, & Barbosa (2006, pp ) had reported their study on determination of quinolones in poultry muscle by LC with UV detection. The analytes were extracted from poultry muscle with 0.3% m-phosphoric acid and acetonitrile (75:25, v/v). The extract was purified with ENV+ (200 mg, 3 ml) SPE cartridges. The chromatographic separation was achieved on a Zorbax Eclipse XDB-C 8 (150 x

24 mm) colomn. The mobile phase consisted of 10 mm citric acid and acetonitrile (88:12, v/v) ph 4.5. The flow rate was 1.5 ml/min. All of the quinolones were eluted within 25 min. The recoveries of quinolones from poultry muscle were 70-85% with RSD below 6%. The limits of quantification for each quinolones in poultry muscle were between 15 and 60 µg/kg. Hermo, Barron & Barbosa (2006, pp ) had reported their study on determination of eight quinolones in pig muscle by LC with UV detection. The analytes were extracted from muscle with m-phosphoric acid and acetonitrile. The extract was purified with ENV+isolute (200 mg, 3 cc) cartridge. The chromatographic separation was achieved on a Zorbax Eclipse XDB-C 8 (150 x 4.6 mm) column. The mobile phase consisted of 10 mm citric acid ph 4.5-acetonitrile (88:12, v/v). The flow rate was 1.5 ml/min. All of the quinolones were eluted within 28 min. The recoveries of eight quinolones from animal tissues were 81-99% with RSD below 12%. The limits of quantification for each quinolones in pig muscle were between 25 and 40 µg/kg. Hassouan, Ballesteros, Zafra, Vilchez & Navalon (2007, pp ) had reported their study on determination of quinolones in pig kidney samples by LC with fluorescence detection. The analytes were extracted from pig kidney with 0.3% m- phosphoric acid and acetonitrile mixture (75:25, v/v). The extract was purified with isolute ENV+ (200 mg, 3mL) SPE cartridges. The separation was accomplished on a Zorbax Eclipse XDB-C 8 (12.5 x 4.6 mm, particle size 5 µm) column. The mobile phase consisted of 10 mm citrate buffer ph 4.5 and acetonitrile (88:12, v/v). The flow rate was 1.5 ml/min. The injection volume was 20 µl. The limits of detection were

25 25 between 1 and 8 µg/kg. The limits of quantification were between 5 and 27 µg/kg. The recoveries of quinolones from pig kidney were 74-91% with RSD below 4%. Christodoulou, Samanidou & Papadoyannis (2007, pp ) had reported their study on determination of ten quinolones in chicken muscle and egg yolk by HPLC with photodiode array detector. The analytes were extracted from muscle with 0.1% TFA in methanol. The extract was purified with LiChrolut RP-18 (200 mg, 3 cc) SPE cartridges. The chromatographic separation was achieved on a PerfectSil Target ODS-3 (250 x 4 mm, particle size 5 µm) column. The mobile phase for gradient elution consisted of solvent A (0.1% TFA) and solvent B (acetonitrile) and solvent C (methanol). The flow rate was 1.2 ml/min. The injection volume was 50 µl. The limits of detection for each quinolones in chicken muscle were between 5.0 and 12.0 µg/kg. The recoveries of ten quinolones from animal tissues were % with RSD below 4.2%. Si-Jun, Cun, Hai-yang, Bing-Yu and Jian-Zhong (2007, pp ) had reported their study on determination of seven quinolones in animal muscle tissues by HPLC with fluorescence detection. The analytes were extracted from muscle with 0.1 M phosphate solution at ph 7.0. The extract was purified with HLB SPE cartridge (60 mg, 3 cc). The chromatographic separation was achieved on a Symmetry C 18 (250 x 4.6 mm, particle size 5µm) column from Waters Co. A gradient program in HPLC system was used with solvent A (0.02% formic acid, ph 2.8) and solvent B (acetonitrile). The injection volume was 100 µl. The limits of detection were µg/kg. The limits of quantification were µg/kg. The recoveries for each quinolone in chicken and pig muscles ranged from 70.4% to 105.8% with RSD below 9.3%.

26 26 Rubies, Vaquerizo, Centrich, Compano, Granados and Prat (2007, pp ) had reported their study on determination of quinolones in bovine muscle by LC with MS detection. The analytes were extracted from muscle with ultrapure water. The extract was purified with Oasis HLB 60 mg cartridges. The chromatographic separation was achieved on a X-Terra MS C 18 (50 x 2.1 mm, particle size 2.5 µm) column from Waters Co. The mobile phase with a gradient elution was used. The mobile phase consisted of solvent A (0.1% acetic acid) and solvent B (acetonitrile:water:acetic acid, 90:10:0.1, v/v/v). The flow rate was 0.3 ml/min. The linear calibration curves in the ranges and ng/g were obtained for all the analytes. Zhao, Jiang, Li, Mi & Shen (2007, pp ) had reported their study on determination of ten quinolones (marbofloxacin, ciprofloxacin, norfloxacin, lomefloxacin, danofloxacin, enrofloxacin, sarafloxacin, difloxacin, oxolinic acid and flumequine) in swine, chicken and shrimp muscle tissues by HPLC with fluorescence detection. The analytes were extracted from muscle with 0.01 M phosphate buffer at ph 7. The extract was purified with HLB SPE (60 mg, 3 cc) cartridge. The chromatographic separation was achieved on a Symmetry C 18 (250 x 4.6 mm, particle size 5 µm) column from Waters Co. The mobile phase for gradient elution consisted of solvent A (0.02% formic acid at ph 2.8) and solvent B (acetonitrile). The flow rate was 1 ml/min. The injection volume was 100 µl. All of the quinolones were eluted within 24 min. The recoveries of ten quinolones from animal tissues were % with RSD below 11.2%. The limits of quantification for each quinolone in difference animal muscle tissues were between 0.3 and 1.0 ng/g.

27 27 Li, Jiang, Zhao, Zhang, Ding & Shen (2008, pp ) had reported their study on determination of fluoroquinolones and sulfonamides in chicken muscle by LC with fluorescence and UV detection. The analytes were extracted from muscle with phosphate buffer at ph 6.0. The extract was purified with Oasis HLB SPE (60 mg, 3 cc) cartridge. The separation was accomplished on a SymmetryShild RP 18 (150 x 3.9 mm, particle size 5 µm) column. The mobile phase for gradient elution consisted of solvent A (0.1% formic acid) and solvent B (acetonitrile). The flow rate was 0.7 ml/min. The injection volume was 100 µl. The recoveries of ten quinolones from animal tissues were % with RSD below 9.3%. The limits of quantification for each fluoroquinolones in chicken muscle were between 0.2 and 4.0 ng/g and 15.0 ng/g for sulfonamides. Galarini, Fioroni, Angelucci, Tovo & Cristofani (2009, pp ) had reported their study on determination of eleven quinolones in animal feed by LC with fluorescence and untraviolet detection. The analytes were extracted from animal feed with the mixture of m-phosphoric acid, water and acetonitrile (0.45:30:70, w/v/v). The extract was purified with Oasis HLB (60 mg, 3 ml) SPE cartridges. The chromatographic separation was achieved on a GEMINI C 18 (250 x 3 mm, particle size 5 µm) column. The mobile phase consisted of solvent A (acetonitrile) and solvent B (o-phosphoric acid 25 mm ph 3). The flow rate was 0.65 ml/min. The injection volume was 20 µl. The limits of detection for each quinolone were in the range 0.04 to 0.8 mg/kg. The limits of quantification for each quinolone were in the range 0.1 to 2.4 mg/kg. The recoveries were from 69 to 98% with RSD below 10%. Herrera, Borges, Delgado, Herrero & Cifuentes (2011, pp ) had reported their study on determination of quinolones in infant and young children

28 28 powdered milk by UPLC with MS detection. The analytes were extracted from infant and young children powdered milk with the mixture of acetonitrile containing 20% TCA. The extract was purified with Oasis HLB (500 mg, 6 ml) SPE cartridges. The chromatographic separation was achieved on a Hypersil Gold C 18 (50 x 2.1 mm, particle size 1.9 µm) column. The mobile phase consisted of solvent A (0.1% formic acid) and solvent B (methanol). The flow rate was 0.3 ml/min. The injection volume was 10 µl. The procedure was developed providing recovery higher than 84% (RSDs lower than 13%) for all analytes. The limits of detection were between 0.04 and 0.52 µg/kg. Yu, Tao, Chen, Pan, Liu, Wang, Huang, Dai, Peng Wang & Yuan (2012, pp ) had reported their study on determination of fluoroquinolones in food of animal origin by HPLC with UV detector. The analytes were extracted from food of animal origin with acetonitrile. The extract was purified with Oasis HLB (60 mg, 3 cc) SPE cartridges. The chromatographic separation was achieved on a Zorbax SB- Aq-C 18 (250 x 4.6 mm, particle size 5 µm) colomn. The mobile phase consisted of methanol/acetonitrile/0.02 M citric acid and 0.03 M ammonium acetate. The flow rate was 0.8 ml/min. The injection volume was 50 µl. The recoveries of quinolones spiked in the tissues as muscle, liver, kidney of swine, bovine, chicken and fish were found between 70.6% and 111.1% with RSD below 15%. The LOD and LOQ of the 15 quinolones were between 3 µg/kg and 10 µg/kg, respectively Yan-Xia, Wei, Xue-Lian & Ming (2012, pp ) had reported their study on determination of fourteen veterinary antibiotics in animal feces by HPLC with photodiode array detection. The analytes were extracted from animal feces with a mixture of methanol and citrate buffer (1:1, v/v). The extract was purified with

29 29 Oasis HLB (200 mg, 6 ml) SPE cartridges. The chromatographic separation was achieved on a inertsil ODS-3 (250 x 4.6 mm, particle size 5 µm) column. The mobile phase A and B were acetonitrile and 1% of acetic acid, respectively. The flow rate was 0.8 ml/min. The LOD and LOQ of the veterinary antibiotics were between 9 and 90 µg/kg and 33 and 297 µg/kg, respectively. The recoveries of veterinary antibiotics from animal feces were 58-93% with RSD below 12.5%. Evaggelopoulou & Samanidou (2013, pp ) had reported their study on determination of seven quinolones in salmon tissue by HPLC with photodiode array detector. The analytes were extracted from salmon tissue with citrate buffer at ph 4.7. The extract was purified with Oasis HLB (200 mg, 6 ml) SPE cartridges. The chromatographic separation was achieved on a Perfectsil ODS (250 x 4 mm, particle size 5 µm) column. The mobile phase consisted of 0.1% TFA and acetonitrile and methanol. The flow rate was 1.0 ml/min. The injection volume was 20 µl. The recoveries of seven quinolones from salmon tissue were % with RSD below 9.4%.

30 30 CHAPTER 3 RESEARCH METHODOLOGY 3.1 Equipment and apparatus 1. Liquid chromatography coupled with UV detector. (HP 1050 series, Hewlett Packard, Germany). 2. Chromolith column RP-18e (Merck, Germany) 3. Analytical balance (AG 204, Mettler Toledo, Switzerland) 4. Ultrasonic bath (690D, Crest, Germany) 5. Micropipette µl (Oxford Benchmate, Japan) 6. Micropipette µl (Socorex, Switzerland) 7. Syringe filter 0.45 µm (Hewlett-Packard, USA) 8. ph meter (Beckman, USA) 9. Magnetic stirrer (C-MAG HS7, IKA, Malaysia) 10.Vortex (VELP Scientifica, Italy) 11. Sep-Pak tc18 60 mg 3 cc (Oasis, Waters, USA) 12. Sep-Pak HLB 60 mg 3 cc (Oasis, Waters, USA) 13. Sep-Pak MAX 60 mg 3 cc (Oasis, Waters, USA) 3.2 Chemicals 1. Marbofloxacin; C 17 H 19 FN 4 O 4, FW (AR Grade Sigma, Germany) 2. Ciprofloxacin; C 17 H 18 FN 3 O 3, FW (AR Grade Fluka, Germany) 3. Enrofloxacin; C 19 H 22 FN 3 O 3, FW (AR Grade Fluka, Germany)

31 31 4. Difloxacin hydrochloride; C 21 H 19 F 2 N 3 O 3.HCl, FW (AR Grade Sigma, Germany) 5. Sarafloxacin hydrochloride trihydrate; C 20 H 17 F 2 N 3 O 3.HCl.3H 2 O, FW (AR Grade Sigma, Germany) 6. Lomefloxacin hydrochloride; C 17 H 19 F 2 N 3 O 3.HCl, FW (AR Grade Sigma, Germany) 7. Methanol; CH 3 OH, FW (AR Grade QREC, New Zealand) 8. Acetonitrile; CH 3 CN, FW (AR Grade QREC, New Zealand) 9. Sodium hydroxide; NaOH, FW (AR Grade APS Ajax finechem, Australia) 10.Di-Ammonium hydrogen citrate; C 6 H 14 N 2 O 7, FW (AR Grade Carlo Erba, Italy) 11.Orthophosphoric acid (85%); H 3 PO 4, FW (AR Grade QREC, New Zealand) 12.Potassium dihydrogen orthophosphate; KH 2 PO 4, FW (AR Grade Merck, Germany) 13.Sodium acetate hydrated; CH 3 COONa.3H 2 O, FW (AR Grade Ajax finechem, Australia) 14.Acetic acid (100%) ; CH 3 COOH, FW (AR Grade Merck, Germany) 15.Trichloroacetic acid; C 2 HCl 3 O 2, FW (AR Grade Merck, Germany) 16.Formic acid (98%); CHOOH, FW (AR Grade Merck, Germany)

32 Experimental Preparation of 0.01 M sodium hydroxide in methanol 0.01 M sodium hydroxide in methanol was prepared by pipetting 93 µl of 10.8 M sodium hydroxide stock solution into a 100 ml of volumetric flask. The volume was adjusted to 100 ml with methanol Preparation of standard solutions 1) 10 ml of 200 mg/l quinolone stock solutions Individual stock solutions were prepared by dissolving 2.0 mg of each quinolones into 10 ml of 0.01 M sodium hydroxide in methanol. The stock solutions were stored in a refrigerator at 4 o C. 2) 5 ml of 5 mg/l quinolone standard solutions Individual standard solutions were prepared by pipetting 125 µl of each quinolones stock into a 5 ml of volumetric flask. The volume was adjusted to 5 ml with deionized water. The standard solutions were stored in a refrigerator at 4 o C. 3) 5 ml of working standard mixture The mixture of working solution was diluted from Individual standard solutions. The volume was adjusted to 5 ml with deionized water Preparation of 10 mm citrate buffer 10 mm citrate buffer was prepared by dissolving g of diammonium hydrogen citrate in 250 ml of deionized water. The ph of citrate buffer was adjusted to the desired ph by using citric acid.

33 Preparation of 10 mm acetate buffer 10 mm acetate buffer was prepared by dissolving g of sodium acetate in 250 ml of deionized water. The ph of phosphate buffer was adjusted to the desired ph with using acetic acid Preparation of 10 mm formate buffer 10 mm formate buffer was prepared by pipetting 500 µl of 5 M formic acid into a 250 ml of volumetric flask. The volume was adjusted to 250 ml with deionized water. The ph of formate buffer was adjusted to the desired ph with using sodium hydroxide Preparation of 30 mm phosphate buffer 30 mm phosphate buffer was prepared by dissolving g of potassium dihydrogen orthophosphate in 250 ml of deionized water. The ph of phosphate buffer was adjusted to 7.0 by using phosphoric acid Preparation of home-made SPE syringes First, the cotton was placed at the bottom of the syringe (3 cc). Second approximately 60 mg of the sorbent (dry weight) was mixed with methanol to form slurry and transferred to the syringe. Finally, the cotton was placed at the top of the syringe to hold the sorbent material beads inside the syringe.

34 Preparation of samples Six different chicken muscle samples were obtained from local supermarkets in Chon Buri province. Chicken muscle samples were finely chopped and homogenized by using kitchen blender. Samples were stored at -20 C until analysis Sample extraction by QuEChERS method Figure 3-1 shows the step of extraction with QuEChERS method for this work ± 0.05 g of chicken muscle was weighed and placed in a 50 ml polypropylene centrifuge tube and spiked with the standard solution. The samples were kept for 15 min in the dark at room temperature. Extraction was carried out using 8 ml of 30 mm phosphate buffer solution at ph 7.0. The samples were mixed by vortex for 1 min. After that, 10 ml of 5% formic acid in acetonitrile was added and the tubes were shaken about 1 min by vortex. Then, 4 g magnesium sulphate anhydrous and 1 g sodium chloride were added and shaken with vortex for 1 min, and centrifuged at 3500 rpm for 5 min. For clean up step, 1 ml of upper acetonitrile extract was transferred in to the dispersive SPE tubes containing 250 mg of sorbent and mixed by vortex for 1 min and centrifuged at 3500 rpm for 5 min. Finally, acetonitrile supernatant extract (500 µl) was transferred to vial and evaporated to dryness under stream of nitrogen at temperature 65 o C. The residue was dissolved in 500 µl mobile phase for injection into the HPLC.

35 35 Figure 3-1 Diagram of extraction with QuEChERS method Sample extraction by SPE method 2 g of chicken muscle was weighed and placed in a 50 ml polypropylene centrifuge tube and spiked with the standard solution. The samples were kept for 15 min in the dark at room temperature. After that, 8 ml of 30 mm phosphate buffer solution at ph 7.0 was added to the sample. The samples were mixed by vortex about 1 min, and centrifuged at 3500 rpm for 10 min. The supernatant was collected. The sample was extracted once more and combined the supernatant. The SPE cartridge was conditioned with 2 ml methanol and 4 ml water. The extract was percolated through the cartridge. The cartridge was washed with 2 ml of methanol/water (1:4, v/v). The analytes were eluted with 3 ml of 2% acetic acid in methanol. The eluent was evaporated to dryness at 65 C under a nitrogen stream and reconstituted in 500

36 36 µl of mobile phase. The sample was filtered through a 0.45 µm of Syringe filter and injected into HPLC Condition for separation of quinolones by HPLC Chromatographic separation by using isocratic elution was performed on a monolith column. The mobile phase consisted of acetonitrile and 10 mm ammonium citrate (ph 3.7) in the ratio 17:83 (v/v). The flow rate at 1 ml/min was used in all experiments. The injection volume was 20 µl. Detection by using timetable program with UV detector at 296 nm for MAR, 277 nm for CIP, LOM and ENR, 280 nm for SAR, and DIF was chosen Optimization of HPLC condition Separation of six quinolones was carried out by using HPLC. The parameters affecting of the separation were studied. 1) Effect of ph The effect of ph on analytical of quinolones was studied. The concentration of citrate buffer was used at 10 mm. The ph of citrate buffer was adjusted from range 3.2 to 4.5. The other conditions of HPLC were mentioned in ) Effect of percentage acetonitrile The effect of percentage acetonitrile on analytical of quinolones was studied by varying between 15 and 18% in mobile phase (acetonitrile and 10 mm citrate buffer ph 3.7). The other conditions of HPLC were mentioned in

37 37 3) Types of buffer Three type of buffer in mobile phase were studied (citrate, acetate and formate buffers). The concentration of buffer was used at 10 mm of each buffer and separations were analyzed under the optimized percentage acetonitrile and ph of buffer. The other conditions of HPLC were mentioned in Optimization of QuEChERS method 1) Effect of types of sorbents The effect of types of sorbent on recoveries of quinolones was studied by using three types of sorbent (silica, florisil, and aminopropyl (NH 2 ) sorbents). The steps of QuEChERS procedure were addressed in ) Effect of amount of sorbent The effect of amount of sorbent on recoveries of quinolones was studied by varying amount of sorbent between 200 and 300 mg under the optimized sorbent types. The steps of QuEChERS procedure were addressed in ) Effect of volume of extraction solvent The effect of volume of extraction solvent on recoveries of quinolones was studied by varying volume between 3 and 10 ml under the optimized sorbent types and amount of sorbent. The steps of QuEChERS procedure were addressed in

38 Optimization of SPE method 1) Effect of SPE types The effect of sorbent types on recoveries of quinolones was studied by using three types of SPE (Oasis HLB, Oasis tc18, and Oasis MAX). 60 mg of each sorbents were studied. The steps of SPE procedure were mentioned in ) Effect of amount of acid in the eluant The effect of amount of acid on recoveries of quinolones was studied by varying amount of acid between 1 and 4% acid under the optimized SPE types. The steps of SPE procedure were mentioned in ) Effect of types of acids in eluent The effect of types of acids in eluent on recoveries of quinolones was studied by varying three types of acids (acetic acid, formic acid, and propionic acid) under the optimized SPE types and amount of acid in the eluent. The steps of SPE procedure were mentioned in Method validation The analytical characteristics of the proposed method for quinolones were investigated under the Food and Drug Administration (FDA) guideline for bioanalytical assay validation (U.S. Food and Drug Administration, 2012). The parameters (limit of detection (LOD), limit of quantification (LOQ), matrix matched calibration curve, precision, and accuracy) were studied. 1) Limit of detection (LOD) The lowest concentration of analyte can be detected with acceptable certainty but can t be quantified with acceptable precision. The limit of detection

39 39 based on the signal-to-noise ratio (S/N) of 3 is operated by comparing ratio between signals of analyte in the samples and signals of noise in the samples. 2) Limit of quantification (LOQ) The lowest concentration of analyte can be quantified with acceptable accuracy and precision. The limit of quantification was calculated with a signal-to-noise ratio (S/N) of 10 is operated by comparing ratio between signals of analyte in the samples and signals of noise in the samples. 3) Matrix matched calibration curve The five points of matrix matched calibration curve were obtained by plotting between peak areas and the concentration of quinolones in chicken muscle samples. The calibration curve was prepared by spiking blank chicken muscle samples with five concentration levels of standard quinolones. Each calibration level was analyzed in triplicate. 4) Precision The precision of the proposed method was evaluated in terms of percentage relative standard deviations (%RSD). For intra-day precision was determined by spiking three concentration levels of standard quinolones in chicken muscle samples. Five replicated of each level were analyzed under the same experimental conditions on the same day. For inter-day precision was determined by the same procedure on three days. 5) Accuracy The accuracy of the proposed method was evaluated in terms of recovery. The recoveries were studied by spiking three concentration levels of standard quinolones in chicken muscle samples and five replicates of each sample

40 40 were analyzed under the same experimental conditions. The percentage recovery can be calculated by the formula (3-1). %Recovery = C spiked C unspiked C added 100 (3-1) Where C spiked, C unspiked and C added are the concentration of quinolones measured in spiked sample, concentration of quinolones in unspiked sample and concentration of added quinolones, respectively Determination of quinolones in sample Chicken muscle samples were prepared under the optimum condition in three replicates. The samples were analyzed with HPLC under the optimum experimental conditions. The amounts of quinolones were calculated by using a matrix matched calibration curve. The average results were reported.

41 41 CHAPTER 4 RESULTS 4.1 Optimization of HPLC condition ph Effect The effect of buffer ph was studied to obtain an optimum buffer ph for subsequent analysis. Using a constant concentration of buffer (10 mm ammonium citrate) and 17% acetonitrile in the mobile phase, and varying ph of citrate buffer, the effect of buffer ph on chromatographic separation in terms of retention time, peak height, and resolution was investigated. Figure 4-1 shows the effect of buffer ph on the chromatographic separation for six quinolones. It can be seen that buffer ph had a little effect on the retention time of MAR, CIP, LOM, and SAR peaks with increasing buffer ph, but the retention time of ENR and DIF increased when buffer ph was increased. The peak height of all analytes slightly decreased when the buffer ph increased. In terms of resolution, it is apparent that resolution increased for LOM and ENR peaks, and SAR and DIF peaks, as the buffer ph was increased. However, buffer ph has minute effect on the resolution for MAR and CIP peaks, CIP and LOM peaks, and ENR and SAR peaks. Moreover, it was found that when the buffer ph was 3.7 or higher, the resolution of LOM and ENR peaks, and SAR and DIF peaks exceeded 1.5. Thus, to make a compromise between resolution and peak height, an optimum buffer ph was chosen to be 3.7.

42 42 Figure 4-1 Effect of buffer ph on chromatographic separation of six quinolones, the chromatographic separations were a citrate buffer/acetonitrile (83/17, v/v) maintained in an isocratic mode. The concentrations of standards were 0.1 mg/l for MAR and CIP, 0.2 mg/l for LOM, ENR, SAR, and DIF. For other chromatographic conditions see research methodology The percentage of organic solvent effect The effect of organic solvent was studied to obtain an optimum percentage of acetonitrile for subsequent analysis. Using a constant concentration and ph of buffer (10 mm ammonium citrate at ph 3.7), and varying percentage of acetonitrile in mobile phase, the effect of percentage of acetonitrile on chromatographic separation in terms of retention time, peak height, and resolution was investigated. Figure 4-2 shows the effect of percentage of acetonitrile on the chromatographic separation for six quinolones. It can be seen that percentage of acetonitrile had effect on the retention time of MAR, CIP, LOM, ENR, SAR and DIF

43 43 peaks with increasing percentage of acetonitrile. The retention time of MAR, CIP, LOM, ENR, SAR and DIF decreased when percentage of acetonitrile was increased. The peak height of all analytes increased when the percentage of acetonitrile increased. In terms of resolution, it is apparent that resolution decreased for all analytes peaks, as the percentage of acetonitrile was increased. However, percentage of acetonitrile has significant effect on the resolution for CIP and LOM peaks and SAR and DIF peaks. Moreover, it was found that when the percentage of acetonitrile was 18, the resolution of CIP and LOM peaks and SAR and DIF peaks were less than 1.5. In addition, at percentage of acetonitrile of 17, all analytes were completely separated with resolution more than 1.5.Thus, to make a compromise between resolution and peak height, optimum percentage of acetonitrile was chosen to be 17. Figure 4-2 Effect of organic solvent on chromatographic separation of six quinolones, the chromatographic separations were a 10 mm ammonium citrate buffer (ph 3.7)/acetonitrile maintained in an isocratic mode. The concentrations of standards were 0.1 mg/l for MAR and CIP, 0.2 mg/l for LOM, ENR, SAR, and DIF. For other chromatographic conditions see research methodology.

44 Effect of buffer type The effect of buffers was studied to obtain the optimum types of buffers (acetate, formate and citrate) at the same ph and concentration for subsequent analysis. Using a constant concentration and ph of buffer (10 mm at ph 3.7) and 17% acetonitrile in the mobile phase, and different types of buffers, the effect of types of buffers on chromatographic separation in terms of retention time, peak height, and resolution was investigated. Figure 4-3 shows the effect of types of buffers on the chromatographic separation for six quinolones. It can be seen that types of buffers had effect on the retention times and resolution of all analytes with varying types of buffers. In terms of retention times, In the case of citrate buffer, the retention times of all analytes are more than formate and acetate buffer. Meanwhile, the retention times of all analytes by using formate buffer are more than acetate buffer. In terms of resolution, it is apparent that citrate buffer provided highest resolution between SAR and DIF peak with resolution Moreover, it was found that when the types of buffers were acetate or formate, the resolution of SAR and DIF peaks less than 1.5. Consequently, to make a compromise between resolution and peak height, citrate buffer was selected to be an optimum buffer.

45 45 Figure 4-3 Effect of types of buffers on chromatographic separation of six quinolones, the chromatographic separations were a buffer (ph 3.7)/acetonitrile (83/17, v/v) maintained in an isocratic mode. The concentrations of standards were 0.1 mg/l for MAR and CIP, 0.2 mg/l for LOM, ENR, SAR, and DIF. For other chromatographic conditions see research methodology. In this study, the mobile phase consisting of acetonitrile and 10 mm citrate buffer at ph 3.7 (17:83, v/v) with isocratic elution provided the best chromatographic separation, and was chosen to be the optimum condition. Under these optimum conditions, the separation of the six quinolones was achieved within less than 7 min (Figure 4-4).

46 46 Figure 4-4 Chromatogram of quinolones standard solution under the optimum condition, the chromatographic separations were a citrate buffer (ph 3.7) /acetonitrile (83/17, v/v) maintained in an isocratic mode. The concentrations of standards were 0.1 mg/l for MAR and CIP, 0.2 mg/l for LOM, ENR, SAR, and DIF. For other chromatographic conditions see research methodology. 4.2 Optimization of QuEChERS method To optimize the QuEChERS method, three study parameters such as sorbent type, sorbent amount, and solvent volume were investigated Sorbent type Three types of sorbents including aminopropyl, florisil, and silica were used in this work to study the effect of sorbent type on extraction of the analytes in terms of recoveries. To evaluate the influence of sorbent type on the recoveries of quinolones, the sorbent amount of 250 mg, and ml of extraction solvent (5% formic acid in acetonitrile) were utilized in all experiments. Figure 4-5 shows effect of sorbent type on chromatograms obtained from analysis of a chicken muscle sample without spiked (a), and spiked at 0.1 mg/l

47 47 of MAR and CIP, 0.2 mg/l of LOM, ENR, SAR, and DIF (b). In Figure 4.5 (a), it can be seen that, during analysis time between 2 to 8 min, silica minimized matrix effect better than florisil and aminopropyl. Thus, in quinolone analysis, silica sorbent was chosen for further experiments. Figure 4-6 illustrates the recoveries of quinolone analysis by using silica sorbent. The recoveries of MAR, CIP, LOM, ENR, SAR and DIF were between 81% and 95%, and relative standard deviations were less than 2.5% except LOM (5.1%). Consequently, silica sorbent was selected for clean-up process for QuEChERS method in this experiment. (a) (b) Figure 4-5 Effect of type of sorbent on chromatograms obtained from analysis of a chicken muscle sample with QuEChERS method without spiked standard (a), and spiked at 0.1 mg/l of MAR and CIP, 0.2 mg/l of LOM, ENR, SAR, and DIF (b) analysis of quinolones with QuEChERS method. The chromatographic separations were a citrate buffer (ph 3.7)/acetonitrile (83/17, v/v) maintained in an isocratic mode. For other chromatographic conditions see research methodology.

48 recovery(%) MAR CIP LOM ENR SAR DIF Figure 4-6 Recoveries of quinolone analysis by using silica sorbent Sorbent amount The effect of sorbent amount was studied to obtain the optimum amount of sorbent for subsequent analysis. For all experiments in this section, ml of extraction solvent (5% formic acid in acetonitrile) was used. To investigate the effect of sorbent amount on the recovery of the anlytes, amount of silica sorbent (200 mg, 250 mg and 300 mg) used in clean-up step was investigated. The influence of silica amount on the recovery of analytes is shown in Figure 4-7. It was found that the recoveries increased when silica amount increased from 200 mg to 250 mg for all analytes excluding SAR. In addition, virtually all recoveries decreased with changing 250 mg to 300 mg of silica sorbent. Recoveries of analytes by using 200 mg, 250 mg, and 300 mg silica sorbent were in range 81-99%, %, and 75-99%, respectively. From these results, 200 mg of silica sorbent was selected to be an optimum amount for using in clean-up process.

49 %recovery 49 silica 200 mg silica 250 mg silica 300 mg MAR CIP LOM ENR SAR DIF Figure 4-7 Effect of silica amount on the recoveries of quinolones Extractant volume The effect of extractant volume was studied to obtain the optimum extractant volume for subsequent analysis. In this section, silica amount was fixed at 250 mg, and 3.00 ml and ml of extractant (5% formic acid in acetonitrile) was investigated. Figure 4-8 shows the effect of extractant volume on recoveries of six quinolones. It was found that recoveries of all analytes significantly increased when 3.00 ml of extractant was changed to ml. Recoveries obtained from 3.00 ml and ml of extractant were between 43% and 71%, and 82% and 104%, respectively. Consequently, ml of extractant was applied to extract quinolones in chicken muscle sample.

50 recovery (%) MAR CIP LOM ENR SAR DIF ml of extractant 3 ml of extractant Figure 4-8 Effect of volume of extractant on the recoveries of quinolones. 4.3 Optimization of SPE method Solid phase extraction or SPE method is a sample preparation. It was selected to be studied in this work in order to compare with QuEChERS method mentioned in previous section. To obtain an optimum SPE method for clean-up and preconcentration steps, three study parameters were investigated. These included SPE type, as well as acid amount and type in eluant SPE type Sorbent type in SPE is an importance to clean-up and preconcentration processes. Three differences in sorbent types (Oasis HLB, Oasis tc18, and Oasis MAX) were used and investigated. To investigate the effect of SPE type on recovery, 60 mg of each sorbents was used. The eluant used all experiments in this section was 2% acetic acid in methanol. Figure 4-9 shows the effect of SPE types on the recoveries of six quinolones. It revealed that HLB sorbent provided the highest recoveries, and MAX

51 recovery (%) 51 sorbents gave the lowest recoveries. The recoveries obtained from HLB, tc18, and MAX sorbents were in range 82% and 105%, 37% and 86%, and 26% and 77%, respectively. For HLB sorbent, % recoveries of MAR, CIP, LOM, ENR, SAR, and DIF were 105, 102, 97, 102, 82, and 84, respectively. Therefore, from these results, HLB was selected for further experiment MAR CIP LOM ENR SAR DIF 0.00 HLB MAX tc18 Figure 4-9 Effect of different type of SPE on the recoveries of quinolones Acid amount in eluant The effect of acid amount in the eluant (methanol) was studied to obtain the optimum amount of acid in eluant for subsequent analysis. Acetic acid was chosen because this acid was commonly used. To investigate acetic acid amount in eluant affecting recovery, percentage of acetic acid in eluant was varied from 1% to 4%. Effect of percentage of acetic acid in eluant on recovery of quinolones is illustrated in Figure From the Figure, tendency of recoveries of all quinolones increased with

52 recovery (%) 52 changing 1% to 2% acetic acid. From 2% to 4% acetic acid, almost recoveries of all analytes slightly decreased, but some exceptions were observed in SAR and DIF. For SAR, recovery decreased from 2% to 3% acetic acid, and increased from 3% to 4%. For DIF, recovery levelled off between 2% and 3%, and increased between 3% and 4%. In addition, it revealed that virtually all recoveries of quinolones obtained from 2% acetic acid in eluant provided gave satisfactory results with the recoveries between 84% and 101%, and relative standard deviations between 0.6% and 2.3%. Consequently, 2% of acetic acid in methanol was chosen as the optimum eluant %acetic 2%acetic 3%acetic 4%acetic MAR CIP LOM ENR SAR DIF Figure 4-10 Effect of acetic acid in the eluant on the recoveries of quinolones Acid type in the eluant The effect of acid type was studied to obtain the suitable acid for subsequent analysis. Three differences in acid types, namely acetic acid, formic acid, and propionic acid were selected and investigated. To investigate the effect of acid

53 recovery (%) 53 type on recovery of quinolones, 2% of each acid in methanol, and HLB sorbent were employed. The influence of acid type on the recovery of analytes is shown in Figure The recoveries of quinolones obtained from acetic acid, formic acid, and propionic acid were in range 86% to 98%, 76% to 98%, and 58% to 90%, respectively. It can be seen that recoveries of most analytes with the exception of DIF obtaining from acetic acid were more than recoveries obtaining from formic and propionic acids. For acetic acid, % recoveries of MAR, CIP, LOM, ENR, SAR, and DIF were 91, 98, 93, 95, 90, and 86, respectively. Therefore, based on these results, it can be concluded that acetic acid was a suitable acid, and chosen for further experiment acetic formic propionic MAR CIP LOM ENR SAR DIF Figure 4-11 Effect of acid type in the eluant on the recoveries of quinolones.

54 Validation of the method Validation of the method was operated according to the US Food and Drugs Administration (FDA) guideline for bioanalytical assay procedure Limit of detection (LOD) and limit of quantification (LOQ) LOD values were calculated by using a signal to noise ratio of 3, while LOQ values were calculated by using a signal to noise ratio of 10. The values of LOD and LOQ of quinolones are shown in Table 4-1. According to Table 4-1, the LOD and LOQ values of quinolones in chicken muscles were ranged µg/kg and µg/kg respectively. As can be seen, all LOD and LOQ values were lower than the MRLs for these drugs in the European Union in the Council Regulation 2377/90. Table 4-1 LOD and LOQ values of quinolones in chicken muscles obtained from this work Analytes MAR CIP LOM ENR SAR DIF LOD (µg/kg) LOQ (µg/kg)

55 Calibration curves of quinolones Calibration curves of each quinolone were constructed by using matrix matched calibration curve. Five points of matrix matched calibration curves were received by spiking 2 g of blank chicken muscles sample with standard quinolones at concentration between 23 µg/kg and 50 µg/kg for MAR and ENR (Figure 4-12, 4-13 respectively), 40 µg/kg and 70 µg/kg for DIF (Figure 4-14), 13 µg/kg and 50 µg/kg for CIP (Figure 4-15), 30 µg/kg and 70 µg/kg for SAR (Figure 4-16), and 24 µg/kg and 50 µg/kg for LOM (Figure 4-17). The calibration curves were carried out by plotting the peak area of analytes versus the concentrations. The obtained results are indicated in Table 4-2. The correlation coefficients (R 2 ) of calibration curves were obtained in range of for six quinolones. Table 4-2 Concentration range and correlation coefficients (R 2 ) of calibration curves of quinolones Analytes Concentration range (µg/kg) R 2 MAR CIP LOM ENR SAR DIF

56 Peak area Peak area y = x R² = Concentration of MAR (µg/kg) Figure 4-12 Calibration curve of MAR (n=3) y = x R² = Concentration of ENR (µg/kg) Figure 4-13 Calibration curve of ENR (n=3)

57 Peak area Peak area y = x R² = Concentration of DIF (µg/kg) Figure 4-14 Calibration curve of DIF (n=3) y = x R² = Concentration of CIP (µg/kg) Figure 4-15 Calibration curve of CIP (n=3)

58 Peak area Peak area y = x R² = Concentration of SAR (µg/kg) Figure 4-16 Calibration curve of SAR (n=3) y = x R² = Concentration of LOM (µg/kg) Figure 4-17 Calibration curve of LOM (n=3) Accuracy and precision The accuracy of the proposed method was evaluated in terms of recovery and the precision (intra-day and inter-day) was assessed in terms of percentage

59 59 relative standard deviation (%RSD). The chicken muscles were spiked with the different analyte concentrations at three levels. Level 1 was 27 µg/kg for MAR, CIP, LOM, and ENR, 42 µg/kg for SAR and DIF. Level 2 was 35 µg/kg for MAR, CIP, LOM, and ENR, 55 µg/kg for SAR and DIF. Level 3 was 45 µg/kg for MAR, CIP, LOM, and ENR, 65 µg/kg for SAR and DIF. Five replicated samples of each level were prepared and analyzed under the same experimental conditions on the same day for intra-day precision. The procedure was repeated on three different days to determine inter-day precision. The recoveries were in range 83.6 to 102.5% at all spiked levels of each quinolones (Table 4-3). %RSD were in range 1.1 to 6.8% and 2.2 to 8.6% for intra-day and inter-day precision respectively (Table 4-4). Table 4-3 Recoveries of quinolones from chicken muscles spiked at three different concentration levels. Analytes Recovery (%) (n=5) Level 1 Level 2 Level 3 MAR CIP LOM ENR SAR DIF

60 60 Table 4-4 Precision of the proposed method from chicken muscles spiked at three different concentration levels. Analytes Intra-day (n=5, %RSD) Inter-day (n=15, %RSD) Level 1 Level 2 Level 3 Level 1 Level 2 Level 3 MAR CIP LOM ENR SAR DIF Analysis of quinolones in chicken muscle samples Chicken muscle samples were obtained from local supermarket in the area of Chon Buri (Thailand). The samples were extracted and cleaned-up by SPE method as mentioned previously. The obtained results are indicated in Table 4-5. It can be seen that MAR and SAR were not found in all samples. For CIP, it was found lower than LOQs for sample A, B, C, E and F, and not found in sample D. For LOM, it was found in sample C at 40 µg/kg and found in sample D, E, and F with concentration lower than LOQs and not found in sample A and B. For ENR, it was found in sample C only at concentration lower than LOQs. For DIF, it was found in sample A, B, D, E, and F at concentration lower than LOQs and not found in sample C.

61 61 Table 4-5 Quinolones levels found in chicken muscle samples from local supermarket in the area of Chon Buri (µg/kg). Samples MAR CIP LOM ENR SAR DIF A - <LOQs <LOQs B - <LOQs <LOQs C - <LOQs 40 <LOQs - - D - - <LOQs - - <LOQs E - <LOQs <LOQs - - <LOQs F - <LOQs <LOQs - - <LOQs Figure 4-18 shows the example of chromatograms of (a) blank chicken muscle sample, (b) blank chicken muscle sample spiked with six quinolones at 27 µg/kg for MAR, CIP, LOM, and ENR, and 42 µg/kg for SAR and DIF. According to Figure 4-18, peaks of MAR, CIP, LOM, ENR, SAR, and DIF were not interfered with matrix interferences from chicken muscle sample.

62 62 a b Figure 4-18 Chromatograms of chicken muscle samples (a) blank chicken muscle sample, (b) blank chicken muscle sample spiked with six quinolones

63 63 CHAPTER 5 DISCUSSION AND CONCLUSION 5.1 Optimization of HPLC condition In this study, HPLC using monolith column with UV detection was used to separate six quinolones (MAR, CIP, LOM, ENR, SAR, and DIF). The effects of mobile phase on analysis of quinolones were optimized. Three parameters of mobile phase were investigated. The ph of buffer, the percentage of organic solvent and the type of buffer in mobile phase were studied ph Effect The effect of buffer ph varying from 3.2 to 4.5 on the chromatographic separation of six quinolones was first investigated. It can be seen that buffer ph had a little effect on the retention times of MAR, CIP, LOM, and SAR except ENR and DIF. The retention times of ENR and DIF increased when buffer ph was increased. This is because quinolones are amphoteric compounds which have carboxylic group (acidic functional groups) and amine group in piperazine ring (basic functional groups) in their structure. pk a1 and pk a2 values of carboxylic group and amine group for MAR, CIP, LOM, ENR, SAR, and DIF were 5.69 and 8.84, 5.76 and 8.68, 5.64 and 8.70, 5.69 and 6.68, 5.7 and 8.68, and 5.64 and 6.45, respectively (He, Soares, Adejumo, Mcdiarmid, Squibb, & Blaney, 2015, Babic, Horvat, Pavlovic, & Macan, 2007). Considering pk a1 of six quinolones, all analytes have not significantly different in pk a1. However, based on the pk a2 values, six analytes can be divided into two groups. The first groups are MAR (pk a2 = 8.84), CIP (pk a2 = 8.68), LOM (pk a2 = 8.70), and SAR (pk a2 = 8.68). The second group are ENR (pk a2 = 6.68) and DIF (pk a2

64 64 = 6.45). For the range of ph study (ph 3.2 to 4.5), carboxylic group in structure of all analytes are presented in non-ionized form. However, comparison of positively charge on nitrogen atom at piperazine ring in structure of quinolones, nitrogen atom in quinolones of the second group (ENR and DIF) has less positively charge than quinolones in the first group because pk a2 values of quinolones in the second group approach to ph range study more than pk a2 values of quinolones in the first group. From this reason, it can be concluded that, in the range of ph study, quinolones in the first group are more polar than quinolones in the second group, in other words, quinolones in the second group is less polar than quinolones in the first group. Consequently, changing in buffer ph affects to retention time of quinolones in the second group, namely ENR and DIF, but does not or less affect to retention time of quinolones in the first group. From the experiment results, the buffer ph influenced not only retention time, but also affected to peak height. Peak height of all analytes slightly decreased when the buffer ph increased. This is because nitrogen atom at piperazine ring in structure of quinolones are likely to tailing peaks and band broadening in reversed phase chromatography due to interactions with hydrogen bonding of nitrogen atom at piperazine ring in structure of quinolones with unprotected silanol group of the stationary phase (Maraschiello, Cusido, Abellan, & Vilageliu, 2001, Lara, Iruela, & Campana, 2013). In acidic conditions, the nitrogen atom at piperazine ring in quinolone structure and unprotected silanol groups of the stationary phase become protonation, and hydrogen bonding between nitrogen atom and silanol groups decreased with decreasing buffer ph from 4.5 to 3.2. With these reasons, increasing in

65 65 peak height and peak symmetry was obtained with decreasing buffer ph. Consequently, an optimum buffer ph of 3.7 was selected The percentage of organic solvent effect The percentage of organic solvent was studied. In this work, acetonitrile was chosen for organic solvent in mobile phase because of low UV absorption (190 nm), low viscosity, and commonly used in HPLC. In the study, percentage of acetonitrile was changed from 15% to 18%, and it was found that retention time of all quinolones decreased with increasing percentage of acetonitrile, and peak height increased. These are because polarity of mobile phase decreased with increasing percentage of acetonitrile in mobile phase, in other word, elution strength increased as increasing percentage of acetonitrile. In addition, low band broadening obtained with increasing percentage of acetonitrile leading increase in peak height. However, the resolution decreased because of less retained analytes. Consequently, based on resolution and analysis time, 17 percentage of acetonitrile in mobile phase was chosen Effect of buffer type In reversed-phase chromatotraphy, buffer typre or counterion could affact chromatographic separation in terms of retention, and peak shape. In this work, three counter-ions, namely acetate, formate, and citrate at ph 3.7, were chosen to study. It can be seen that citrate counter-ion provided more retention of protonated basic analytes than acetate and formate counter-ions. In addition, peak shape in terms of peak height obtaining from citrate counter-ion is better than peak shape obtaining from acetate and formate counter-ions. Citrate ion has three negatively charged positions coming from carboxylic groups in its structure, but acetate and formate ions

66 66 have one negatively charged site in their structure. With these reasons, citrate counterion disrupts the analyte solvation shell and increase the analyte hydrophobicity more than acetate and formate ions (Kazakevich, Lobrutto, & Vivilecchia, 2004, Phechkrajang, 2010). Considering resolution, citrate counter-ion was chosen because all resolutions were baseline separation. 5.2 Optimization of QuEChERS method The QuEChERS method was attracted in the last few years because this method combined few step and reducing the time need to complete the extraction and clean up processes. The QuEChERS procedure according to literature review (Lucatello, Cagnardi, Capolongo, Ferraresi, Bernardi, & Montesissa, 2015) for determination of quinolones in animal muscle tissues has been adapted in this work for chicken muscle samples. The parameters of QuEChERS method were studied. The sorbent type, sorbent amount, and the extractant volume were optimized Sorbent type For dispersive solid phase extraction step (or clean-up step) in QuEChERS method, three types of sorbents (aminopropyl (NH 2 ), florisil, and silica sorbent) used for eliminating matrix in clean-up step were investigated. Chromatograms from analysis of a chicken muscle sample without spiking standard quinolones are shown in Figure 4-5 (a). It was found that silica and florisil reduced matrix effect better than aminopropyl during analysis time between 2 and 8 min. Figure 4-5 (b) shows the chromatograms obtaining from analysis of a chicken muscle sample spiking with standard quinolones. It can be seen that no interference peaks were found when silica sorbent was used. However, matrix interference was observed

67 67 when aminopropyl sorbent was employed, and there were no standard quinolone peaks by using florisil as a sorbent. Considering polarity of sorbents, polaritiy of florisil sorbent is higher than silica sorbent, and aminopropyl sorbent is lowest polarity (Varelis, 2009). In the case of florisil sorbent, it can be used as a strong polar sorbent, and can adsorb polar matrix from samples (such as fatty acids and organic acids) and quinolones, which have polar groups (carboxylic and amine groups) in their structure. Silica sorbent is lower polarity than florisil sorbent. It can remove polar matrix like florisil. However, from the experimental results, silica sorbent did not adsorb or adsorbed quinolones a little. For aminopropyl sorbent, this sorbent is lowest polarity. It can be seen that some polar matrix were not eliminated. The results showed that silica sorbent can use for cleanup step better than florisil and aminopropyl sorbents. Therefore, silica sorbent was selected to study percentage of recovery. The recoveries of quinolones obtaining from silica sorbent were between 81% and 95% Sorbent amount The amount of silica sorbent for dispersive solid phase extraction in QuEChERS method was optimized. According to chapter 4, the effect of the amount of silica sorbent on the recoveries of quinolones was studied. The experiment was evaluated by varying amount of silica sorbent (200 mg, 250 mg and 300 mg) for clean-up process in QuEChERS method. According to Figure 4-7 in chapter 4, it can be seen that recoveries of quinolones slightly increased when silica amount increased from 200 mg to 250 mg for all analytes excluding SAR. In addition, virtually all recoveries decreased with changing 250 mg to 300 mg of silica sorbent because some

68 68 quinolones were adsorbed on the excess silica sorbent. Therefore, 200 mg of silica sorbent was selected to be an optimum amount Extractant volume The volume of extraction solvent for extracting quinolones in chicken muscle samples was optimized for QuEChERS method ml and ml of extractant (5% formic acid in acetonitrile) were investigated. According to Figure 4-8 in chapter 4, it was found that recoveries of all analytes significantly increased when 3.00 ml of extractant was changed to ml because extraction efficiency increased with increasing volume of 5% formic acid in acetonitrile. Recoveries obtained from ml of extractant were between 82% and 104%. In this work, high recoveries were required for determination of quinolones residue in chicken muscle samples. Consequently, ml of extractant was applied to be an optimum volume. 5.3 Optimization of SPE method The SPE method was used to clean-up and preconcentration for determination of quinolones in samples. In this work, the SPE method was applied to extract quinolones residue in chicken muscle samples. The SPE type (sorbent), acid amount in the eluant, and acid type in the eluant were optimized SPE type The SPE was used to preconcentrate and clean-up quinolones residue in chicken muscle sample. According to chapter 4, the sorbent type in SPE including Oasis HLB, Oasis tc18, and Oasis MAX were investigated. It was found that HLB sorbent provided the higher recoveries than Oasis tc18 and Oasis MAX. This is

69 69 because the Oasis HLB is made by a polymeric macroporous poly(divinylbenzene-co- N-vinylpyrrolidone). The benefits of polymeric reversed phase sorbents are higher loading capacities than general silica based reversed phase sorbent thus the amounts of sorbent lower are needed (Janusch, Scherz, Mohring, Stahl, & Hamscher, 2014, Bailac et al., 2004). Furthermore, the Oasis HLB sorbent provides superior reversed phase sorbents with a hydrophilic N-vinylpyrrolidone for improved retention of polar analytes. Meanwhile, Oasis MAX is a strong mixed mode anion exchange. The Oasis MAX sorbents have positively charged functional groups that retain negatively charged anions. In addition, the Oasis tc 18 has overall similar retention capacity as C 18 sorbent but the trifunctional bonding on solid support (According to Figure 2-13 in chapter 2) makes tc 18 better hydrophobic stability than C Acid amount in eluant The effect of acid amount in the methanol on the recoveries of quinolones was evaluated. According to chapter 4, the experiment was investigated by varying percentage of acetic acid in methanol from 1% to 4%. It was found that recoveries of all quinolones increased with changing 1% to 2% acetic acid. From 2% to 4% acetic acid, almost recoveries of all analytes slightly decreased. All recoveries of quinolones obtaining from 2% acetic acid in eluant provided satisfactory results with the recoveries between 84% and 101%. Because 2% of acetic acid in methanol is sufficient amount of acid for elution of quinolones on Oasis HLB sorbent by disruption the hydrogen bonding between polar group of quinolones and sorbents (Janusch, Scherz, Mohring, Stahl, & Hamscher, 2014).

70 Acid type in the eluant The effect of acid types (acetic acid, formic acid, and propionic acid) on the recoveries of quinolones was evaluated. To study the effect of acid type on recovery of quinolones, 2% of each acid in methanol was used. According to Figure 4-11, it can be seen that the recoveries of quinolones obtaining from acetic acid, formic acid, and propionic acid were in range 86% to 98%, 76% to 98%, and 58% to 90%, respectively. The poor recoveries of quinolones were obtained when propionic acid was used because of lowest acid strength (Fajardo, Canto, Brown & Nesic, 2007). In addition, the recoveries of most analytes with the exception of DIF obtaining from acetic acid were more than recoveries obtaining from formic acid. Consequently, acetic acid was a suitable acid, and chosen for further experiment. 5.4 Comparison of the QuEChERS method and SPE method In terms of LOD, it was found that LODs of quinolones obtaining from QuEChERS method, and SPE method were in range 100 µg/kg and 300 µg/kg, and 5 µg/kg and 14 µg/kg, respectively. In general, the quantification of quinolones residue in biological samples with QuEChERS method has been achieved by high sensitivity detector such as mass spectrometry and fluorescence detection (Agui, Gracia, Blanco, & Campana, 2011, Lucatello, Cagnardi, Capolongo, Ferraresi, Bernardi, & Montesissa, 2015). Accordingly, QuEChERS method is unsuitable procedure for the analysis of quinolones residue with UV detector because the LODs of quinolones from QuEChERS method is insufficient for quantification of the quinolones below their MRLs. From the experimental results, it can be seen that the LODs of quinolones by using SPE procedure are lower than their MRLs (European

71 71 commission, 2010). Consequently, SPE method was chosen for the analysis of quinolones residues in chicken muscle samples. 5.5 Validation of the method The performance characteristics of the proposed method were achieved by the validation procedure according to the criteria by the Food and Drug Administration (FDA) guideline for bioanalytical assay validation (U.S. Food and Drug Administration, 2012). Six quinolones were detected at the lower level with the propose method. The analytes were carried out under the optimized SPE and HPLC methods. It has been reported that the analytes signal could be affected by by matrix effect of biological samples (Tsai et al., 2009) (Agui, Gracia, Blanco, & Campana, 2011) (Moema, Nindi, & Dube, 2012). Thus, to provide reliable results from chicken muscle samples, the matrix matched calibration curves were chosen in analysis to compensate for interference effects from matrix. The matrix matched calibration curves were constructed by spiking blank chicken muscles sample with standard quinolones at five levels of concentrations. The five points of calibration curves were plotted between peak area and concentration of quinolones. LOQs of target quinolones in chicken muscle samples were lower than maximum residue limits (MRLs) of international organizations. The correlation coefficients (r 2 ) of the matrix matched calibration curves higher than were obtained for all quinolones thus indication of good linearity of the method within the concentration range. The percentage relative standard deviations (%RSD) of intra-day and inter-day precision at three concentration levels for each quinolones lower than 9% was obtained in all cases. The recoveries of quinolones from chicken muscles at three concentration

72 72 levels higher than 83% were obtained. The results of validation were achieved within the FDA recommendation. 5.6 Analysis of quinolones in chicken muscle samples Determination of quinolones residue in chicken muscle samples was achieved by using the proposed method. The chicken muscle samples were bought from local supermarkets in the area of Chon Buri (Thailand). Next, the chicken muscle samples were extracted according to the proposed procedure and the samples were analyzed by HPLC using monolith column with UV detection. According to Chapter 4 in Table 4-5, the amounts of quinolones were detected lower than their MRLs of each quinolones. Consequently, the chicken muscle samples from local supermarket in the area of Chon Buri (Thailand) are safety for consumer. 5.7 Conclusions The screening method was developed and validated for the determination of six quinolones in chicken muscle samples. The SPE method was applied to clean up samples before analysis with monolithic high-performance liquid chromatography. The six quinolones were detected within 7 min. The time was used for separation of quinolones faster than previous work (Zhao, Jiang, Li, Mi, Li & Shen, 2007, Lombardo, Gamiz,, Blanco,, & Campana, 2011, Moema, Nindi, & Dube, 2012) because high effective of monolithic column was used. The proposed method was validated according to the international guidelines and good validation data were achieved for calibration curve, LOD, LOQ, recovery and precision (intra-day, interday). The method was applied to determine quinolones in chicken muscle sample

73 73 from different supermarket within Chon Buri. The low concentration of quinolones was detected in chicken muscle samples. These concentrations of quinolones in samples are lower than the MRLs established by the European Union (European commission, 2010).

74 74 REFERENCES Agui, M. L., Gracia, L. G., Blanco, C. C., and Campana, A. M. G. (2011), Comparison of different sample treatments for the analysis of quinolones in milk by capillary-liquid chromatography with laser induced fluorescence detection, Journal of Chromatography A, 1218, Alegre, M. R., Vicente, J. P., Romero, J. E., and Broch, S. C. (2010). Analysis of selected veterinary antibiotics in fish by micellar liquid chromatography with fluorescence detection and validation in accordance with regulation 2002/657/EC, Food Chemistry, 123, Amran, M. S., Hossain, M. R., Baki, M. A., Amjad, F. M., Sultana, S., and Hossain, M. A. (2011). Analysis of Lomefloxacin by Spectrophotometry: Development of Simple Quantitative Analytical Method, Bangladesh Pharmaceutical Journal, 14, Anastassiades, M., and Lehotay, S. J. (2003). Fast and Easy Multiresidue Method Employment Acetonitrile Extraction/Partitioning and dispersive Solid Phase Extraction for the Determination of Pesticide Residues in Produce, Journal of AOAC, 86, Andersson, M. I., and Macgowan, A. P. (2003). Development of the quinolones, Journal of Antimicrobial Chemotherapy, 51, Babic, S., Horvat, A. J. M., Pavlovic, D. M., and Macan, M. K. (2007). Determination of pk a values of active pharmaceutical ingredients, Trends in Analytical Chemistry, 26,

75 75 Bailac, S., Barron, D., and Barbosa, J. (2006). New extraction procedure to improve the determination of quinolones in poultry muscle by liquid chromatography with ultraviolet and mass spectrometric detection, Analytica Chimica Acta, 580, Bailac, S., Ballesteros, O., Lozano, E. J., Barron, D., Nebot, V. S., Navalon, A., Vilchez, J.L., and Babosa, J. (2004), Determination of quinolones in chicken tissues by liquid chromatography with ultraviolet absorbance detection, Journal of Chromatography A, 1029, Cabrera, K. (2004). Applications of silica-based monolithic HPLC columns, Journal of Separation Science, 27, Christie, W. W. (1992). Solid phase extraction columns in the analysis of lipids, Advances in Lipid Methodology, 1, Christodoulou, E. A., Samanidou, V. F., and Papadoyannis, I. N. (2007). Validation of an HPLC-UV method according to the European Union Decision 2002/657/EC for the simultaneous determination of 10 quinolones in chicken muscle and egg yolk, Journal of Chromatography B, 859, Chrom, S. R. O. (2001). Speed and performance in monolithic form chromolith HPLC columns, Bratislava, Slovakia. Clemente, M., Hermo, M. P., Barron, D., and Barbosa, J. (2006). Confirmatory and quantitative analysis using experimental design for the extraction and liquid chromatography UV, liquid chromatography mass spectrometry and liquid chromatography mass spectrometry/mass spectrometry determination of quinolones in turkey muscle, Journal of Chromatography A, 1135,

76 76 Delepine, B., Pessel, D.H., and Sanders, P. (1998). Simultaneous determination of six quinolones in pig muscle by liquid chromatography-atmospheric pressure chemical ionization mass spectrometry, The Analyst, 123, Elghany, W. A. A., and Madian, K. (2011). Control of Experimental Colisepticaemia in Broiler Chickens Using Sarafloxacin, Life Science Journal, 8, Evaggelopoulou, E. N., and Samanidou, V. F. (2013). HPLC confirmatory method development for the determination of seven quinolones in salmon tissue (Salmo salar L.) validated according to the European Union Decision 2002/657/EC, Food Chemistry, 136, European Commission, (2010). On pharmacologically active substances and their classification regarding maximum residue limits in foodstuffs of animal origin, Reguration No. 37/2010, Official Journal of the European Communities, 15, Fajardo, Canto, Brown, and Nesic, (2007). Effect of organic acids in CO 2 corrosion, corrosion conference & expo, 07319, Faria, A. M., Collins, C. H., and Jardim, C. S. F. (2009). State-of-the-Art in Immobilized Polymer Stationary Phases for High-Performance Liquid Chromatography, Journal of the Brazilian Chemical Society, 20, Ferenc, A. Z., and Biziuk, M. (2006), Solid phase extraction techinique-trend, Opportunities and Applications, Polish Journal of Environmental Studies, 15, Foods Program Science and Research Steering Committee (SRSC) (2012), Guidelines for the Validation of Chemical Methods for the FDA Foods Program. U.S.

77 77 Gigosos, P. G., Revesado, P. R., Cadahia, O., Fente, C. A., Vazquez, B. I., and Franco, C. M. (2000). Determination of quinolones in animal tissues and eggs by high-performance liquid chromatography with photodiode-array detection, Journal of Chromatography A, 871, Galarini, R., Fioroni, L., Angelucci, F., Tovo, G. R., and Cristofani, E. (2009). Simultaneous determination of eleven quinolones in animal feed by liquid chromatography with fluorescence and ultraviolet absorbance detection, Journal of Chromatography A, 1216, Hassouan, M. K., Ballesteros, O., Taoufiki, J., Vılchez, J. L. Aguilera, M. C., and Navalon, A. (2007). Multiresidue determination of quinolone antibacterials in eggs of laying hens by liquid chromatography with fluorescence detection, Journal of Chromatography B, 852, Hassouan, M. K., Ballesteros, O., Zafra, A., Vilchez, J. L., and Navalon, A. (2007). Multiresidue method for simultaneous determination of quinolone antibacterials in pig kidney samples by liquid chromatography with fluorescence detection, Journal of Chromatography B, 859, He, K., Soares, A. D., Adejumo, H., Mcdiarmid, M., Squibb, K., and Blaney, L. (2015). Detection of a wide variety of human and veterinary fluoroauinolone antibiotics in municipal wastewater and wastewater-impacted surface water, Journal of Pharmaceutical and Biomedical Analysis, 106, Hermo, M. P., Barron, D., and Barbosa, J. (2005). Determination of residues of quinolones in pig muscle comparative study of classical and microwave extraction techniques, Analytica Chimica Acta, 539,

78 78 Hermo, M. P., Barron, D., and Barbosa, J. (2006). Development of analytical methods for multiresidue determination of quinolones in pig muscle samples by liquid chromatography with ultraviolet detection, liquid chromatographymass spectrometry and liquid chromatography-tandem mass spectrometry, Journal of Chromatography A, 1104, Herrera, A. V. H., Borges, J. H., Delgado, M. A. R., Herreo, M., and CiFuentes, A. (2011). Determination of quinolone residues in infant and young children powdered milk combining solid-phase extraction and ultra-performance liquid chromatography-tandem mass spectrometry, Journal of Chromatography A, 1218, Hoof, N. V., Wasch, K. D., Okerman, L., Reybroeck, W., Poelmans, S., Noppe, S., and Brabander, H. D. (2005). Validation of a liquid chromatography tandem mass spectrometric method for the quantification of eight quinolones in bovine muscle, milk and aquacultured products, Analytica Chimica Acta, 529, Hoven, R. V. D., Wagenaar, J. A., and Walker, R. D. (2000). In vitro activity of difloxacin against canine bacterial isolates, Journal of Veterinary Diagnostic Investigation, 12, Hunter, R. P., Koch, D. E., Coke, R. L., Carpenter, L. W., and Isaza, R. (2007). Identification and comparison of marbofloxacin metabolites from the plasma of ball pythons (Python regius) and blue and gold macaws (Ara ararauna), Journal of Veterinary Pharmacology and Therapeutic, 30, Janusch, F., Scherz, G., Mohring, S. A. I., Stahl, J., and Hamscher, G. (2014), Comparison of different solid-phase extraction materials for the

79 79 determination of fluoroquinolones in chicken plasma by LC-MS/MS, Journal of Chromatography B, 951, Johnston, L., Mackey, L., and Croft, M. (2002). Determination of quinolones and fluoroquinolones in fish tissues and seafood by high-performace liquid chromatography with electrospray ionization tandem mass spectrometric detection, Journal of Chromatography A, 982, Jun, Z. S., Cun, L., Yang, J. H., Yu, L. B., and Zhong, S. J. (2007). Simultanious determination of 7 quinolones residues in animal muscle tissue by high performance liquid chromatography, Chinese Journal of Analytical Chemistry, 35, Karbiwnyk, C. M., Carr, L. E., Turnipseed, S. B., Andersen, W. C., and Miller, K. E. (2007). Determination of quinolone residues in shrimp using liquid chromatography with fluorescence detection and residue confirmation by mass spectrometry, Analytica Chimica Acta, 596, Kazakevich, Y. V., Lobrutto, R., and Vivilecchia, R. (2004). Reversed-phase highperformance liquid chromatography behavior of chaotropic counteranions, Journal of Chromatography A, 1064, Khargharia, S., Barua, C. C., Mohan, P., and Bhattacharya, M. (2005). Pharmacokinetic Studies of Enrofloxacin in Yak after Intramuscular Administration, Iranian Journal of Pharmacology and Therapeutics, 4, Lombardo, A. M., Gamiz, G. L., Blanco, C. C., and Campana, G. M. A. (2011). Comparison of different sample treatments for the analysis of quinolones

80 80 in milk by capillary-liquid chromatography with laser induced fluorescence detection. Journal of Chromatography A, 1218, Lara, F. J., Iruela, M. D. O., and Campana, A.M.G. (2013). On-line exchange solidphase extraction coupled to liquid chromatography with fluorescence detection to determine quinolones in water and human urine, Journal of Chromatography A, 1310, Li, C., Jiang, H. Y., Zhao, S. J., Zhang, S. X., Ding, S. Y., Li, J. C., Li, X. W., and Shen, J. Z. (2008). Simultaneous determination of fluoroquinolones and sulfonamides in chicken muscle by LC with fluorescence and UV detection, Chromatographia, 68, Lucci, P., Pacetti, D., Nunez, O., and Frega, N. G. (2012). Chromatography The Most Versatile Method of Chemical Analysis, Rijeka, Croatia. Lucatello, L., Cagnardi, P., Capolongo, F., Ferraresi, C., Bernardi, F., and Montesissa, C. (2015). Development and validation of an LC-MS/MS/MS method for the quantification of fluoroquinolones in several matrices from treated turkeys. Food Control 48, Mansoor, M. A. (2002). Encyclopedia of life sciences, London: Nature Publishing Group. Maraschiello, C., Cusido, E., Abellan, M., and Vilageliu, J. (2001). Validation of an analytical procedure for the determination of the fluoroquinolone ofloxacin in chicken tissues, Journal of Chromatography B, 754, Moema, D., Nindi, M. M., and Dube, S. (2012), Development of a dispersive liquidliquid microextraction method for the determination of fluoroquinolones in

81 81 chicken liver by high performance liquid chromatography, Analytica Chimica Acta, 730, Meloan, C. E. (1999). Chemical separation: principles, techniques, and experiment, New York: John wiley & Sons INC. Phechkrajang, C. M. (2010). Chaotropic effect in reversed-phase HPLC: A Review, Mahidol University Journal of Pharmaceutical Science, 37, 1-7. Poudel, B. K., & Thapa, P. (2013). In vitro interactions of ciprofloxacin with selected drugs and excipients, Journal of Science, Engineering and Technology, 9, Qiao, F., & Sun, H. (2010). Simultaneous extraction of enrofloxacin and ciprofloxacin from chicken tissue by molecularly imprinted matrix solid-phase dispersion, Journal of Pharmaceutical and Biomedical Analysis, 53, Ramos, M., Aranda, A., Garcia, E., Reuvers, T., and Hooghuis, H. (2003). Simple and sensitive determination of five quinolones in food by liquid chromatography with fluorescence detection, Journal of Chromatography B, 789, Rubies, A., Vaquerizo, R., Centrich, F., Compano, R., Granados, M., and Prat, M. D. (2007). Validation of a method for the analysis of quinolones residues in bovine muscle by liquid chromatography with electrospray ionisation tandem mass spectrometry detection, Talanta, 72, Sabir, A. M., Moloy, M., and Bhasin, P. S. (2013). HPLC method development and validation : A review, International Research Journal of Phamacy, 4, Stubbings, G., and Bigwood, T. (2009). The development and validation of a multiclass liquid chromatography tandem mass spectrometry (LC MS/MS) procedure for the determination of veterinary drug residues in animal tissue

82 82 using a QuEChERS (Quick, Easy, Cheap, Effective, Rugged and Safe) approach, Analytica Chimica Acta, 637, Tsai, W. H., Chuang, H. Y., Chen, H. H., Huang, J. J., Chen, H. C., Cheng, S. H., and Huang, T. C. (2009), Application of dispersive liquid-liquid microextraction and dispersive micro-solid-phase extraction for the determination of quinolones in swine muscle by high-performance liquid chromatography with diode-array detection, Analytica Chimica Acta, 656, Varelis, P. (2009). Advances in high-performance liquid chromatography and its application to analysis of food and beverages, in: Otles, S. (Ed.), Handbook of food analysis instruments. Florida: CRC press. Voermans, M., Soest, J. M. V., Duijkeren, E. V., and Ensimk, J. M. (2006). Clinical efficacy of intravenous administration of marbofloxacin in a Staphylococcus aureus infection in tissue cages in ponies, Journal of Veterinary Pharmacology and Therapeutic, 29, Wang, J., Macneil, J. D., and Kay, J. F. (2012). Chemical analysis of antibiotic residues in food, New Jersey: John wiley & Sons INC. World Health Organization. (1998). Use of Quinolones in Food Animals and Potential Impact on Human Health, Report of a WHO meeting, WHO, Geneva, Switzerland. Xia, L. Y., Wei, L., Lian, Z. X., and Ming, Y. (2012). Simultaneous determination of fourteen veterinary antibiotics in animal feces by solid phase extraction and high performance liquid chromatography, Chinese Journal of Analytical Chemistry, 40,

83 83 Yu, H., Mu, H., and Hu, Y. M. (2011). Determination of fluoroquinolones, sulfonamides, and tetracyclines multiresidues simultaneously in porcine tissue by MSPD and HPLC DAD, Journal of Pharmaceutical Analysis, 2, Yu, H., Tao, Y., Chen, D., Pan, Y., Liu, Z., Wang, Y., Huang, L., Dai, M., Peng, D., Wang, X., and Yuan, Z. (2012). Simultaneous determination of fluoroquinolones in foods of animal origin by a high performance liquid chromatography and a liquid chromatography tandem mass spectrometry with accelerated solvent extraction, Journal of Chromatography B, 886, Zhao, S., Jiang, H., Li, X., Mi, T., Li, C., and Shen, J. (2007). Simultaneous determination of trace levels of 10 quinolones in swine, chicken, and shrimp muscle tissues using HPLC with programmable fluorescence detection, Journal of agricultural and food chemistry, 55,

84 APPENDIX 84

85 85 HPLC instrument Figure 1 High performance liquid chromatography coupled with UV detector (HP 1050 series, Hewlett Packard, Germany). Figure 2 Chromolith column RP-18e (Merck, Germany) Calculation in this method 1. Mean (X) = X 1+X 2 +X 3 + +X i n Where X is value of each sample and n is number of total sample