CHAPTER 5 NANOMATERIALS AND CLASSICAL MATRICES FOR EFFICIENT GLUCOSE BIOSENSOR - A COMPARATIVE STUDY

Transcription

1 CHAPTER 5 NANOMATERIALS AND CLASSICAL MATRICES FOR EFFICIENT GLUCOSE BIOSENSOR - A COMPARATIVE STUDY Overview of the Chapter In this chapter we have presented a comparative study of immobilization techniques and materials for fabrication of glucose biosensor. Conventional matrices and membranes have been compared with current-day nano sized materials for optimum enzyme immobilization. Comparative analysis demonstrates covalently bound enzyme onto nanomaterials to be the most suitable method and matrix for enzyme immobilization. The characteristic compatible curvature and nanoelectrode behavior of nanomaterials were further exploited for enhancing biosensing capabilities of glucose biosensor. Present comparative analysis might serve as a set of design criteria to help engineers fabricate an efficient biosensor. The chapter is designed as, 5.1 Introduction 5.2 Experimental Synthesis of Nanoparticles Enzyme Immobilization Electrostatic Immobilization of GOx on AuNPs Covalent Immobilization of GOx on Fe 3 O 4 and AuNPs Physical adsorption of GOx on PVDF and NC membranes Entrapment of GOx in calcium alginate beads Entrapment of GOx in polyacrylamide gel Estimation of GOx immobilization efficiency Activity analysis Enzyme Leakage Biosensor fabrication 5.3 Results and Discussion Characterization of Nanoparticles Optimization of parameters for enzyme activity 84

2 5.3.3 GOx Immobilization Efficiency Effect of immobilization method/material on enzyme activity Analysis of Free and Immobilized Enzyme Adsorption of Enzyme on membranes Entrapment of Enzyme Electrostatic and covalent immobilization of Enzyme onto NPs Thermal Stability Analysis Biosensor performance Analysis Voltammetric Response EIS Studies Current Response Detection Limit Sensitivity Reproducibility and Precision Interference Stability Response time 5.4 Conclusion 5.1 Introduction Biosensor history dates back to 1962 when Clark and Lyons proposed the concept of enzyme electrodes for glucose determination wherein glucose oxidase enzyme carried by hydrogel was sandwiched between two semipermeable membranes [26]. Since then, there have been extensive research efforts towards the development of an efficient glucose biosensor. The efficiency of a biosensor, i.e., the sensitivity, stability (operational and storage stability) and response time, is very much subjected to the kind of support material used and the method of immobilization. Broadly, enzyme immobilization can be categorized as bound or entrapped. The bound enzyme can be a) physical or ionic adsorption [136, 137], b) covalently bound to a support matrix or c) covalently cross-linked enzyme molecules [138, 139]. Entrapment on the other hand can be a) membrane encapsulation or b) matrix entrapment. The most suitable support material and immobilization method vary depending on the enzyme and particular application. Key parameters for selection of suitable material and method are binding capacity 85

3 of the material, stability, retention of enzyme activity and minimizing leakage of enzyme after immobilization on the material. The membranes used in glucose biosensors have evolved continuously to minimize interference from reactive species through selective permeability to certain species only. Usage of semi permeable membrane ensures high selectivity and wide linear range of glucose detection, however it results in large response time of the order of minutes. The latter is attributed to the delay in electron transfer from enzyme active centre to electrode due to increased distance between the electrode and the enzyme reaction centre. Introduction of mediators facilitates faster electron transfer from enzyme active centre to the electrode surface resulting in reduction of response time from few min [57, 140] to as low as few seconds [37, 141, 142]. In addition to this, mediators lower the required applied potential for redox reactions thereby eliminating or reducing the interference from other oxidzable species at the applied potential. However, a major problem faced with the usage of mediators is its diffusibility into the bulk regime. Direct immobilization of enzymes onto the matrices at electrode surface eliminates the need of usage of mediators at the same time reduces the response time [50, 143]. Further developments in terms of suitable choice of the membranes like polycarbonate membrane, polyvinylpyrrolidone) (PVP) polymer [144], Nafion membrane [145], etc. resulted in extending the linear range of the biosensor. Linearity of a biosensor is governed by rate of diffusion of the electroactive species. Diffusion rates were controlled by selective functionalization or coating of membranes such that substrate diffusivity is decreased and hence increased linearity. Yang et al. coated the outer membrane of biosensor with Nafion polymer (having negatively charged functional groups SO - 3 ) and could extend the linear range to 25mM glucose and response time to as low as 10s [146]. Moreover, the repulsion of negatively charged interfering species (ascorbic acid and uric acid) from the SO - 3 groups of Nafion polymer minimizes the interference [146, 147]. Further improvement in linearity (upto 30 mm) was achieved by using cellulose membrane as outer membrane [148] closely followed by polyurethane (upto 31 mm) and a much higher linearity upto 37.7 mm was achieved with polyvinylchloride coating solutions as outer membrane [149]. In addition to linearity, current response and sensitivity of the biosensor was enhanced with the introduction of conducting polymers like polypyrrole and polyaniline [ ]. However, replacing membranes with nanomaterials like gold, nickel, copper, silica, carbon nanotubes etc. have revolutionized the biosensor research by reducing the response time to 2-10 s and sensitivities ranging from 5μA mm 1 cm 2 to as high as 135μA mm 1 cm 2. The governing 86

4 factors being the shape and size of the nanomaterials in addition to other physical and chemical properties. In this chapter a comparative study and analysis with design and loading of enzyme on classical and nanomaterial was performed to enable suitable choice of methods and materials for biosensor fabrication. 5.2 Experimental Synthesis of Nanoparticles The Nanoparticles of gold and magnetite were synthesized using the chemical methods as discussed in chapter 4. These synthesized NPs have been used as matrices for immobilization of enzyme, in addition to the conventional membranes purchased from market Enzyme Immobilization Glucose oxidase (GOx) was immobilized on different platforms using varied immobilization methods including covalent, electrostatic, entrapment and crosslinking. A simplified schematic representation of the electrostatic and covalent immobilization of GOx onto Au and Fe 3 O 4 NPs is shown in Figure 5.1. Figure 5.1: Schematic representation of methods of immobilization GOx onto Au and Fe 3 O 4 NPs. 87

5 Electrostatic Immobilization of glucose oxidase on AuNPs The citrate stabilized gold nanoparticles (AuNPs) were dispersed in a solution of oppositely charged polyelectrolyte, poly diallyl dimethyl ammonium chloride (PDADMAC) (see Figure 5.2). In order to achieve the uniform functionalization, 3mL fresh solution of the gold nanospheres(10 pmol/ml) was stirred with 1 ml PDADMAC in 3mL water for 2 h. The polyelectrolyte modified AuNPs were then centrifuged three times at rpm in order to remove excess polyelectrolyte. To the pellet, 20U of Glucose oxidase (GOx) was added, and mixture was incubated overnight with constant stirring at 4 o C. Mixture was centrifuged at rpm at 4ºC for 20 minutes to separate free enzyme (GOx) from the immobilized (GOx- AuNPs). The supernatant was stored for checking the activity of unbound enzyme while the pellet containing the GOx-AuNPs was washed repeatedly using phosphate buffer, ph 7.4. Figure 5.2: Schematic illustration of nanoparticles synthesis and functionalization Covalent Immobilization of Glucose oxidase on nanoparticles (Fe 3 O 4 and AuNPs) The amino functionalization of the Fe 3 O 4 NPs was achieved by coating with 3- aminopropyltriethoxysilane (3-APTES) while that of AuNPs was done with 3- aminopropanethiol (3-APT) (see Figure 5.2). This was followed by addition of 4ml of 88

6 glutaraldehyde solution (10% glutaraldehyde) to the NPs solution with continuous stirring for 1 h. Centrifugation was performed to remove excess reagents. To the pellet, 20U of Glucose oxidase (GOx) was added, and mixture was incubated overnight with constant stirring at 4 o C. Free enzyme (GOx) was separated from the immobilized (GOx-NPs) by centrifuging the solution at 12,000 rpm at 4ºC for 20 minutes. The supernatant was stored for checking the activity of unbound enzyme while the pellet containing the GOx-NPs was washed thrice using phosphate buffer, ph 7.4. The GOx-NPs complex thus obtained was redispersed and stored in 1 ml phosphate buffer solution ph 6.5 at 4 o C Physical adsorption of GOx on PVDF and NC membranes 10 µl of GOx solution (20U/10 µl) was immobilized onto 2 cm 2 piece of Polyvinylidene Fluoride (PVDF)and nitrocellulose e(nc) membranes (0.45 µm) slowly allowing each drop to dry before adding next drop. The membrane was air dried at room temperature for 1-2 hours and stored at 4 o C for further use Entrapment of GOx in calcium alginate beads Buffer solution containing 20 units of GOx was mixed well with 3% Na-alginate solution. This was followed by dropwise addition of the same in 0.2M CaCl 2 solution with the help of a syringe kept just above the surface of calcium chloride solution to ensure sphericity of the beads. The spherical beads of enzyme entrapped in calcium alginate gel were left undisturbed for 2 hrs for hardening. Afterwards the excess calcium chloride solution was decanted off and was replaced by autoclaved milliq water and stored at 4 o C till further use Estimation of GOx immobilization efficiency After GOx immobilization on different support, support matrices were washed with phosphate buffer (ph 6.5). The wash out buffer solution was collected in a tube and the amount of free enzyme was estimated. In case of entrapment methods, i.e., the calcium alginate beads were cut open and resuspended in phosphate buffer while in case of polyacrylamide gel, the pieces were crushed and resuspended in phosphate buffer followed by enzyme activity analysis. Here the amount of enzyme per bead (or gel piece) was calculated based on the total number of beads or gel pieces formed. Percentage immobilization of enzyme was calculated as 89

7 5.2.4 Activity analysis Enzyme activity was assayed using dinitrosalicylic acid (DNS) colorimetric method for temperatures ranging from 25 o C - 65 o C. Sample containing 2% glucose was incubated with free enzyme and immobilized enzymes in separate tubes in a water bath set at a fixed temperature. The expected reaction and byproducts are mentioned below: β D Glucose + Glucose oxidase FAD β D Gluconolactone + Glucose oxidase FADH 2 Glucose oxidase FADH 2 O 2 Glucose oxidase FAD H 2 O 2 At fixed intervals, 450 µl of sample is pipetted into fresh eppendorf tubes and 450 µl of dinitrosalicylic acid (1% DNS) was added to each of the tubes. These tubes were then immersed in a hot water bath, preset at 90 C, for minutes followed by addition of 150 µl of sodium potassium tartarate (40%) to each of the tubes. The expected reaction and by products are as below: β D Glucose + 3,5 dinitrosalicylic acid 3 amino, 5 nitrosalicylic acid + Gluconic acid Finally, the absorbance was recorded at 575 nm to estimate the amount of 3-amino,5- nitrosalicylic acid dye formed and hence the concentration of glucose present in each of the samples. The above process was repeated for temperatures between 25ºC to 65ºC. Finally, relative activity of the immobilized enzyme onto different substrates (NC and PVDF membrane, calcium alginate beads, polyacrylamide gel, Fe 3 O 4 NPsand AuNPs) as well free enzyme was calculated at each temperature in comparison to the optimum temperature Enzyme Leakage The stability of immobilized enzyme in terms of percentage loss or leakage of enzyme was assessed by comparing the immobilized enzyme activity at 0 hour and after 72 hours. During the 72 hour storage time, immobilized enzyme was washed thrice daily with phosphate buffer, ph 6.5. Percentage leakage of GOx GOx activity 0 h GOx activity 72 h GOx activity 0 h 90

8 5.2.6 BioSensor Fabrication A screen-printed electrode (SPE) was modified for detection of glucose with gold as both working and auxiliary electrode while Ag served as the reference electrode. The working electrode surface was activated by adding 10 µl of 3APT (10mM). The electrode surface was further modified by addition of 10 l of glutaraldehyde (2.5%), followed by addition of amino functionalized NPs (10pmol/ml) and allowed it to dry for 2 hrs. Finally, 20U of Glucose oxidase was layered on the nanoparticles modified electrode surface. The fabricated biosensor was rinsed with water, dried and stored in PBS (ph 6.5) at 4 C for further use. Figure 5.3 shows a simplified illustration of various functionalization stages involved in biosensor fabrication Figure 5.3: Schematic illustration of various functionalization stages involved in biosensor fabrication Characterization of fabricated biosensor The fabricated biosensors were characterized after each modification through Cyclic voltammetry (CV) and impedance analysis using an Electrochemical workstation CHI660D (CH Instruments). All electrochemical measurements have been conducted on a threeelectrode screen printed system, in phosphate buffer (0.1 M, ph 7.4) as the electrolyte. All the experiments were carried out at ambient temperature. The morphology of synthesized NPs was examined by transmission electron microscope (JEOL 2100F). Ultraviolet and visible (UV-Vis) absorption spectra of the aqueous dispersion of NPs were recorded on ND 1000 UV Vis Spectrophotometer (Nanodrop technologies) operated at 12 V. 91

9 5.3 Results and Discussion Parameters known to influence the activity of immobilized enzymes were examined, specifically, the matrix and method of immobilization. Traditional support materials for immobilization of enzymes like membranes/thin film (PVDF and NC) and polymeric encapsulation (Na-alginate beads and polyacryalmide gel) were compared with nanoparticles (gold and magnetite). The synthesized nanoparticles (Au and Fe 3 O 4 NPs) were characterized physically and chemically followed by activity analysis of covalently and electrostatically immobilized enzyme (GOx) onto these nanoparticles Characterization of Nanoparticles Gold and magnetite nanoparticles were characterized using UV-Visible spectroscopy and transmission electron microscopy (TEM). UV-Vis spectra of nanoparticles are shown in Figure 5.4. Gold nanoparticles show a characteristic absorption peak at 523 nm (Figure 5.4 A) while magnetite nanoparticles (Figure 5.4 B) shows absorption maximum at 400 nm. The shift in the peak absorption wavelength, in comparison with bulk particles, gives a qualitative idea of the size of the synthesized particles. Absorption maxima correlates with the energy difference (ΔE) or the band gap of the material under study. ΔE for any material can be calculated from the following equation E = n2 h 2 8ml 2 Where, l corresponds to the size of the particles, h is Planck s constant, m is mass of the particle and n is the principal quantum number. Also, E = hc E 1 l 2 Where c is velocity of light and λ is the wavelength of light absorbed. Smaller is the size of the particle larger will be the band gap or ΔE and hence smaller would be the wavelength of absorption. The band gap for bulk magnetite is 0.14 ev,[155] while gold being metal has overlapping valence and conduction band or in other words no band gap. Substituting the value of the absorption wavelength of gold and magnetite particles, i.e., 523 nm and 400 nm respectively, in equation (2) we get, E Au = /523 = 2.36 ev 92

10 E magnetite = /400 = 3.09 ev Since the energy band gap of the synthesized materials (both gold and magnetite particles) is larger than that of the corresponding bulk materials hence the synthesized particles are of much smaller dimension, probably structures with nanoscale dimensions. Figure 5.4: UV visible absorption spectrum and micrographs of (A) citrate-stabilized AuNPs and (B) Fe 3 O 4 NPs The size distribution and morphology of these nanoparticles were further confirmed by transmission electron microscopy. Figure 5.4A and 5.4B shows the transmission electron micrograph of the spherical gold and magnetite nanoparticles. TEM micrograph shows a narrow size distribution of the nanoparticles, with Fe 3 O 4 NPs having an average diameter of ~70 nm while AuNPs are of ~20 nm in diameter. 93

11 5.3.2 Optimization of parameters for Enzyme Activity Structural or conformational changes due to immobilization of enzyme leads to the reduction in its catalytic activity. Optimum temperature and ph for the enzymatic activity was evaluated by studying the the rate of conversion of glucose with time at different temperatures and ph. The amount of glucose left in the sample is estimated using DNS assay. Figure 5.5A shows relative activity of free enzyme GOx in phosphate buffer for ph value ranging from 5.0 to 9.0 at 37ºC. GOx shows optimum activity at ph 6.5 and more than 80% of the enzyme activity is retained for ph ranging between 5.0 and 8.0 while above ph 8.0 activity reduces drastically with only 40% activity remaining at ph 8.5 Figure 5.5: Relative activity of free GOx in solution with variation in (A) ph and (B) Temperature Moreover, enzyme activity at physiological ph (i.e.7.0) is 98% of that at optimum ph (ph 6.5). Thermal stability analysis done over a temperature range of 25ºC to 60ºC (Figure 5.5B) showed 40ºC to be the optimum temperature GOx Immobilization Efficiency Efficiency of immobilization of enzyme was estimated for different methods and matrices 1) physical adsorption onto NC and PVDF membranes 2) entrapment inside calcium alginate beads and polyacrylamide gel 3) covalent immobilization onto magnetite and gold nanoparticles and 4) electrostatic immobilization onto citrate functionalized gold nanoparticles. Figure 5.6 shows percentage immobilization of enzyme onto different matrices. 94

12 As seen from the bar diagram, when equal amount of enzyme (20 U) was used for immobilization, percentage immobilization is maximum in entrapment methods (~95%) closely followed by covalent immobilization on gold and magnetite nanoparticles (~92%) and 88% in case of electrostatic method while lowest immobilization percentage was observed in case of physical adsorption methods, i.e., on NC and PVDF membrane (~74%). The enhanced percentage of covalent as well as electrostatic immobilization on nanoparticles could be attributed to highly reactive nature of nanoparticles because of high surface energy. Moreover, the nanoparticles not just have high surface energy, rather high surface area as well leading to high enzyme loading capacity. Figure 5.6: Comparative analysis of conventional matrices with nanomaterials for GOx immobilization in terms of immobilization and leakage percentage, respectively Percentage loss or leakage of enzyme is maximum in case of physical methods (~15%) while in case of electrostatic immobilization onto nanoparticles enzyme leakage is slightly reduced (~8%). This could be attributed to weak Van der Waals forces in case of physical adsorption in comparison to the latter case with electrostatic interaction of enzyme with nanoparticles. However, covalent bonding being much stronger results in minimum leakage of enzyme (~2%). In case of entrapment methods with optimum pore size or crosslinking of polymer enzyme leakage was found to be approximately ~4%. Table 5.1 shows a Comparative analysis of efficiency of different immobilization methods. Percentage leakage of enzyme can be minimized by decreasing the pore size of beads and membranes or increasing the glutaraldehyde concentration in case of surface immobilization, however, this would lead to diffusional limitation and loss of enzyme activity. A balance between 95

13 percentage immobilization, leakage of enzyme and effect of immobilization on activity of enzyme is the deciding factor while choosing a particular method and matrix for immobilization. Table 5.1: Comparative analysis of efficiency of different immobilization methods Immobilization Method % Immobilization % Leakage Entrapment 95 4 Physical Adsorption Electrostatic Interaction 88 8 Covalent Attachment 92 2 In general, if the application is of single time usage (use and throw biosensors) then electrostatic immobilization on nanoparticles could be the method of choice while for long term usage applications covalent immobilization of enzyme onto nanoparticles would be the preferred method Effect of immobilization method/material on enzyme activity Activity Analysis of Free and Immobilized Enzyme Figure 5.7 shows variation of glucose concentration with time for free enzyme and GOx Figure 5.7: Variation of glucose concentration with time for free enzyme GOx and GOx immobilized onto various matrices 96

14 immobilized onto different matrices at optimum temperature (40 C). It can be seen that rate of change of glucose concentration with time is generally lower in case of immobilized enzyme than free enzyme suggesting decreased enzymatic activity in immobilized state than free solution form. Rate of conversion of glucose incase of electrostatic immobilization of GOx onto gold nanoparticles is close to that of free enzyme. The probable reasons for the observed difference in activity of GOx in each case are explained below Adsorption of Enzyme on Membranes Activity of GOx after adsorption of enzyme onto NC and PVDF membrane is calculated from the rate of change of glucose concentration as shown in Figure 5.7. The increased activity on NCM as compared to PVDF could be attributed to the nitro groups present on the NC membrane that results in activity enhancement on immobilization as reported by Tiller et. al. [156]. Moreover, NC membrane being hydrophilic in nature results in charge charge interaction and weak secondary Van der Waals interactions with the enzyme that will not cause much damage to the structure of the enzyme as compared to the strong hydrophobic interactions in case of PVDF membrane that results in decreased enzymatic activity Entrapment of Enzyme The effect of entrapment, of glucose oxidase in polyacrylamide gel and calcium alginate beads, on its catalytic activity is shown in Figure 5.7. It is observed that polyacrylamide gel entrapped enzyme showed larger activity than calcium alginate beads. This could be attributed to the fact that polyacrylamide gel is non-ionic in nature and hence the enzyme properties would be minimally affected. However, in case of calcium alginate gel, the positively charged divalent calcium ions crosslinking the negatively charged gluconic acid residues of alginate polymer, might affect the GOx enzyme properties whose active centre is negatively charged [136]. In addition to the above, the non-ionic character of polyacrylamide gel will not have much effect on diffusion of charged substrates and products, when used in an industry for production or for biosensor [157] Electrostatic and covalent immobilization of Enzyme onto nanoparticles Electrostatic immobilization of GOx onto gold nanoparticles showed higher activity than covalent immobilization onto gold and magnetite nanoparticles (Figure 5.6). Covalent immobilization of enzyme onto amino functionalized gold nanoparticles resulted in slightly lesser activity than in case of electrostatic immobilization on citrate stabilized gold 97

15 nanoparticles of same dimension (average diameter ~20 nm). This could be attributed to the stronger covalent bonds that might deform the structure of the enzyme and hence a reduced activity. However, the reduction in activity was less because of the amino functionalization of the NPs. Moreover, in case of covalent immobilization GOx immobilized onto gold nanoparticles showed higher activity than on magnetite nanoparticles. This could be explained on the basis of curvature of the nanoparticles depending on their size and adaptability (retention of conformation) of the enzyme on immobilization. Synthesized magnetite nanoparticles are of larger size (average diameter ~70 nm) while gold nanoparticles are of much smaller size (average diameter ~20 nm) which is more suitable for immobilization of GOx enzyme (size ~5-6 nm). This shows that size of nanoparticles is one of the key parameters that influence the interaction between protein and nanoparticles. Lundqvist et. al. had shown through a similar study on silica nanoparticles of different sizes the effect of particle curvature on protein activity [158]. It was shown that secondary structure of the proteins is strongly affected by particle curvature. You et al. further showed that not just the curvature rather size, shape, surface chemistry and charge of the NPs are responsible for the adsorption patterns and enzyme structure and function [159] Thermal Stability Analysis Thermal stability of immobilized enzyme was studied over a temperature range of 25 o C - 65 o C at ph 6.5 for each of the support matrix and method. Figure 5.8A and 5.8B shows the relative activity (relative to maximum activity at optimum temperature) of free and immobilized GOx onto conventional and nanomaterials respectively. As seen from Figure 5.8 maximum activity of the immobilized GOx is at 40 C except in case of entrapment of GOx in polyacrylamide gel and calcium alginate beads where optimum temperature was lower i.e. 37 C and 25 C respectively. Generally, at temperatures greater than optimum temperatures the activity of immobilized enzyme was found to be less than free enzyme but in case of covalent immobilization onto nanoparticles (both gold and magnetite) an enhanced thermal stability of the enzyme was observed. As discussed in the previous section this could be attributed to small size and appropriate curvature of the nanoparticles. However, in case of electrostatic immobilization of GOx onto nanoparticles at higher temperatures activity was lower than free enzyme. This suggests that curvature and size of nanoparticles alone is not responsible for 98

16 enhanced stability rather cumulative effect of covalent linkage in addition to these factors accounts for the observed trend. This is further corroborated by an exhaustive study on gold nanoparticles functionalized with different amino acids by You et. al.. It was shown that both electrostatic and hydrophobic interactions between the functionalized gold nanoparticles and proteins are responsible for stability of protein-nanoparticle complex. It was further observed that hydrophobicity binding affinity with nanoparticles and a slower rate of denaturation of proteins. Comparison of results for immobilization efficiency, leakage of enzyme, enzyme activity and thermal stability, indicates that the nanoparticles served as better platform for immobilization as compared to traditional methods including membrane or gel entrapment. Hence, further analysis of biosensing efficiency was done only for covalently immobilized GOx onto gold and magnetite nanoparticles. Figure. 5.8: Comparative thermal stability profiles of free GOx in solution and GOx immobilized onto (A) conventional matrices, (B) nanomaterials Biosensor performance analysis GOx-AuNPs and GOx-Fe 3 O 4 NPs conjugates were further used for biosensor application. Figure 5.3 represents simplified representation of the GOx immobilized onto the modified gold electrode with gold/magnetite nanoparticles for efficient detection of glucose. Finally, evaluation of the biosensing capabilities of the biosensors constructed by immobilization of GOx-AuNPs and GOx-Fe 3 O 4 NPs onto the working gold electrode was done in terms of 99

17 voltammetric response characteristic, impedance measurements, current response, linearity, detection limit, sensitivity, interference, stability, response time, reproducibility and precision Voltammetric Response The voltammetry response of glucose biosensor fabricated using GOx immobilized onto Fe 3 O 4 NPs-gold electrode and onto AuNPs-gold electrode is shown in Figure 5.9(A-E). The voltammetric studies were done in phosphate buffer ph 7.4 without using any mediator to evaluate the direct electron transfer reaction kinetics as facilitated by nanoelectrode behavior of Fe 3 O 4 NPs and AuNPs. Figure 5.9A represents the characteristic response of bare electrode (curve a), magnetite NPs (curve b) and gold NPs (curve c) modified electrodes in phosphate buffer at a scan rate of 100mV/s. Bare gold electrode shows a prominent reduction peak at a potential of 0.31 V however no visible oxidation peak is observed. Furthermore, modification of gold electrode surface with AuNPs is marked by a negative shift of 0.03V, at a reduction peak potential of 0.28 V with 2.6 times enhancement in peak current. A well-defined oxidation peak is observed at 0.042V. However, Fe 3 O 4 NPs modified electrodes show a much lower background current that is even less than bare gold electrode with a reduction peak at 0.49 V and peak current 3.2 times less than AuNPs modified electrode. This is due to difference in material characteristics of gold and magnetite indicative of semiconductor type behaviour of magnetite nanoparticles. Figure 5.9B and 5.9C represent the change in behavior after immobilization of GOx onto magnetite NPs and gold NPs modified electrode surfaces at scan rates of 20mV/s, 40mV/s, 60mV/s, 80mV/s and 100mV/s as shown by curve a-e respectively. As observed, there is approximately two fold reduction in current response pertaining to the insulator type behavior of the GOx enzyme. As seen from Figure 5.9B reduction and oxidation peaks of glucose oxidase enzyme are observed at 0.24V and V respectively at a scan rate of 20mV/s in case of Fe 3 O 4 NPs modified electrode. The peak-topeak potential separation ( Ep) was 0.229V at a formal potential (E ) of 0.015V. However, in case of GOx-AuNPs modified electrodes the redox peak potential of GOx is observed at 0.284V and 0.223V as seen in Figure 5.9C. Theoretically, for a one electron transfer reversible reaction the potential difference between oxidation and reduction peaks should be 0.059V. Since in the present case it is a two electron transfer reaction hence the Ep should be 0.059V/2 i.e V. The peak to peak separation potential (0.031V) of redox reaction of GOx in case of AuNPs modified electrode indicates 100

18 Figure 5.9: Cyclic voltammograms of(a)bare electrode (curve a),fe 3 O 4 NPs modified gold electrode (curve b), and the Au NPs modified gold electrode (curve c)in 0.1M phosphate buffer (ph 7.4) solution at scan rate of 100 mv/s. (B) Cyclic voltammogramin 0.1 M, phosphate buffer solution (ph 7.4) at different scan rates (20, 40, 60, 80, and 100, mv/s) for GOx-Fe 3 O 4 NPsgoldelectrode and (C)GOx-Au NPsgold electrode. (D)Linear variation of cathodic and anodic peak currents for GOx-Fe 3 O 4 NPsgold electrode, and (E) GOx-Au NPsgold electrode, respectively. reversible processes while Fe 3 O 4 NPs modified electrodes has Ep as 0.229V, depicting quasi-reversible nature of electrochemical processes at GOx-Fe 3 O 4 NPs electrode surface. Moreover, the peak current is much higher and comparatively sharper peaks are observed in case of GOx-AuNPs modified electrodes as compared to GOx-Fe 3 O 4 NPs modified electrodes. 101

19 Figure 5.9D and 5.9E shows linear variation of anodic and cathodic peak current with increasing scan rate ( ) for both electrodes, demonstrating a direct electron transfer behavior instead of diffusion controlled process that is marked by linear variation of Ipa and Ipc with 1/2. The resistance to electron flow from AuNPs and Fe 3 O 4 NPs was further evaluated through impedance spectroscopy EIS Studies The electron transfer resistance (R ct ) controls the electron transfer kinetics of the redox probe at the electrode interface. The value of R ct depends on the dielectric and insulating properties at the electrode/electrolyte interface. The linear portion at lower frequency of the Nyquist plots corresponds to the diffusion-limited process whereas the semicircle at higher frequencies corresponds to the electron-transfer limited process. R ct as calculated from the Figure 5.10 shows that the resistance to electron transfer was lowered when bare electrode (R ct : 140Ω, curve a) was modified with gold nanoparticles (R ct : 18Ω, curve d) whereas the R ct value increased to 30 Ω after immobilization of GOx on the surface of the modified electrode (curve e). The above results suggest that the gold nanoparticles are acting as nanoelectrodes and are helping in promoting the electron shuttling between the electrode surface and redox center of enzyme whereas after modification of bare electrode surface with Fe 3 O 4 NPs the resistance to electron flow (R ct : 63 Ω curve b) was even greater than bare Figure 5.10: Electrochemical impedance spectra showing the Faradaic component (Z ) against the capacitive component (Z ) for bare gold electrode (curve a), Fe 3 O 4 NPs modified gold electrode (curve b), GOx-Fe 3 O 4 NPs-gold electrode (curve c), AuNPs modified gold electrode (curve d), and the GOx-AuNPs-gold electrode (curve e) in solution containing 1mMK 3 [Fe(CN) 6 ] in 0.1 M phosphate buffer 102

20 electrode due to semiconductor character. R ct value was further increased to 115 Ω after immobilization of GOx (curve c). The increased R ct after GOx immobilization in both cases (curve c and curve e) confirms the successful immobilization of GOx to the electrode acting as a barrier for the electron transfer Current Response Current response of the fabricated biosensors as a function of glucose concentration at an applied voltage of 0.33V is shown in Figure 5.11A. The applied potential of 0.33V was chosen based on the cyclic voltammetric studies at 100mV/s discussed above. Figure 5.11: Amperometric response of (A) GOx-Fe 3 O 4 NPs-gold electrode and (B) GOx-AuNPs-gold electrode, to different concentrations of glucose at 0.35V in ph 7.4 phosphate buffer solution The biosensor prepared by immobilization of GOx onto the Fe 3 O 4 NPs modified gold electrode showed 1.7 times lower current response in comparison to GOx-AuNPs-gold electrode. This could be attributed to the higher impedance of Fe 3 O 4 NPs as compared to AuNPs as discussed above. The calibration plot (Figure 5.11A) showed a linear behavior over 1 M to 7mM glucose concentration for GOx-AuNPs-gold biosensor whereas 5 M to 5mM for GOx-Fe 3 O 4 NPs-gold biosensor beyond which nonlinearity set-in (Figure 5.11B) Detection Limit The lower detection limit (in A range) of the fabricated biosensors was recorded as the glucose concentration that can be recorded accurately with an acceptable Signal-to-Noise ratio 103

21 is found to be 5 M for Fe 3 O 4 NPs modified electrodes as opposed to 1 M in case of AuNPs modified electrode Sensitivity Sensitivity for the GOx-AuNPs-gold electrode is μa mm -1 cm -2 as compared to 14.3 μa mm -1 cm -2 GOx-Fe 3 O 4 NPs-gold electrode. We feel that observed increased sensitivity and current response is a result of biocompatible nature of amino functionalized AuNPs and the improved conductivity or lower impedance. The biocompatibility aspect was analyzed in terms of K m that serves as an indicator of the enzyme substrate kinetics. The nonlinear curve fitting of the current response curves using Michaelis Menten equation, I = I max C K m,app + C yields the value of K m. Where, I and I max represent the current at a given glucose concentration (C) and maximum current respectively, and K m,app is the apparent Michaelis- Menten constant. K m,app was calculated to be 6.54 mm for Fe 3 O 4 NPs modified electrode and 3.8 mm for AuNPs modified electrode. Low value of K m,app and hence enhanced enzymatic activity could be attributed to cumulative effect of two forces - biocompatible aspect of amino functionalized AuNPs and better size compatibility with the enzyme. The size of GOx being 5-6 nm while AuNPs are of ~ 20 nm and Fe 3 O 4 NPs are ~ 70 nm, hence the maximum retention of 3D conformation of GOx in case of AuNPs as opposed to much larger Fe 3 O 4 NPs Reproducibility and Precision The precision of the biosensor was evaluated by 10 continuous measurements of 1mM glucose solution at ph 6.5. The experimental values showed a low RSD value of 1.43% for GOx-AuNPs-gold and 1.76 % for GOx-Fe 3 O 4 NPs-gold biosensor indicating that GOx is firmly retained onto the electrode surface even after several repeated use of the electrode Interference Ascorbic acid and uric acid present in blood are the known interferants in blood glucose sensing. However, other interferants like maltose, galactose and xylose can be present in high concentrations in blood as a result of therapies given by healthcare professionals. At a low applied working potential of 0.35V, the fabricated biosensor showed no interference from the above mentioned interferants at their normal physiological concentration ranges in blood. 104

22 Stability The stability of fabricated biosensors was recorded for 60 days. Figure 5.12 shows the current response at room temperature for 1.0 mm glucose (storage at 4 C when not in use). Negligible loss in activity wasobserved during the first 10 days. After 60 days, the response current was found to be 73% with GOx-AuNPs-gold and 56 % with GOx-Fe 3 O 4 NPs-gold biosensor. This long-term stability of the biosensor can be credited to excellent microenvironment provided by nanoparticles surface resulting in reduced leakage and denaturation of immobilized enzyme. Figure 5.12: Storage stability of the fabricated glucose biosensor Response Time It was observed that with increasing concentration of glucose, the current increases and 95% of the steady state response was achieved in less than 4 sec and 8 sec for GOx-AuNPs-gold and GOx-Fe 3 O 4 NPs-gold electrodes respectively. The low response time (in seconds) of the fabricated biosensors as compared to few minutes for membrane or gel based biosensors, confirms good electro-catalytic property of the nanoparticles based biosensor. The reduced size of nanoparticles minimizes diffusional limitation and ensures a rapid electron transfer between the active sites of immobilized GOx and electrode surface. 5.4 Conclusion In the end, we would like to stress that industrial applications exploiting catalytic capabilities of enzymes are highly dependent on method and matrix for immobilization. With ever 105

23 increasing industrial needs, the research publications improvising the method of immobilization or designing newer matrices are also increasing dramatically. After a careful comparative analysis of different immobilization studies done by us it was observed that the minimum denaturation of enzyme in case of electrostatic immobilization is an attractive feature, however covalent interactions of enzyme on nanoparticle surface offers the much desired industrial need of storage and operational stability to the biosensor. Moreover, covalent immobilization onto suitably functionalized nanoparticles, depending on properties of the chosen enzyme in general and specially that of active site residues [102], results in optimum immobilization and maximization of catalytic activity. In other words, synergistic effect of appropriate functionalization and size of nanoparticles promotes thermal stability as well as operational stability and high sensitivity of the biosensor. Nonetheless, the electrostatic or covalent method of immobilization remains a commercial decision. 106