2.4.3. FTIR Analysis

The infrared spectra were prepared using the KBr pellet technique, by thoroughly mixing 3 μL of a particle suspension with 0.2 g of KBr and pressing at 5 tonf using a hydraulic press (Carver® Inc., Wabash, IN, USA). The samples were dried in a desiccator overnight and analyzed by the SpectrumTM One FTIR Spectrometer (Perkin Elmer, Waltham, MA, USA) at room temperature in the 4000–400 cm<sup>−</sup><sup>1</sup> range at an operation number of 20 scans, a resolution of 4.0 cm<sup>−</sup>1, and a scanning interval of 1 cm<sup>−</sup>1.

#### 2.4.4. Particle Counter Analysis

Particle concentration was measured using a particle counter (Spectrex Corp., Redwood, CA, USA) in a round-shaped 150 mL transparent glass bottle with a wall thickness of 2 mm. A total of 10 μL of the sample was added to the bottle with 99 mL of water (HPLC grade, Bio-Lab Ltd., Jerusalem, Israel) under continuous stirring. Particle counting was performed with a laser diode at a wavelength of 650 nm.

#### 2.4.5. Dynamic Light Scattering (DLS) Analysis and Zeta-Potential Measurements

The DLS analysis and zeta-potential measurements were performed using a Litesizer 500 type BM10 instrument (Anton Paar GmbH, Graz, Austria) at 25 ◦C. For measurement of hydrodynamic diameters, the samples were diluted to 1:150, 1:300, and 1:600 with HPLC-grade water, placed into a semi-micro quartz cell, and analyzed using a laser at a wavelength of 660 nm and a side scatter of 90◦. Zeta-potential was measured in diluted colloidal solutions at a particle concentration of 1.33 × 10<sup>4</sup> mL−1, which was determined as described in Section 2.4.4. The solutions were injected into an omega-shaped cuvette and analyzed at an operating voltage of 200 V.

#### 2.4.6. X-ray Diffraction (XRD) Analysis

The phase composition of synthesized particles was studied by XRD analysis using a Rigaku SmartLab SE X-ray powder diffractometer with Cu K α radiation (λ = 0.154 nm) for phase identification. Full-pattern identification was carried out by a SmartLab Studio II software package, version 4.2.44.0 from the Rigaku Corporation (Tokyo, Japan). Materials identification and analysis were performed by the ICDD base PDF-2 Release 2019 (Powder Diffraction File, ver. 2.1901). XRD patterns were obtained using 40 kV, 30 mA by Θ/2 Θ (Bragg-Brentano geometry) in the 2 Θ range of 10–90◦ (step size 0.03◦ and speed 4◦/min). The crystallite size was calculated using quantitative analysis based on the Halder–Wagner method, with the help of the program Powder XRD plugin of SmartLab Studio II x64 v4.2.44.0.

#### *2.5. Assay of Enzyme-Like Activities of the Synthesized HCFs in Solution*

PO-like activity of the HCFs was measured by the colorimetric method, with *o*dianisidine and ABTS as chromogenic substrates in the presence of H2O2. One unit (U) of PO-like activity was defined as the amount of HCF releasing 1 μmol H2O2 per 1 min at 30 ◦C under standard assay conditions. To estimate special enzyme-like activity (U/mg), the HCFs were dried. The tested solution/suspension was prepared by weighing the solid substance and adding water until the needed concentration was obtained.

The assay of PO-like activity with *o*-dianisidine: 10 μL of the aqueous suspension of HCF (1 mg mL−1) was incubated in a glass tube with 1 mL of 0.17 mM o-dianisidine in water (as a control), and with the same substrate in the presence of 8.8 mM H2O2 (as a substrate for PO). The addition of NPs to the substrate stimulated the development of an orange color over time, indicating an enzymatic reaction. The enzyme-mimetic activity could be assessed qualitatively with the naked eye and was measured quantitatively with a spectrophotometer. After incubation for an exact time (1–10 min) at 30 ◦C, and upon the appearance of the orange color, the reaction was stopped by the addition of 0.26 mL 12 M HCl. The generated color was determined at 525 nm using a spectrophotometer. The millimolar extinction coefficient (ε) of the resulting pink dye in the acidic solution was 13.38 mM−1·cm<sup>−</sup>1.

The assay of PO-like activity with ABTS: 10 μL of the aqueous suspension of HCF was incubated in a 1 mL quartz cuvette with 1 mM ABTS in water (as a chromogenic substrate for oxidase), and with the same substrate in the presence of 12 mM H2O2 (as a substrate for PO-like HCF). The addition of HCF to the corresponding substrate (ABTS for oxidase-like HCF, ABTS with H2O2 for PO-like HCF) stimulated the development of a green color over time, indicating an enzymatic reaction. The enzyme-mimetic activity could be assessed with the naked eye and was measured quantitatively with a spectrophotometer. The speed of appearance of a green color was monitored at 420 nm over time using a spectrophotometer, thus enabling calculation of the enzyme-like activity. The coefficient ε of the resulting green dye was 36.0 mM−1·cm<sup>−</sup>1.

### *2.6. Sensor Evaluation*

#### 2.6.1. Apparatus and Measurements

The amperometric sensors were evaluated using constant–potential amperometry in a three-electrode configuration with an Ag/AgCl/KCl (3 M) reference electrode, a Ptwire counter electrode, and a working graphite electrode. Graphite rods (type RW001, 3.05 mm diameter) from Ringsdorff Werke (Bonn, Germany) were sealed in glass tubes using epoxy glue for disk electrode formation. Before sensor preparation, the graphite electrode (GE) was polished on emery paper and on a polishing cloth using decreasing sizes of alumina paste (Leco, Germany). The polished electrodes were rinsed with water in an ultrasonic bath.

Amperometric measurements were carried out using a potentiostat CHI 1200 A (IJ Cambria Scientific, Burry Port, UK) connected to a personal computer, performed in a batch mode under continuous stirring in an electrochemical cell with a 20 mL volume at 25 ◦C.

All experiments were carried out in triplicate trials. Analytical characteristics of the proposed electrodes were statistically processed using the OriginPro 8.5 software. Error bars represent the standard error derived from three independent measurements. Calculation of the apparent Michaelis–Menten constants (*KMapp*) was performed automatically by this program according to the Lineweaver–Burk equation.

#### 2.6.2. Immobilization of HCFs and the Enzyme onto Electrodes

The HCFs and enzymes were immobilized on the GEs using the physical adsorption method.

For the development of the HCF or PO-based electrode, 5 μL of HCF or 5 μL of enzyme solution was dropped onto the surface of bulk GEs. After drying for 10 min at room temperature, the layer of HCF or enzyme on the electrode was covered with 10 μL of Nafion. The modified electrodes were rinsed with corresponding buffers and kept in these buffers at 4 ◦C until used.

To fabricate the glucose oxidase (GO)-based biosensor, 8 μL of GO solution (5 U/mL) was dropped onto the dried surface of the gCuHCF-modified GE. The dried composite was covered by a Nafion membrane. The coated bioelectrode was rinsed with water and stored in 50 mM phosphate buffer, pH 6.0, until used.

#### **3. Results and Discussion**

#### *3.1. gHCFs-Modified Electrodes for Hydrogen Peroxide Sensing*

According to the literature, chemically synthesized HCFs (chHCFs) of Fe (III), Mn (II) and Cu (II) demonstrate significant PO-like activity in solution and on electrodes [13–16,29,31]. In the current work, several gHCFs were obtained via Fc*b*2 from the corresponding salts (Fe, Cu, Pd, Ce, Mn, et al.) and from K4Fe(CN)6, a product of K3Fe(CN)6 reduction by L-lactate in the presence of an enzyme (Figure 1). Our first task was to screen the obtained gPBAs for their sensitivity to H2O2 on amperometric graphite electrodes (GEs) and to select the best compounds as PO mimetics. For this purpose, the optimal conditions for the amperometric experiments were investigated. The amperometric characteristics of the control GE (not modified with gHCF) as a chemosensor for H2O2 were tested using cyclic voltammetry (CV) analysis. Selection of the optimal pH, working potential and scan rate was carried out according to the CV results (data not shown).

Under the experimentally chosen optimal conditions (50 mM NaOAc buffer, pH 4.5 and −50 mV as the working potential), numerous electrodes modified with the synthesized HCFs were screened for their ability to decompose hydrogen peroxide. A low working potential is necessary in order to avoid the effect of possible interfering substances on the electrode's response in the presence of oxygen. This requirement is relevant for the construction of biosensors and their exploitation for the analysis of real samples (food products, biological liquids, and others).

The electrocatalytic activities of the synthesized HCFs immobilized on the surface of GEs were tested by CV and chronoamperometry, as described in Section 2.6.1. The amperometric responses of different HCF/GEs to the added H2O2 were compared. Following the chronoamperograms, calibration curves were plotted for H2O2 determination by the developed electrodes (Figure 2 and Figure S1). The linear ranges and sensitivities of the electrodes modified with HCF were calculated. The analytical characteristics of the developed HCF/GEs, as deduced from the graphs (Figure 2 and Figure S1), are summarized in Table 1.

Modification of GEs with the gHCFs improved the efficiency of electron transfer due to the increase in the electrochemically accessible electrode surface area. It is worth mentioning that in comparison to native PO, several gHCF/GEs displayed higher current responses (*Imax*) to H2O2 at substrate saturation and higher sensitivities (Table 1). The enhancement of current outputs and sensitivities of the electrodes modified with other gHCFs were insignificant. Thus, gCuHCF, gFeHCF, gPdHCF and gCeHCF, when immobilized on graphite electrodes, demonstrated higher PO-like activities in comparison with other gHCFs, as well as with native PO and chemically synthesized chCuHCF (Table 1, Figure 2 and Figure S1). For the most effective electrode (gCuHCF/GE), the current response (*Imax*) to H2O2 at substrate saturation was five-fold higher, and the sensitivity was 29-fold higher than those of the PO/GE (Table 1).

**Figure 2.** Amperometric characteristics of the modified electrodes: chronoamperograms (**a**), dependences of the response on increasing concentrations of H2O2 (**b**), and calibration graphs (**c**) for PO/GE (1), gFeHCF/GE (2), and gCuHCF/GE (3). Conditions: working potential −50 mV versus Ag/AgCl (reference electrode), 50 mM NaOAc buffer, pH 4.5 at 23 ◦C.


**Table 1.** Comparative analytical characteristics of HCFs as artificial peroxidases on graphite electrodes.

As seen, the results presented in Table 1 supported the gCuHCF/GE as the most effective PO mimetic. It was therefore studied in more detail.

Many of the reported H2O2-sensitive PBA-based sensors have sensitivities similar to the developed gCuHCF/GE sensor (1620 A <sup>M</sup>−1m−2) [40]. For example, a PBmodified glassy carbon electrode (GCE) demonstrated sensitivity of 2000 A M−1m−<sup>2</sup> [51], MnPBA/GCE—1472 A M−1m−<sup>2</sup> [37]. Graphite-paste electrodes, modified with Ni-FePBA and Cu-FePBA, showed sensitivities of 1130 and 2030 A <sup>M</sup>−1m−2, respectively [52]. Diamond-boron doped (DBD) electrodes, modified with PB and Ni-FePBA, demonstrated sensitivities of 2100 and 1500 A <sup>M</sup>−1m−2, respectively [53].

Other H2O2-sensitive sensors that contain PBA, coupled with other nanomaterials (carbon, graphene, natural polysaccharides, or synthetic polymers), demonstrated significantly higher sensitivities (from 3–5-fold [16,27,29,32–34] up to 300-fold [54]) compared with the gCuHCF/GE. The main peculiarities of the described sensors were high stability, sensitivity, and selectivity towards H2O2 in extra-wide linear ranges. These properties led to the successful use of the PBAs in oxidase-based biosensors [29–33,35,36,40,51,54–56].

The results obtained by us indicated that the gCuHCF and other gHCFs may have a potential for use as PO-like composites for the construction of amperometric oxidasebased biosensors.

#### *3.2. Study of Structure, Morphology, and Size of the gCuHCF Composite*

The size, morphology, and composition of any materials, especially of NPs, are considered as their basic parameters. A number of noninvasive label-free methods were developed for the characterization of different materials: scanning electron microscopy (SEM), transmission electron microscopy (TEM), dynamic light scattering (DLS), Fourier transform infrared spectroscopy (FTIR), X-ray diffraction (XRD) analysis, Raman spectroscopy, atomic force microscopy (AFM) and other approaches. FTIR spectroscopy allows rapid acquisition of a biochemical fingerprint of the sample under investigation, giving information on its main biomolecule content. DLS allows the rapid determination of diffusion coefficients and also provides information on relaxation time distribution for the macromolecular components of complex systems and their hydrodynamic diameters. XRD provides information regarding the crystallographic structure of a material based on incident X-ray irradiation of the material and measurement of scattering angles and intensities of X-rays leaving the sample. SEM produces images of a sample by scanning the surface with a focused beam of electrons and gives information about the surface topography and composition of the sample. The diversity and ambiguity of green-synthesized materials necessitate the use of multiple techniques for valid characterizations. In our study, the synthesized catalytically active organic-inorganic composite gCuHCF was examined using FTIR, DLS, XRD and SEM (see Section 2.4).

### 3.2.1. FTIR Characterization

The FTIR spectrum of the sample is presented in Figure S2. The FTIR spectrum was described in detail in our previous work [40]; it demonstrates the presence of the following groups: O-H, N-H, C≡N, C-H, C-O, C-N, Fe-C≡N and H2O-Cu-CN. Hydroxyl groups were identified by the bands at 3456 and 3050 cm<sup>−</sup>1, which are related to O-H stretching vibrations; and at 1398 cm<sup>−</sup>1, which corresponds to O-H bending [57]. Amine groups were determined by the bands at 3437 and 2994 cm<sup>−</sup><sup>1</sup> (primary amine stretching) and at 1638 cm<sup>−</sup><sup>1</sup> (assigned to N-H bending) [57]. The 2875 cm<sup>−</sup><sup>1</sup> band was attributed to C-H stretching; the 1476, 790 and 719 cm<sup>−</sup><sup>1</sup> bands corresponded to C-H bending [57]. The signals at 1122 and 1109 cm<sup>−</sup><sup>1</sup> can be explained by stretching vibrations of C-O and C-N groups, respectively [57]. The presence of C≡N groups was confirmed by the band at 2105 cm<sup>−</sup>1, reflecting stretching vibrations of this group [58]. The bands in the fingerprint region in the 509–667 cm<sup>−</sup><sup>1</sup> range can be related to Fe-CN linear bending, and the band at 468 cm<sup>−</sup><sup>1</sup> to Fe-C stretching [58]. The 2010 cm<sup>−</sup><sup>1</sup> band indicates the presence of a H2O-Cu-CN moiety [58]. The results of FTIR showed the presence of copper cyanoferrate particles enveloped by an organic layer with hydroxyl and amine groups, probably of protein origin.

### 3.2.2. DLS Studies

The main results of the DLS measurements were described in detail in our previous work [40]. The DLS demonstrated heterogeneous mean hydrodynamic diameters of the particles in a tested gCuHCF. It is worth mentioning that very large differences in hydrodynamic diameters were found for various dilutions of the sample. In the most concentrated sample, only one size fraction was detected. There were probably larger agglomerates of particles in the concentrated suspensions that could not be measured by the designated instrument since the upper limit of measurement was 10,000 nm. After dilutions under gentle agitation, large aggregates disintegrated, and two fractions of particles were obtained.

In concentrated suspensions, the hydrodynamic diameter in the smaller particle fraction was 445 nm, whereas after dilution of the sample, two fractions were detected. The polydispersity index exceeded 10% for all dilutions. This result proved that the tested sample was not monodispersed. The zeta-potential was negative, estimated as −20.9 mV. This value characterizes the suspension state of gCuHCF as the threshold of delicate dispersion.

Particle concentration and mean size of the gCuHCF fraction, estimated by the particle counter, were 2.00 × 10<sup>6</sup> mL−<sup>1</sup> and 3.04 ± 1.98 μm, respectively.

#### 3.2.3. X-ray Diffraction (XRD) Analysis

The XRD pattern of the particles is shown in Figure S3. Diffraction peak positions and their relative intensities reflect the cubic crystalline structure of gCuHCF. Parameters of the crystal cell were calculated from the XRD pattern data (Table 2). The crystal cell belongs to a cubic type with the parameter a = 7.071 Å. Crystallite size was estimated as 156 ± 13 Å.

**Table 2.** Crystal cell parameters of gCuHCF.

