Interaction Study of Anti‐E. coli Immobilization on 1DZnO at Nanoscale for Optical Biosensing Application

Developing low‐cost biosensing platforms for robust detection response and sensitivity at low concentrations is of great interest. This work reports synthesizing 1D ZnO nanostructured materials (1DZnO) with controllable properties utilizing a metal catalyst‐assisted vapor phase growth technique (VLS). The obtained materials are functionalized with (3‐aminopropyl) trimetoxysilane (APTMS) and immobilized with anti‐Escherichia coli enteropathogenic (EPEC) antibodies. Characterization results show changes in the optical and structural properties of 1DZnO that are correlated with the biofunctionalization methodologies. Further, the biofunctionalization process is assessed on 1DZnO surface platforms to obtain acceptable antibody immobilization efficiencies (52%, 96%, and 100%) using a low‐concentration antibody solution (30 µg ml−1). Special techniques such as focused ion beam micromachining and scanning tunneling electron microscopy are proposed to appreciate the semiconductor biofunctionalization layer around 1DZnO and explain the physics of the interaction process. It is found that morphology obtained from distinct synthesis methods, solvents, and functionalization agents can generate functional groups for biomolecule attachment. Remarkably, it is demonstrated that biofunctionalization on 1DZnO takes place all over a single nanostructure. This work presents a proof‐of‐concept focused on generating pathogen sensing platforms using 1DznO semiconducting materials, providing new insights into bio‐analytes interaction with structures at the nanoscale.


Introduction
The standard techniques for detecting pathogens associated with acute diarrheic diseases such as Escherichia coli or Salmonella are generally performed with biochemical and molecular methodologies, providing advantages such as high sensitivity and accuracy. However, the actual pandemic SARS-CoV-2 has established the need to develop new ergonomic, reliable, and easy-to-use technologies that can overcome the high demand for rapid identification of infectious agents. [1] Although molecular methods are the golden standard, the times for sample processing, the use of expensive equipment, reactive, and the need for qualified personnel are limited resources that become a constant bottleneck when the identification demand is rapidly increasing. Biosensors are selective and sensitive devices for detecting biological analytes at low concentrations. Therefore, developing more efficient biosensing devices with improved mechanisms for detection are alternatives that contribute substantially to identifying current and emergent pathogens in the medical, food, water, and environmental sectors. Although there are still limitations for emerging biosensors, they have a reasonable projection for industrial and high-scale processing with suitable detection limits, and their improvement has become the aim of intense research. [2][3][4][5][6] On the other hand, innovations in growth techniques of ZnO-based nanostructures have been a milestone for developing biosensing platforms. The possibility of obtaining nanostructured materials with different morphological features, adaptive physicochemical properties, and high surface-tovolume ratios plays a crucial role in the effective interaction with various biological analytes. 1D ZnO nanostructured materials (1DZnO) have characteristic features that make them excellent candidates for biosensing applications. For instance, ZnO's non-center symmetric structure allows obtaining an extensive collection of nano-morphologies from this compound, such as nanowires, nanorods, and nanotubes, reported for producing sensitive and selective biosensors. [7][8][9][10][11][12][13] These morphologies offer increased surface areas compared to the Developing low-cost biosensing platforms for robust detection response and sensitivity at low concentrations is of great interest. This work reports synthesizing 1D ZnO nanostructured materials (1DZnO) with controllable properties utilizing a metal catalyst-assisted vapor phase growth technique (VLS). The obtained materials are functionalized with (3-aminopropyl) trimetoxysilane (APTMS) and immobilized with anti-Escherichia coli enteropathogenic (EPEC) antibodies. Characterization results show changes in the optical and structural properties of 1DZnO that are correlated with the biofunctionalization methodologies. Further, the biofunctionalization process is assessed on 1DZnO surface platforms to obtain acceptable antibody immobilization efficiencies (52%, 96%, and 100%) using a low-concentration antibody solution (30 µg ml −1 ). Special techniques such as focused ion beam micromachining and scanning tunneling electron microscopy are proposed to appreciate the semiconductor biofunctionalization layer around 1DZnO and explain the physics of the interaction process. It is found that morphology obtained from distinct synthesis methods, solvents, and functionalization agents can generate functional groups for biomolecule attachment. Remarkably, it is demonstrated that biofunctionalization on 1DZnO takes place all over a single nanostructure. This work presents a proof-of-concept focused on generating pathogen sensing platforms using 1DznO semiconducting materials, providing new insights into bio-analytes interaction with structures at the nanoscale.
bulk counterpart, improving their biosensing performance. In addition, its high isoelectric point provides good selectivity for the attachment of biomolecules to the surface. Other properties such as the optical response, piezoelectricity, low cost, and biocompatibility with the human body have allowed their use in bioapplications such as antimicrobial coatings, drug delivery, and imaging. [14] The use of 1DZnO as active elements in sensors derives from the need to achieve lower-threshold detection levels. For instance, surface plasmon resonance phenomena and characteristic photoluminescence reported in 1DZnO metal composites have turned these optical systems into promising biosensing prospects. [15] Yet, 1DZnO features such as orientation, size, structure, density, and chemical surface modification must be considered to establish efficient biomolecule immobilization methodologies.
To illustrate the relevance of 1DZnO materials (nanorods/ nanowires) towards the development of optical biosensors, a literature search was carried out using the Web of Science database with the keywords: biosensor, ZnO, and nano ZnO. The obtained results are depicted in Figure 1a. This graph shows the great interest from the scientific community in developing materials with bio-transductive properties and the limited implementation of ZnO-based nanomaterials for high-impact disease detection. As these materials offer tunable surface and optical properties, they could be implemented in microfluidic devices and lab-on-chip technology, opening a wide range of opportunities when functionalized with different biological recognition elements.
Strategies to develop efficient biosensor devices are diverse; hence, the fabrication process can be optimized at different levels. Many research works have tried to increase biosensor sensitivity through the site direction of antibodies and/or the use of labels for the case of immunosensors (i.e., virus, bacteria, and cancer cell detection). [16][17][18][19][20][21][22] Other approaches search binding matrices compatible with susceptible signal responses (i.e., gold and platinum electrodes). [23,24]

www.advmatinterfaces.de
Despite there are different methodologies for the binding of bifunctional molecules, the use of silane as a coupling agent has been a reproducible and widely accepted method for the generation of sensor surfaces because it offers advantages such as biomolecule stability, high immobilization density, reduced biofouling, and leach-proof biomolecule binding. [12,[25][26][27] Strategies for functionalization of ZnO platforms are mainly based on silane treatments with (3-mercaptopropyl)trimetoxysilane (MTS), (3-aminopropyl)triethoxysilane (APTES), and (3-aminopropyl)trimetoxysilane (APTMS), among others. [25,26,28,29] These can be applied to different ZnO nanostructured scaffolds such as nanocomposites, nanoparticles, nanorods, thin film transistors (TFTs), and nanowires. [8,[29][30][31][32] In some cases, the morphological correlation of ZnO surfaces with the antibody immobilization efficiency has not been completely elucidated. Currently, there are diverse configurations to detect foodborne pathogens using different biosensing platforms. However, although many works achieve the identification of bacterial infectious agents such as E. coli [5,[33][34][35] and Salmonella, [31,36,37] for the development of nanomaterials-based biosensors, the research, and optimization of the biofunctionalization strategies (reagents, concentrations, contact times, among others) have not been thoroughly studied. These parameters can affect the biosensing platform by reducing or increasing the immobilization efficiency of the biorecognition agent. This needs to be evaluated by experimental studies to analyze morphology interaction, surface chemical modification, and how these together can affect the response signals from nanomaterials. Therefore, the comprehension of antibody immobilization mechanisms on ZnO nanostructured morphologies could directly impact the enhancement of biosensors by improving surface chemical modification. It could help to achieve homogeneous immobilization of biorecognition elements without crosslinkers and labels. In addition, using low antibody concentrations is another crucial parameter to optimize for nano-biosensors that provide a single and accurate response. This work evaluates biofunctionalization strategies of 1DZnO-based nanomaterials with different morphological features for their application in optical responsive biosensing platforms. Moreover, as far as the authors know, for the first time, a complete characterization of the biofunctionalization process, including analytic (Bradford protein assays) and advanced structural techniques (focused ion beam [FIB] micromachining and scanning tunneling electron microscopy [STEM] compositional mapping), was performed to confirm the antibody immobilization up to a single ZnO nanostructure. The present work shows the importance of the diverse morphologies of 1DZnO for an improved immobilization strategy and the use of different solvents but also explains the interaction mechanism using structural and optical in-depth analysis that could help researchers from multidisciplinary branches working with biosensing applications.

1DZnO Morphological, Structural, and Optical Characterization
The differences between the preparation of substrates (presence or absence of seed layer) and the synthesis conditions (fluxes and substrate position) allowed the fabrication of ZnO nanostructured platforms with different morphologies. The scanning electron micrographs (SEM) of the synthesized nanomaterials are presented in (Figure 2), where we can observe the three morphologies studied in this work. (Figure 2a-c) shows the formation of both 1DZnO nanostructures (i.e., nanowires) and 2D structures (i.e., wall-like structures). We use the label 1DZnO_1 to refer to this morphology, which was synthesized using substrates without a ZnO seed layer, that is, on Si substrates directly coated with an Au catalyst. The flow rate applied was 36 sccm of Ar:O 2 mixture as the carrier gas. Figure 2c shows that the tip of 1DZnO structures present a contrast concerning the backbone matrix. In previous structural studies, we have shown that this is due to the formation of an Au-ZnO growth interface at the tip, which represents strong evidence that the growth mechanism of 1D-nanostructures is assisted by liquid metal catalyst (vapor-liquid-solid: VLS). [38] On the other hand, the growth of 2D (wall-like) structures is evidence of a vapor-solid-solid (VSS) growth mechanism without the participation of the metal catalyst. Hence, the growth in samples where we did not employ a seed layer is a combination of VLS and VSS mechanisms. Although there is high tension in the growth interface between the substrate and the nanostructured ZnO matrix due to the structural differences (Si cubic, ZnO hexagonal structures), the growth conditions of the reactor allowed us to obtain a homogeneous coating on the entire surface substrate even without the use of seed layer. The length and diameter of these samples were 8.1 ± 2.7 µm and 136 ± 11 nm, respectively. Figure 2d-f shows the morphology of the 1DZnO_2 sample. For this case, silicon substrates coated with ZnO seed layer films and Au catalyst film were used. ZnO seed layers with a thickness of 200 ± 10 nm were deposited by magnetron sputtering, obtained under a downstream position with a flux of 36 sccm. A homogeneous coating was observed over the entire substrate surface composed of ZnO nanostructures perpendicularly aligned to the surface. It is also appreciated that the length of the structure is smaller (0.8 ± 0.1 µm) compared to the other systems.
We attribute this morphology to a combination between the working pressure and flux employed, which propitiates conditions where the gaseous precursors are supplied to the growth interface in a controlled manner at a stable growth rate. (Figure 2g,h) shows the last morphology employed, 1DZnO_3. In this case, seed-layer substrates were used downstream at a drag flux of 72 sccm. Due to the increased drag flow, the rate at which the precursors were delivered to the substrate surface was considerably increased. Therefore, the nanostructures obtained were more extended than in the other cases (8.3 ± 3.4) µm. Besides, from SEM micrographs, it is appreciated that a large amount of catalyst nanoparticles did not participate in the growth of the ZnO matrix as they are supported by the seed layer and not by axial ZnO structures (Figure 2h). Figure 2i presents a histogram comparison between the length and diameter distributions present in the synthesized samples. The analysis was performed using the Image J software to analyze micrographs of different sample areas using log-normal distributions. These graphs allowed us to detect slight variations in the diameters obtained for the three morphologies (130-140) nm. This is because the same type of Au catalyst film was used in all www.advmatinterfaces.de cases, which consequently determines the radial confinement of ZnO nanostructures. Table 1 summarizes the morphological features obtained for each synthesis condition employed.
The X-ray diffractograms obtained in a grazing angle configuration (GIXRD) are presented in Figure 3a. The observed changes in the morphology of 1DZnO impact the nature of the X-ray diffraction pattern (ICSD-193696 and ICSD-611625) for ZnO and Au, respectively. It was possible to identify the hexagonal phase of ZnO (wurtzite) and the cubic phase of Au. For the case of 1DZnO_1, where there is a random orientation of structures and the presence of walls, a polycrystalline behavior is observed in the diffractograms, and the peak of maximum intensity was associated with the family of planes (100). On the other hand, for the 1DZnO_2 sample, high intensity in the plane reflection (103) was observed. In this case, given the measurement configuration, we have substantiated that the appearance of this peak corresponds to the preferential orientation of the (002) plane, and therefore, the presence of structures is highly aligned with each other and perpendicular to the substrate plane. For a more detailed description of the diffraction conditions for this kind of system, please refer to our recent publication. [39] For the 1DZnO_3 sample, we also have a polycrystalline behavior where the reflection of higher intensity corresponds to the (002) plane. However, as observed in 1DZnO_1, the small signal of the peak (103) suggests that the nanostructures are randomly oriented. This observation is consistent with the images shown in Figure 2g,h. To corroborate the preferential orientation of the samples more rigorously, τ TC texture coefficients were calculated for the three morphologies ( Figure 3b). This graph shows that the morphology with the highest texture coefficient for the plane (103) is 1DZnO_2, the one in which the structures are vertically oriented. In the

www.advmatinterfaces.de
case of 1DZnO_1, the planes with preferential orientation correspond to the present inclinations with respect to the plane (002) and, therefore, could cause growth in random directions. For 1DZnO_3, the preferential orientation (002) could correspond partly to the seed layer orientation and not entirely to the ZnO nanostructures on the surface.
To study the optical properties of the system, we performed photoluminescence (PL) measurements of the synthesized nanostructured platforms. The obtained spectra are shown in Figure 3c. We can observe a clear difference between the 1DZnO_1 spectrum and the spectra from the other two samples. For 1DZnO_1, the appearance of two emission bands can be seen, while only a single band is observed for the other samples. These two emission bands are characteristic of ZnO and are named according to the nature of the radiative recombination that originates them. The first is a band in the near violet-UV region associated with excitonic band-band recombination, called near-band emission (NBE). The second emission band is centered in the spectral region and is attributed to permitted states within the forbidden gap generated by intrinsic defects in the ZnO structure, called the deep-level emission band (DLE). To make an association of the various luminescence centers present in the material, a deconvolution process of the PL spectra was performed using the Fityk software. This way, Gaussian curves were fitted to recover the measured spectra experimentally. Figure 3c shows the fitting curves in dotted lines. We can observe that the luminescence centers of the three morphologies are similar, with minimal shifts. The centers at 2.54-2.57 eV (cyan) and 2.39-2.35 eV (green) are associated with zinc vacancies (V Zn − ) and oxygen vacancies (V O ), respectively. [39] The yellowish-green emission, 2.16-2.21 eV, has been attributed in the literature to a complex between oxygen vacancies and Zn interstitials (V O Zn i ). [40] The orange emission (2.00-2.02 eV) in samples 1DZnO_2 and 1DZnO_3 has been attributed to oxygen interstices (O i ). [41] Finally, in 1DZnO_1, the presence of the NBE band requires the addition of centers in the near-band region, where neutral zinc vacancies (V Zn ) and interstitial Zn (Zn i ) have been related to emissions between 3.11-3.00 eV. [42] On the other hand, the plots presented in Figure 3d show that the peaks shift of both bands is minimal between samples, which confirms that the physical origin of luminescence defect bands is the same and the ratio between their intensities gives a fundamental distinction in the optical signal. The difference can be attributed to www.advmatinterfaces.de the relative contribution of defects in the 1DZnO crystalline structure, where zinc-related radiative recombination with energy closer to the band gap is favored for sample 1DZnO_1 (upstream position). Once the synthesis characterization of 1DZnO morphologies was finished, the spectroscopic analysis was carried out, and details about the bio-functionalization steps are presented in the following section.

Spectroscopic Analysis of the Biofunctionalization Process
As demonstrated in previous works, Fourier transform infrared spectroscopy (FTIR) has allowed following the changes that occur at the material surface after processes that involve chemical modifications. [12,32,43] A Proof-of-concept of antibody immobilization was developed to find the morphology that permits an improved biofunctionalization. Figure 4 shows the FTIR spectroscopic analysis of surface chemical modification for the three 1DZnO samples. In general, the different siloxane signals result from the processes that occur after interacting with APTMS. The siloxane signals (Si-O-Si) and the silicon wafer (Si-O) could show a signal addition. As observed, the hydroxylation and silanization steps influence the existence of characteristic signals. This implies that the interaction with a biorecognition agent could vary due to the availability of bonds generated from the chemical modification. The sample 1DZnO_1 (Figure 4a) showed a wall-like morphology, where two characteristic signals of Zn-O are observed at 411 and 576 cm −1 ( Figure S1a, Supporting Information). This can be explained by the presence of the vibrational mode E1(TO) at 411 cm −1 related to nanowalls and the mode E1(LO) associated with the nanowires that coexist in the sample at 576 cm −1 . [44][45][46] However, it is suggested that due to the presence of these wall-like structures, the generation of functional groups could not be homogeneous, and at the same time, the density of 1DZnO_1 structures is lower compared to the other morphologies. This shows that the hydroxylation

www.advmatinterfaces.de
solution, at the activation step, could have an abrasive effect on the surface, which is observed at the reduction signal of Zn-O (Figure 4a). Since the generated OH − groups are not enough in the sample, the FTIR does not observe the formation of siloxane, suggesting that APTMS binding was not appropriately achieved. Therefore, it was expected that 1DZnO_1 morphology could not present suitable signals for immunoglobulins detection after the immobilization process, as observed in Figure 4a, where only signals from the phosphatebuffered saline (PBS) buffer were detected, as well as remanent solvents from previous treatments with acetone. Figure 4b corresponds to the biofunctionalization strategy of 1DZnO_2 with shorter and homogeneous nanowires. It is observed that the signal is consistent with the Zn-O bond at 549 cm −1 , and after silanization, the signals related to Si-O and Si-O-Si (siloxane) are observed at 1062 and 1108 cm −1 , respectively (Figure S1b, Supporting Information). Some overtones are present in the observable signals, related to the different vibrational forms of APTMS binding. In addition, signals referent to the amine group N-H at 1582 cm −1 and contributions of methyl groups C-H at 2851 and 2920 cm −1 indicates a proper functionalization. [47,48] This could be related to the homogeneous morphology and higher surface area of 1DZnO_2, which allows a better interaction of the chemical treatments on this sample. Finally, the stretching peak O-H at 3337 cm −1 enables the visualization of surface hydroxylation. It is worth mentioning that the hydroxylation solution does not decrease the signal of Zn-O, which can be related to the material morphology and how the hydroxylation solution can interact with the surface. [43] Regarding antibody immobilization, the most important signals identified were the amide 1 stretching C=O at 1693 cm −1 and the amide 2 bending N-H at 1536 cm −1 , implying a better antibody attachment on 1DZnO_2. [49,50] Moreover, the confirmation of immunoglobulins' presence can be observed with the signals related to the disulfide bridges from the IgG hinge region at 520 cm −1 (S-S) and 682 cm −1 (C-S). [32,49] The S-S peak has a slight shift and broad signal, which can be attributed to the proximity between disulfide bridges and the ZnO signal. A similar case was observed for the amide 1 region, where a prominent peak was related to the used solvent. Since both signals (amide 1 and acetone) have the C=O bond, they have a broad but detectable signal due to overtones and the modification of peaks at the region 1000-1200 cm −1 . Hence, it is concluded that 1DZnO_2 morphology could play an essential role in improving biofunctionalization, impacting the antibody binding to the 1DZnO matrix. Figure 4c corresponds to the described morphology of long nanowires without walls (1DZnO_3). An intense signal from the stretching bond Zn-O is observed at 545 cm −1 . However, the functionalization modifies the absorption peak intensity, which could result from the hydroxylation process of nanowires that probably have better contact with the seed layer instead of the 1DZnO support. Therefore, the signal at 1010 cm −1 is primarily related to Si-O ( Figure S1c On the other hand, for the interaction with the recognition element (antibody), the signal related to the siloxane is hindered by the PBS buffer signal, which has activity at the same region. In addition, peaks related to amides 1 and 2 are observed; however, the antibody attachment is weak. The information about functional group comparison from the three 1DZnO morphologies is summarized in Table 2. Based on this, sample 1DZnO_2 showed the best characteristics for immobilization; the higher homogeneity allows a steady development of the chemical reactions for surface modification that generates functional groups. This implies that the ZnO nanowires' orientation, density, and size are features to be considered to establish strategies for antibody immobilization focused on immunosensors.
At the first stage, 1DZnO samples were treated using the same processing times for hydroxylation, silanization, and antibody immobilization, to evaluate the role of morphology on generating sensor surfaces. However, the results suggest that biofunctionalization can be improved for the selected morphology (1DZnO_2) because a prolonged reaction time with APTMS could develop agglomerations, limiting the number of available functional groups for antibody attachment. A polar solvent such as acetone could produce excessive polymerization or multilayer formation that could cause steric hindrance or interactions by hydrogen bonds. Moreover, it is known that the interaction of APTMS with polar solvents (acetone or ethanol) could show spontaneous solvolysis; however, the generated monolayer can be used if the residence times are controlled. [26] Therefore, the enhancement for the second stage of biofunctionalization was developed specifically on 1DZnO_2 samples, considering the reduction of contact times for hydroxylation and silanization. All presented studies were performed using the 1DZnO_2 morphology for the following sections. Figure 5 shows the FTIR analysis of 1DZnO_2 subjected to the new biofunctionalization strategies. Figure 5a presents an increment of the characteristic signal of the Zn-O bond related to the interaction between nanowires and the chemical reactions associated with surface modification. The most important feature is the presence of the Si-O-Si bond (1154 cm −1 ) and N-H signal (1463 cm −1 ), which indicates the binding of the coupling agent APTMS. In addition, signals associated with acetone are observed at 1743 cm −1 (C=O bond), where a constructive addition occurs between methyl groups from acetone and silane. As a result, the peaks at 2854 and 2926 cm −1 show a higher intensity signal. The 1DZnO_2 and Si wafer spectra were included as a control, where it is mainly observed the signal associated with the Si-O bond derived from the substrate (Si) and ZnO. This new strategy has the presence of CH 3 OH in the hydroxylation solution, which increases the dipolar moment and could contribute to the generation of a higher amount of OH − functional groups. Since the reaction between KOH + CH 3 OH from the hydroxylation solution produces a potassium methoxide (CH 3 OK), the contact times were adjusted for 1DZnO_2 to avoid possible abrasive impacts on the surface material. [51] Figure 5b depicts the effect of the improved strategy on the immobilization process under three functionalization conditions with solvents such as toluene (Tol) and acetone for 5 min (Ac5) and 20 min (Ac20).
The characteristic peaks of proteins are observed, and the presence of amides 1 (C=O) and 2 (N-H) indicates the existence of immobilized antibodies. [49,50] Moreover, signals associated www.advmatinterfaces.de with disulfide bonds at 668 cm −1 also confirm the presence of immunoglobulins. The samples Ac5 and Ac20 show similar signals since they use the same solvent (acetone). For the case of sample Tol-APTMS/IgG, observable differences are detected at the disulfide bond peak. Based on these results, it is proved that FTIR is a nondestructive technique for monitoring the antibody presence on 1DZnO_2. In addition, complementary approaches to verify immobilization efficiency and the role of the interaction in the biofunctionalization process are presented.

Atomic Force Microscopy (AFM)
One of the accurate techniques used to follow morphological changes is atomic force microscopy (AFM), which was selected to observe changes at the different stages of biofunctionalization of 1DZnO_2 samples. Figure 6a shows a highly uniform morphology with sharp shapes over the surface.
Muhammad A. K. et al. reported that the average roughness is between 55.282 and 400 nm for 1DZnO nanostructures. [52] This concurs with the 141.1 nm (Ra) value obtained in this work. However, the topography of samples varies in function of the biofunctionalization steps. Figure 6b shows the changes after the interaction with APTMS. Some island-like morphologies are observed, which modify the general form of the sample, and these conformations are related to the APTMS polymerization for the multilayer formation, which has the amine groups available for interaction with IgGs.
Then, Figure 6c corresponds to the antibody immobilization stage, where a surface topography of multilayers with curved drops has been associated with immobilized antibodies on the surface, which is concordant with our results. [12,53] Derived from the different processes on 1DZnO_2, the average roughness increased to 159 nm (Ra). The AFM analysis could represent a simple and practical validation strategy to evaluate the antibody immobilization, which also supports the results obtained by FTIR ( Figure 5).

Photoluminescence Response of Biofunctionalization
As mentioned earlier, biosensing devices focus on identifying easily interpretable response signals. That is the case of fluorescence, infrared, photoluminescence, Raman, and surface plasmon resonance-based optical biosensors. [54] In this work, the optical response of the sensing platform was monitored by room-temperature PL spectroscopy. Figure 7a shows the normalized emission spectra of 1DZnO_2 samples for the different functionalization strategies described in the Experimental Section. It is possible to observe the changes in the radiative recombination of the material when it is functionalized with a bioselective layer of antibodies using different solvents (acetone 5/20 min, toluene 10 min). All the functionalization methodologies produced a redshift of the broadband emission in the visible region for samples without antibody immobilization (Figure 7b). As the contact time with solvents was increased, the red shift was more pronounced (Ac 20). Redshifts in emission have been reported in the work of Bauer et al. for APTMScoated ZnO nanostructures and have been associated with the migration of defects in the ZnO lattice due to silanization at room temperature. [29] In this figure, we have included inset photographs of the sample emissions when irradiated with a low-power UV lamp, and more information is provided in Figure S2, Supporting Information. The changes in emission are visible to the naked eye, highlighting that IgG does not show emission by itself, as it does when immobilized on 1DZnO_2 nanostructures. The samples Tol and Ac5 showed a similar PL behavior, with acetone as a suitable and less toxic solvent for functionalization than toluene. Finally, the PL response of Ac20 (observed at the inset) could be related to a higher contact time with APTMS, which can be derived from a higher molecule agglomeration. The PL measurement related to immunoglobulins over 1DZnO_2 was included as well. In addition to the emission shift, the biofunctionalization procedure also increased the broadness of the emission band, which can be attributed to the production of new energy levels associated with antibody immobilization. Figure 7c presents an increase in the relative signal of the NBE band (400 nm) from 1DZnO_2 subjected to the different functionalization methods. [55] This figure shows an increased NBE signal for samples treated with acetone for a prolonged time (Ac20) compared to those treated with toluene. The changes in the response signal intensity resulting from the biofunctionalization of the ZnO semiconducting surface can be attributed to the effect of hydroxylation on the compound structure. Deficiencies or influence of the oxygen vacancies as a product of the different chemical modifications and  It is also suggested that 1DZnO surface coating with a polymeric APTMS layer can modify the emission response and intensity due to surface passivation or electron injection (APTMS-lowest unoccupied molecular orbital (LUMO) towards the ZnO conduction band, enhancing the radiative recombination efficiency in the NBE region. [56][57][58] Therefore, the functionalization strategy using acetone as solvent is an option to modulate between non-radiative and near-excitonic recombination. As observed, it is vital to correlate the measurable signal responses (visible PL emission) to the chemical surface changes after biofunctionalization (antibody attachment) on 1DZnO_2.
From these results, it should be noted that obtaining a bioselective layer on the surface of the semiconductor matrix does not hinder the optical response of the material. In this way, the ZnO transduction signal can be used to detect interest analytes, which can be applied to generate optical immunosensors that can associate PL variations (peak shift or intensity changes) to the concentration of target analytes. In fact, some already published reports have successfully implemented the nanostructured ZnO photoluminescence response as a transducing signal in immunosensing-based platforms. Some of them are presented in Table 3.
However, direct validation of the immobilization efficiency of antibodies on silanized nanostructured ZnO surfaces is still required. A comprehensive analysis of the interaction mechanism between 1DZnO/APTMS and anti-E. coli antibodies could be significant in developing modern-generation biosensors.

Analysis by Transmission Electron Microscopy (TEM)
Generating highly sensitive biosensors requires constant validation during all the processes for the fabrication of biofunctionalized platforms. In this work, a suggested strategy is using the FIB technique ( Figure S3, Supporting Information) to obtain samples for analysis by TEM/STEM. (Figure 8) corresponds to the micrographs of biofunctionalized Tol samples analyzed by TEM (Figure 8a-e) and their study by energy dispersive spectroscopy (EDS) (Figure 8f-l) in STEM mode. Due to its synthesis nature (VLS), gold is observed over the tip of the nanostructure. The panels (a)-(e) show the presence of APTMS in the form of island-like structures over the complete nanowire. Additionally, panel (e) depicts dust-like forms related to the PBS buffer used for antibody immobilization reaction. This was also verified by the elemental scanning EDS, from which the subsequent antibody elements were detected ( Figure S4, Supporting Information). This technique allowed us to confirm the biofunctionalization even from a single nanowire. It is known from a previous report that the desired condition is the surface binding and condensation of APTMS to achieve horizontal polymerization and siloxane formation. [12] This is correlated to the islandlike structures observed from AFM micrographs (Figure 6b), which confirmed the presence of APTMS.
Moreover, the analysis by EDS and elemental mapping is a tool able to validate undoubtedly the attachment of immobilized antibodies over the 1DZnO_2 surface. Panel (f) shows a general mapping of elements to validate their presence or absence as components of the nanostructures. For instance, panel (g) allows the identification of Zn inside the nanowires. In contrast, the characteristic elements from antibody amino acids from immobilization were observed outside of 1DZnO_2, also identified by the presence of phosphorous (P), carbon (C), and sulfur (S) at panels (h), (i), and (j), respectively. Then, panels (k) and (l) are related to the silicon and oxygen of the different molecules. These results are also correlated to the findings obtained from FTIR ( Figure 5), where characteristic signals from disulfide bridges (S-S) were also observed and used as a corroboration of the immobilization process.
In Figure 9a-e, the presence of APTMS around the 1DZnO_2 Ac5 sample can be better appreciated since the

www.advmatinterfaces.de
formation of a monolayer is consistent, reinforcing the presence of silane. Panel (e) shows more clearly how the polymerization is carried out along the ZnO nanostructure, which is an advantage of using such materials, which provides a broader contact area. This could imply that silane concentration for generating sensor surfaces is an important parameter to consider. The uncontrolled increase in concentration suggests a possible agglomeration and cluster formation that could saturate the nanostructures, causing a decreased availability of functional groups for antibody binding. The corresponding elemental scanning EDS is shown in Figure S5, Supporting Information.
In addition, using acetone as a solvent agent can be an alternative compared to more pollutant solvents, such as toluene, if the reaction time and spontaneous solvolysis are controlled. Figure 9f shows the general mapping of elements in the sample 1DZnO_2 Ac5 from FIB analysis. Panel (g) indicates the presence of sulfur (S) from the antibody hinge region, confirming the antibody immobilization as well. Then, panels (h) and (i) correspond to the Zn and O elements which are components of nanostructures. Similarly, the presence of P from the PBS buffer (panel (j)) and Si as a contribution of APTMS (panel (k)) confirm that biofunctionalization was carried out appropriately over the entire extension of the nanowire. Remarkably, the APTMS binding can be observed around all the nanowires in this figure, implying that this strategy provides higher functional groups available for antibody attachment to generate more homogeneous and reproducible sensor surfaces. Finally, panel (l) shows the presence of Au as the metal catalyst for  VLS growth. It is essential to mention that this strategy allowed us to detect and confirm the presence of a small number of immobilized antibodies over the surface (30 εg ml −1 ). Since a quantitative identification of immobilization is crucial and necessary, the measurement of the attached immunoglobulins on 1DZnO_2 is presented in the following section.

Immobilization Efficiency by Protein Quantification
The results obtained from structural and optical characterization were correlated to antibody attachment quantification after the immobilization step. First, a calibration curve with the bovine serum albumin (BSA) standard was prepared by using the Bradford method ( Figure S6, Supporting Information). The obtained equation was used to correlate the UV-vis absorbance measurement (OD 590nm ) with the protein concentration to calculate immobilization efficiencies (Equation (1)). Antibody immobilization efficiency equation. [32] % de antibodies 100 final stock wash 1 wash 2 wash 3 initial stock 100 The concentration of initial antibody stock (30 εg ml −1 ) and after contact with biofunctionalized 1DZnO_2 (Tol, Ac5, and

www.advmatinterfaces.de
Ac20) was measured, as well as the remanent antibody concentration from washes. Data obtained from protein quantification are presented in Table 4. It is observed that samples 1DZnO_2 Tol and Ac5 showed higher immobilization efficiencies, 96% and 100%, respectively, obtained from the calculation of antibody concentrations from stock and washes after the contact time, which was near zero. It implies that 28.8 εg ml −1 (1DZnO_2 Tol) and 30 εg ml −1 (1DZnO_2 Ac5) antibodies are immobilized on 1DZnO_2 nanostructures, which supports the results presented before by TEM and FIB analysis, also substantiating that acetone is a suitable alternative that provides higher immobilization efficiency with lower contact times. Although the samples functionalized with acetone-APTMS for 20 min (1DZnO_2 Ac20) showed a proper signal intensity observed by FTIR (Figure 5b), it has a lower immobilization efficiency (52.6%). In addition, it is possible to correlate the changes in NBE (intensity) and DLE (shift) bands observed in PL measurements (Figure 7a-c) with the antibody attachment efficiency determined by the Bradford method, thus having an optical signal to monitor the biofunctionalization efficiency indirectly. As previously mentioned, increasing the contact time with the silane can generate agglomerations resulting in higher solvolysis. This suggests that the reaction occurs faster in solvents such as acetone than in toluene. Hence, it is necessary to reduce contact times to provide a functionalization strategy compatible with nanostructured materials. As observed for 1DZnO_2 Ac20, Figure 9. a-e) Biofunctionalization analysis from TEM microscopy of 1DZnO_2 Ac5 samples (dash lines were added as a visual help to distinguish the 1DZnO and APTMS, yellow-APTMS, and red-one single nanostructure with Au tip). f-l) Elemental scanning by EDS in STEM mode.
www.advmatinterfaces.de the availability of functional groups for antibody binding was decreased due to interference from excessive APTMS reaction. ND: Not detectable; this implies that protein concentration is 0.
It is substantiated that the biofunctionalization strategies play a fundamental role in the generation and improvement of sensor surfaces where the control of nanomaterial morphology, silane polymerization, and contact times directly impact the immobilization efficiency of the biological recognition element.

Conclusion and Perspectives
In this work, we reported the analysis of 1DZnO nanostructured materials with different morphological properties for developing biofunctionalized sensor surfaces. The three 1DZnO nanostructures were functionalized with APTMS to subsequently immobilize anti-E. coli EPEC antibodies to select the best for immobilization. At first, the functionalization process was monitored by FTIR, concluding that 1DZnO_2 morphology showed a clear presence of the characteristic signals of amide I (C=O), amide II (N-H), and disulfide bridges (S-S, C-S), which confirmed a proper biofunctionalization. In this work, we have established that nanomaterial structure, orientation, density, and size of 1DZnO are features to be considered to develop strategies for antibody immobilization focused on immunosensors.
Since the present contribution was focused on improving the 1DZnO matrix for developing sensing surfaces compatible with optical immunosensors, two different functionalization solvents (acetone and toluene) and various contact times (5, 10, and 20 min) were tested. The characterization techniques mentioned above were used to confirm the biofunctionalization process to determine structure variations derived from the chemical surface modification. Moreover, a micromachining step through FIB was proposed as an advanced technique to focus the nanometric area to be analyzed. Differences between the use of acetone and toluene were observed mainly from PL, STEM, and FIB analysis. Notably, this strategy allowed direct confirmation of the biofunctionalization process at the nanoscale level highlighting that the biofunctionalization process takes place all over a single nanostructure, supported by complementary techniques.
Regarding photoluminescence, the changes in characteristic PL values and their correlation to the chemical surface modification can be used as a signal response for developing optical biosensors. On the other hand, acetone is a suitable biofunctionalization agent for 1DZnO nanostructures but needs rigid control and reduced contact times (5 min) to achieve proper silane polymerization and functional group availability for immobilization. This was confirmed by performing protein quantification tests, where an immobilization efficiency of 100% for sample 1DZnO_2 Ac5 (acetone 5 min) was obtained using a low antibody concentration (30 µg ml −1 ). Therefore, this work can provide an advantage for the compatibility of these sensing platforms with lab-on-chip and microfluidics technologies.
Currently, there is still a lack of reports that evidence the interaction mechanism between the material (1DZnO) with the biofunctionalization agents and immobilization strategies, which represents an opportunity for the research area to develop future work on the biosensors field. The present research denotes the importance of the high surface areas provided by 1D nanostructures to be applied for antibody immobilization at low concentrations. Determination of the effective surface area, the optimization of biofunctionalization and immobilization steps, and the selection of suitable antibody concentrations for more complex matrixes are still counted for further research (Figure 10) to improve the compatibility of materials at the nanoscale for the development of more sensitive and accurate biosensors.

Experimental Section
1DZnO Sample Preparation and Characterization: 1DZnO nanostructured materials were prepared by a metal catalyst-assisted vapor phase growth technique, which had been reported in previous works. [39,59] Crystalline n-type silicon substrates with <100> orientation were used as supports for these nanostructures. To evaluate the role of ZnO seed layers in the growth of 1DZnO nanostructures, ZnO sputtering coated Si substrates and uncoated Si wafers (pristine) were used as different types of supports. Subsequently, all substrates were coated with 4 nm Au thin films to act as metal catalysts during growth. The synthesis process was carried out in a tube furnace at 950 °C and 30 mbar for 60 min. Inside this tube, the previously prepared substrates were placed together with the powdered precursors. A mixture of ZnO:C (1:1 w/w) powder was used as the precursor material to obtain gaseous Zn species at lower temperatures by a reduction reaction. By controlling the flow rates of the gases (Ar, O 2 ) and the position of the substrates (upstream, downstream), it was possible to control the morphological characteristics of the obtained nanostructures: 1DZnO_1, 1DZnO_2, and 1DZnO_3. SEM images were obtained with JEOL JSM-7600F equipment for morphological characterization. The structural characterization was carried out using the GIXRD technique (Rigaku Ultima IV diffractometer).
Surface Activation: Method 1: To generate (OH − ) groups, the hydroxylation strategy was based on using a solution KOH: H 2 O (1:10 v/v). Afterward, the samples were submerged in this solution for 30 min and then rinsed with deionized water. For the silanization process, a 3% APTMS solution (Sigma Aldrich) dissolved in acetone was prepared, Calculated with an initial stock antibody concentration of 30 εg ml −1 .
Adv. Mater. Interfaces 2023, 10, 2300167 www.advmatinterfaces.de and the samples after the hydroxylation process were submerged inside for 40 min, rinsed with acetone and placed under thermal treatment at 110 °C ± 3° for 90 min. Method 2: To improve the biofunctionalization strategy for the selected morphology (1DZnO_2), the employed parameters were modified in function of the functionalization agents: acetone (Ac) and toluene (Tol), and contact times (5, 10, and 20 min) shown in Table 5. The initial step for all the strategies was a hydroxylation with 0.1M KOH diluted in CH 3 OH, 10 min contact time, and a washing step with CH 3 OH.
Antibody Immobilization: To assess the compatibility of functionalized 1DZnO, the antibody immobilization procedure was developed using polyclonal antibodies of two types: anti-total and anti-flagella from E. coli EPEC. Two methods were used for the immobilization process, where the 1DZnO structures were placed in contact with a solution of 100 εg ml −1 anti-total E. coli EPEC (Method 1) and 30 εg ml −1 anti-flagella E. coli EPEC antibodies (Method 2) in PBS, at 4 ˚C for 60 min. After incubation was completed, 1DZnO samples were washed with PBS (5 min) and rinsed three times (5 min each) to avoid unspecific adhesions.
FTIR and AFM Characterization: FTIR was used to monitor the chemical modifications performed on 1DZnO. Measurements were carried out in a Vertex 70 Bruker equipment in the middle region (400−4000 cm −1 ). To observe the surface changes during the biofunctionalization, morphological analyses were performed in Nanosurf AFM equipment (tapping mode).
Photoluminescence Characterization: PL spectroscopy was utilized to monitor changes in the optical response of the platforms after each stage of the biofunctionalization process. Spectra Pro 2500i spectrometer and Kimmon Koha He-Cd laser (325 nm / 6 mW) were used to measure the emission spectra of the samples.
TEM and FIB Characterization: FIB Jeol JEM-9320FIB equipment (Ga: source, 30kV) was used to prepare the samples. Additionally, Pt deposition was performed to enhance resolution. Compositional mapping by EDS in STEM mode measurements was carried out to evaluate the biofunctionalization results. JEOL JEM-201 microscope (LaB 6 filament) was used to obtain sample micrographs.
Quantitative Confirmation of Immobilization: To confirm the number of antibodies attached to the surface of 1DZnO_2, the antibody concentrations were quantified by the Bradford protein assay (BioRad) in a microplate reader (Multiskan™ GO Microplate Spectrophotometer Thermofisher Scientific). A calibration curve was plotted from different concentrations of BSA protein standard (BioRad) at an interval of 75-2000 εg ml −1 . Data were fitted by linear regression to obtain the equation correlating the UV absorbance measurement (OD 590nm ) with the protein concentration. Protein quantification from antibody stocks before and after immobilization (washes) was performed to obtain the antibody immobilization efficiencies after each strategy.

Supporting Information
Supporting Information is available from the Wiley Online Library or from the author.

Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.