A Novel NiFe2O4/Paper-Based Magnetoelastic Biosensor to Detect Human Serum Albumin

For the first time, a novel NiFe2O4/paper-based magnetoelastic (ME) biosensor was developed for rapid, sensitive, and portable detection of human serum albumin (HSA). Due to the uniquely magnetoelastic effect of NiFe2O4 nanoparticles and the excellent mechanical properties of the paper, the paper-based ME biosensor transforms the surface stress signal induced by the specific binding of HSA and antibody modified on the paper into the electromagnetic signal. The accumulated binding complex generates a compressive stress on the biosensor surface, resulting in a decrease in the biosensor’s static magnetic permeability, which correlates to the HSA concentrations. To improve the sensitivity of the biosensor, the concentration of NiFe2O4 nanofluid and the impregnated numbers of the NiFe2O4 nanofluid-impregnated papers were optimized. The experimental results demonstrated that the biosensor exhibited a linear response to HSA concentrations ranging from 10 μg∙mL−1 to 200 μg∙mL−1, with a detection limit of 0.43 μg∙mL−1, which is significantly lower than the minimal diagnosis limit of microalbuminuria. The NiFe2O4/paper-based ME biosensor is easy to fabricate, and allows the rapid, highly-sensitive, and selective detection of HSA, providing a valuable analytical device for early monitoring and clinical diagnosis of microalbuminuria and nephropathy. This study shows the successful integration of the paper-based biosensor and the ME sensing analytical method will be a highly-sensitive, easy-to-use, disposable, and portable alternative for point-of-care monitoring.


Introduction
Diabetic nephropathy is a chronic complication of diabetes, affecting human endocrine metabolism [1]. There is increasing incidence of diabetic nephropathy in China, and it has become the second cause of end-stage renal disease, second only to various forms of glomerulonephritis. Albuminuria is the first clinical manifestation of diabetic nephropathy, initially intermittent, and then becoming persistent. Elevated albumin levels in urine (20-200 mg·L −1 ), a condition known as albuminuria, may manifest as renal dysfunction [2]. Detection of the levels of human serum albumin (HSA) or micro human serum albumin (mHSA) in urine can allow earlier diagnosis of albuminuria and can help control the condition [3,4]. HSA is an important human plasma biomacromolecule, accounting for about 60% of total plasma protein, with a molecular weight of 67 kDa [5,6]. Therefore, it is necessary to accurately detect and quantitatively analyze HSA in clinical applications and biochemistry experiments.
where H is the magnetic field strength and B is the magnetic induction strength.
Here, a NiFe 2 O 4 /paper-based ME biosensor, which composed of cellulose paper infiltrated with NiFe 2 O 4 as the transducer platform, and the anti-HSA antibody as the biorecognition molecule, was first developed for HSA detection. For the first time, NiFe 2 O 4 was doped into the paper, thus changing its properties, so that ordinary paper had magnetoelastic properties. The successful immobilization of anti-HSA antibody provided sites for the highly efficient binding of HSA, which caused the stress variation of the paper-based ME biosensor, inducing the change in the µ. The performance of the paper-based ME biosensors was then assessed for different concentrations of HSA. Accumulation of the biomacromolecule HSA on the surface caused a decrease of µ. The experimental results showed excellent performance of the paper-based ME biosensor for HSA detection, with advantages of simple preparation method, low cost, high sensitivity, strong specificity, and a large linear range.   were purchased from Shanghai Fanke Biological Technology Co., Ltd. The water used in the experiments was obtained from an ultra-pure water manufacturing system (URT-11-10T).

Preparation of the Paper
Different filter papers have been used in the development of paper-based sensors due to their wicking ability [31,32]. Among them, Whatman chromatography paper has been the most widely used. In this study, Whatman chromatography paper No.1 with a thickness of 180 µm and grammage of 88 g/m 2 was selected as the basis for construction of the NiFe 2 O 4 /paper nanocomposite. Paper of different sizes (16 mm × 1 mm, 4 mm × 4 mm) was prepared for compounding with NiFe 2 O 4 to obtain NiFe 2 O 4 /paper nanocomposite, as shown in Figure 1A(a). The paper was characterized by SEM, EDS, and FTIR spectra. The fabrication process of NiFe2O4/paper nanocomposite was as follows: (1) Prepared paper samples of different sizes were immersed in the NiFe2O4 nanofluid for one hour until fully saturated, as shown in Figure 1A(b). (2) The impregnated paper was dehydrated at 70 °C for 10 min on an electric heating plate to achieve a drying effect, as shown in Figure 1A(c). (3) The dried paper sheet was stored in an environment with constant temperature and humidity. The resulting NiFe2O4/paper nanocomposite samples were characterized by SEM, EDS, and FTIR spectra. The fabrication process of the NiFe2O4/paper nanocomposite, (a) paper, (b) the NiFe2O4 nanofluid-impregnated paper, (c) the dehydrated NiFe2O4/paper; (B) The fabrication process of the CYS-AuNPs, (a) AuNPs, (b) CYS-AuNPs; (C) The fabrication process of the NiFe2O4/paper-based ME biosensor, (a) CYS-AuNPs/NiFe2O4/paper-based sensors, (b) anti-HSA/ CYS-AuNPs/NiFe2O4/paper-based sensors, (c) BSA/anti-HSA/CYS-AuNPs/NiFe2O4/paper-based biosensors, (d) HSA/BSA/ anti-HSA/CYS-AuNPs/NiFe2O4/paper-based biosensors.

Fabrication of the Paper-Based ME Sensor
First, gold nanoparticles (AuNPs) were prepared. To do this, 250 nL of 0.1 M HAuCl4 solution was added to 100 mL of deionized water and heated to boiling. Then, 600 nL of 0.25 M Na3C6H5O7 solution was rapidly added to the boiling solution; the color of the solution changed from pale yellow to deep red in a short time, followed by continued heating for 30 min to ensure complete reduction. The obtained solution was centrifuged at 12,000 rpm for 15 min and then washed three times. Finally, the AuNPs were re-dispersed into deionized water and stored at 4 °C ( Figure 1B(a)).
Second, the prepared AuNPs were functionalized. A 4 mM cysteamine solution was mixed with the solution of AuNPs at a volume ratio of 1:9, and then incubated at room temperature for 12 h to obtain thiolated AuNPs (CYS-AuNPs), as shown in Figure 1B(b). The AuNPs and CYS-AuNPs were characterized by TEM and Dynamic Light Scattering (DLS).
Third, the paper-based ME sensors were prepared. Samples of 4 mm × 4 mm NiFe2O4/paper nanocomposites were horizontally immersed in the solution of CYS-AuNPs for 2 h, and the CYS-AuNPs were fixed on the cellulose fiber and cellulose fiber network of the paper by physical deposition. Finally, the NiFe2O4/paper nanocomposites with immobilized CYS-AuNPs were dried in a stream of nitrogen to obtain the paper-based ME sensors ( Figure 1C(a)).

Antibody Immobilization
The anti-HSA was diluted into PBS buffer (0.01 M, pH = 7.4) to prepare for 50 μg•ml −1 antibody solution [26,33,34]. The anti-HSA antibody was activated through the 4 mg•mL −1 -4 mg•mL −1 The fabrication process of NiFe 2 O 4 /paper nanocomposite was as follows: (1) Prepared paper samples of different sizes were immersed in the NiFe 2 O 4 nanofluid for one hour until fully saturated, as shown in Figure 1A(b). (2) The impregnated paper was dehydrated at 70 • C for 10 min on an electric heating plate to achieve a drying effect, as shown in Figure 1A(c). (3) The dried paper sheet was stored in an environment with constant temperature and humidity. The resulting NiFe 2 O 4 /paper nanocomposite samples were characterized by SEM, EDS, and FTIR spectra.

Fabrication of the Paper-Based ME Sensor
First, gold nanoparticles (AuNPs) were prepared. To do this, 250 nL of 0.1 M HAuCl 4 solution was added to 100 mL of deionized water and heated to boiling. Then, 600 nL of 0.25 M Na 3 C 6 H 5 O 7 solution was rapidly added to the boiling solution; the color of the solution changed from pale yellow to deep red in a short time, followed by continued heating for 30 min to ensure complete reduction. The obtained solution was centrifuged at 12,000 rpm for 15 min and then washed three times. Finally, the AuNPs were re-dispersed into deionized water and stored at 4 • C ( Figure 1B(a)).
Second, the prepared AuNPs were functionalized. A 4 mM cysteamine solution was mixed with the solution of AuNPs at a volume ratio of 1:9, and then incubated at room temperature for 12 h to obtain thiolated AuNPs (CYS-AuNPs), as shown in Figure 1B Third, the paper-based ME sensors were prepared. Samples of 4 mm × 4 mm NiFe 2 O 4 /paper nanocomposites were horizontally immersed in the solution of CYS-AuNPs for 2 h, and the CYS-AuNPs were fixed on the cellulose fiber and cellulose fiber network of the paper by physical deposition. Finally, the NiFe 2 O 4 /paper nanocomposites with immobilized CYS-AuNPs were dried in a stream of nitrogen to obtain the paper-based ME sensors ( Figure 1C(a)).

Antibody Immobilization
The anti-HSA was diluted into PBS buffer (0.01 M, pH = 7.4) to prepare for 50 µg·ml −1 antibody solution [26,33,34]. The anti-HSA antibody was activated through the 4 mg·mL −1 -4 mg·mL −1 EDC/NHS mixed solution at room temperature for 30 min. The carboxyl group of the antibody was activated into NHS ester, which make the bonding with the amino group on the CYS-AuNPs be more efficiently [35].
The paper-based ME sensors were immersed into the activated antibody solution at room temperature for 1 h, so that the antibodies were uniformly immobilized on the CYS-AuNPs ( Figure 1C(b)). Then, the antibody-modified paper-based ME sensors were removed from the solution and rinsed with PBS to remove physically adsorbed antibodies. Next, the paper-based ME sensors were treated with 0.1 % BSA for 30 min to block the uncombined and non-specific sites ( Figure 1C(c)). Then, the paper-based ME sensors were rinsed with PBS and dried with nitrogen, to obtain the paper-based ME biosensor ready for HSA detection.

Signal Measurement
The magnetic field strength of the NiFe 2 O 4 /paper nanocomposites was measured using a Gauss meter. The detection design is shown in Figure 4a. The NiFe 2 O 4 /paper is attached to the tweezer, with one end clamped by adhesive tape. With a tunable external magnetic field produced by adjusting the distance between the NiFe 2 O 4 /paper and a magnet, the tip deflection and bending angle at the free end can be measured. Loading and unloading (hysteresis test) can be achieved by the placement and removal of magnets.
The hysteresis loop and saturation magnetic induction strength (B s ) of NiFe 2 O 4 nanoparticles, NiFe 2 O 4 /paper nanocomposites, and paper-based ME biosensors were investigated at room temperature using a vibrating-sample magnetometer (SQUID-VSM, Quantum Design, American) with an applied field of 1.1 T. The VSM test condition requires in-plane dimensions that do not exceed 4 mm × 4 mm, so samples with in-plane dimensions of 4 mm × 4 mm were used to detect the hysteresis loop. During the test of the H we found that long strip samples could more easily detect deflection than squares, so we tested 16 mm × 1 mm samples of the same area to optimize the concentration of NiFe 2 O 4 nanofluid. In addition, different concentrations of HSA solution were prepared by serial dilution with PBS. The paper-based ME biosensor was immersed in different HSA solutions at room temperature for 1 h, then removed and dried in a stream of nitrogen. Then, the hysteresis loop of the paper-based ME biosensors after detecting different concentrations of HSA ( Figure 1C(d)) was investigated. The µ of the materials at maximum magnetic field strength (H m ) was calculated by Equation (1).

Characterization of NiFe 2 O 4 Nanoparticles and AuNPs
The NiFe 2 O 4 nanoparticles were characterized using a high-resolution transmission electron microscope (JEM2010-TEM, JEOL, Tokyo, Japan). Figure 2 shows the TEM images of NiFe 2 O 4 nanoparticles at different multiples. Since NiFe 2 O 4 nanoparticles have ferrimagnetic properties, the nanoparticles have a large surface energy and are prone to aggregation. As can be seen from   It can be seen from the TEM images that both AuNPs and CYS-AuNPs are spherical, AuNPs are evenly dispersed, and some AuNPs modified with CYS exhibit slight aggregation. The particle size distributions of AuNPs and CYS-AuNPs were analyzed using a dynamic light scattering analyzer (ZS90-DLS, Malvern, UK). According to DLS analysis, as shown in Figure 3b,d, the particle size distribution of AuNPs was concentrated at about 22.48 nm, with an average particle size of 25.21 nm, while the particle size distribution of CYS-AuNPs was concentrated at about 27.55 nm, with an increase in average particle size to 31.59 nm, indicating the effective modification of CYS on AuNPs.

Optimization of NiFe 2 O 4 Nanofluid Concentration and the Number of Impregnation Times
The amount of NiFe 2 O 4 nanoparticles embedded into the paper was associated with the concentration of NiFe 2 O 4 nanofluid and the number of times the paper was impregnated by the NiFe 2 O 4 nanofluid. To improve the sensitivity of the NiFe 2 O 4 /paper nanocomposite, we optimized the working concentration of the NiFe 2 O 4 nanofluid. Figure 4a is the schematic diagram of the testing device. The tip deflection was measured for the NiFe 2 O 4 /paper impregnated with different concentration of NiFe 2 O 4 nanofluids as shown in Figure 4b. The NiFe 2 O 4 /paper impregnated with 25 wt% NiFe 2 O 4 nanofluid had a deflection angle that was significantly larger than that of the paper soaked in 5 wt% and 15 wt% solutions. The mass percentage of NiFe 2 O 4 nanoparticles was measured in the NiFe 2 O 4 /paper samples as 38.89%, 47.62%, and 52.17%, respectively, for corresponding NiFe 2 O 4 nanofluid concentrations of 5%, 15%, and 25%. With the NiFe 2 O 4 nanofluid concentration increased from 5 wt% to 25 wt%, the H around the NiFe 2 O 4 /paper tip continued to increase, as shown in Figure 4c. The result shows that the response reached the largest value of approximately 110.7 mT at a NiFe 2 O 4 nanofluid concentration of 25 wt%. Since it is difficult to prepare NiFe 2 O 4 nanofluid at higher mass percentages, 25 wt% was selected as the optimal NiFe 2 O 4 nanofluid concentration. soaked in 5 wt% and 15 wt% solutions. The mass percentage of NiFe2O4 nanoparticles was measured in the NiFe2O4/paper samples as 38.89%, 47.62%, and 52.17%, respectively, for corresponding NiFe2O4 nanofluid concentrations of 5%, 15%, and 25%. With the NiFe2O4 nanofluid concentration increased from 5 wt% to 25 wt%, the H around the NiFe2O4/paper tip continued to increase, as shown in Figure 4c. The result shows that the response reached the largest value of approximately 110.7 mT at a NiFe2O4 nanofluid concentration of 25 wt%. Since it is difficult to prepare NiFe2O4 nanofluid at higher mass percentages, 25 wt% was selected as the optimal NiFe2O4 nanofluid concentration. Next, we optimized the number of times of impregnation with the NiFe2O4 nanofluid. The NiFe2O4 nanofluid was used to impregnate the paper 1,3,5,7,9 times, each time for 1 h. The paper was removed from the NiFe2O4 nanofluid and then dried at 70 °C for 10 min before the next dip. Figure 5a shows the response of NiFe2O4/paper samples treated with varying times of impregnation. It can be clearly seen that the NiFe2O4/paper impregnated seven times exhibited the most obvious deflection for H of 110 mT. When the number of impregnations increased from 1 to 9, the tip deflection of NiFe2O4/paper and the mass percentage of NiFe2O4 nanoparticles increased first and then decreased, as shown in Figure 5b. The results show the largest response of approximately 9 mm and 77.55% for seven cycles of impregnation. The adsorption capacity of the paper reached saturation and nanoparticles would be lost with increased of impregnation cycles, explaining the decrease in the deflection and mass percentage after nine impregnation steps. Therefore, the optimal impregnation was set as seven times. Next, we optimized the number of times of impregnation with the NiFe 2 O 4 nanofluid. The NiFe 2 O 4 nanofluid was used to impregnate the paper 1,3,5,7,9 times, each time for 1 h. The paper was removed from the NiFe 2 O 4 nanofluid and then dried at 70 • C for 10 min before the next dip. Figure 5a shows the response of NiFe 2 O 4 /paper samples treated with varying times of impregnation. It can be clearly seen that the NiFe 2 O 4 /paper impregnated seven times exhibited the most obvious deflection for H of 110 mT. When the number of impregnations increased from 1 to 9, the tip deflection of NiFe 2 O 4 /paper and the mass percentage of NiFe 2 O 4 nanoparticles increased first and then decreased, as shown in Figure 5b. The results show the largest response of approximately 9 mm and 77.55% for seven cycles of impregnation. The adsorption capacity of the paper reached saturation and nanoparticles would be lost with increased of impregnation cycles, explaining the decrease in the deflection and mass percentage after nine impregnation steps. Therefore, the optimal impregnation was set as seven times.
Sensors 2020, 20, 5286 8 of 15 deflection of NiFe2O4/paper and the mass percentage of NiFe2O4 nanoparticles increased first and then decreased, as shown in Figure 5b. The results show the largest response of approximately 9 mm and 77.55% for seven cycles of impregnation. The adsorption capacity of the paper reached saturation and nanoparticles would be lost with increased of impregnation cycles, explaining the decrease in the deflection and mass percentage after nine impregnation steps. Therefore, the optimal impregnation was set as seven times.

Characterization of Paper and NiFe2O4/Paper Nanocomposite
The paper surface was imaged and appeared smooth (Figure 6b). The surface appearance of the paper and the NiFe2O4/paper nanocomposite was characterized with a scanning electron microscope (SU3500-SEM, Hitachi, Japan). It can be seen from Figure 6a that the cellulose fibers of the paper appeared intricately overlapped to form a porous and closed cellulose fiber network. The length-diameter ratio of the fibers in the cellulose paper was relatively large, with a diameter of about 10 μm [36]. This indicates that the structure of cellulose paper is stable, providing many

Characterization of Paper and NiFe 2 O 4 /Paper Nanocomposite
The paper surface was imaged and appeared smooth (Figure 6b). The surface appearance of the paper and the NiFe 2 O 4 /paper nanocomposite was characterized with a scanning electron microscope (SU3500-SEM, Hitachi, Japan). It can be seen from Figure 6a that the cellulose fibers of the paper appeared intricately overlapped to form a porous and closed cellulose fiber network. The length-diameter ratio of the fibers in the cellulose paper was relatively large, with a diameter of about 10 µm [36]. This indicates that the structure of cellulose paper is stable, providing many attachment sites for NiFe 2 O 4 nanoparticles. As can be seen from the macrostructure photograph of the NiFe 2 O 4 /paper (Figure 6d), the paper changed from white to brown after impregnating with the NiFe 2 O 4 nanofluid, and the impregnation is uniform. According to the SEM images (Figure 6c), the NiFe 2 O 4 nanoparticles were uniformly distributed on the surface of cellulose fibers and diffused into the porous regions of the cellulose fiber networks. attachment sites for NiFe2O4 nanoparticles. As can be seen from the macrostructure photograph of the NiFe2O4/paper (Figure 6d), the paper changed from white to brown after impregnating with the NiFe2O4 nanofluid, and the impregnation is uniform. According to the SEM images (Figure 6c), the NiFe2O4 nanoparticles were uniformly distributed on the surface of cellulose fibers and diffused into the porous regions of the cellulose fiber networks. FTIR measurements were conducted using a Fourier-transform infrared spectrometer (Bruker Optik GmbH Tensor 27-FTIR, Ettlingen, Germany). It is evident from the FTIR spectrum presented in Figure 6e that single bond vibration corresponding to cellulose can be detected in analysis of both paper and NiFe2O4/paper samples. A wide band of 3300 cm −1 (1 in Figure 6e) was observed in both samples, which was attributed to the hydrogen bonding of O-H tensile vibration, indicating the hydrophilic tendency of cellulose paper. The band at 2898 cm −1 (2) belongs to the aliphatic C-H stretching vibration. The low-intensity bands located at 1642 cm −1 (3), 1426 cm −1 (4), 1364 cm −1 (5), 1158 cm −1 (6) and 1107 cm −1 (7) are respectively the bending vibration of OH bond in adsorbed water, CH2 symmetrical bending, CH bending, C-O-C stretching and reverse bending. The strong band at 1027 cm −1 (8) is related to C-C, C-OH and C-H ring and side group vibrations, and is assigned to the amorphous region of cellulose. The small band at 894 cm −1 (9) is attributable to C-O-C, C-C-O and C-C-H distortion and stretching vibrations. The band stands for the conformational deformation of the glycoside 4 C1 ring in cellulose and the glycosidic bond between the glucuronic acid ring [37]. In addition, the bands located at 563 cm −1 and 416 cm −1 (10,11) were assigned to the stretching vibration of the metal oxide bond (Fe-O, Ni-O) at the tetrahedral/octahedral position of the ferrite structure as shown in Figure 6e [38], indicating the successfully immobilization of NiFe2O4 nanoparticles on the surface of cellulose fibers and diffusion of these nanoparticles into the porous regions among cellulose fiber networks. A summary of infrared band signals associated with bond vibrations within cellulose and metal oxide is highlighted in Table 1.  Table 1.   Table 1.
Sensors 2020, 20, 5286 9 of 15 FTIR measurements were conducted using a Fourier-transform infrared spectrometer (Bruker Optik GmbH Tensor 27-FTIR, Ettlingen, Germany). It is evident from the FTIR spectrum presented in Figure 6e that single bond vibration corresponding to cellulose can be detected in analysis of both paper and NiFe 2 O 4 /paper samples. A wide band of 3300 cm −1 (1 in Figure 6e) was observed in both samples, which was attributed to the hydrogen bonding of O-H tensile vibration, indicating the hydrophilic tendency of cellulose paper. The band at 2898 cm −1 (2) belongs to the aliphatic C-H stretching vibration. The low-intensity bands located at 1642 cm −1 (3), 1426 cm −1 (4), 1364 cm −1 (5), 1158 cm −1 (6) and 1107 cm −1 (7) are respectively the bending vibration of OH bond in adsorbed water, CH 2 symmetrical bending, CH bending, C-O-C stretching and reverse bending. The strong band at 1027 cm −1 (8) is related to C-C, C-OH and C-H ring and side group vibrations, and is assigned to the amorphous region of cellulose. The small band at 894 cm −1 (9) is attributable to C-O-C, C-C-O and C-C-H distortion and stretching vibrations. The band stands for the conformational deformation of the glycoside 4 C 1 ring in cellulose and the glycosidic bond between the glucuronic acid ring [37]. In addition, the bands located at 563 cm −1 and 416 cm −1 (10,11) were assigned to the stretching vibration of the metal oxide bond (Fe-O, Ni-O) at the tetrahedral/octahedral position of the ferrite structure as shown in Figure 6e [38], indicating the successfully immobilization of NiFe 2 O 4 nanoparticles on the surface of cellulose fibers and diffusion of these nanoparticles into the porous regions among cellulose fiber networks. A summary of infrared band signals associated with bond vibrations within cellulose and metal oxide is highlighted in Table 1. Figure 7 presents the performance of the optimized NiFe 2 O 4 /paper nanocomposite with in-plane dimensions of 16 mm × 1 mm under tunable magnetic field. The distribution of the magnetic field was measured by Gauss meter, as presented in Figure 7a. The magnetic intensity decreased with increased distance between the NiFe 2 O 4 /paper and the magnet. The optimized NiFe 2 O 4 /paper were driven by an increasing magnetic intensity, and the snapshots were exhibited in Figure 7d. With the increase of the magnetic field, the deflection and bending angle of the sensor tip increased, reaching peak values at 9 mm and 30 • when the magnetic field was 110 mT, as shown in Figure 7b,c. As the magnetic field gradually decreased to 0 mT, the deflection and bending angle of the sensor tip returned to 0.3 mm and 1.1 • , which was close to the initial value. The small hysteresis effect may be ascribed to the big flexing deformation in excess of the elastic deformation of NiFe 2 O 4 /paper, resulting in irreversible deformation.  Figure 7 presents the performance of the optimized NiFe2O4/paper nanocomposite with in-plane dimensions of 16 mm × 1 mm under tunable magnetic field. The distribution of the magnetic field was measured by Gauss meter, as presented in Figure 7a. The magnetic intensity decreased with increased distance between the NiFe2O4/paper and the magnet. The optimized NiFe2O4/paper were driven by an increasing magnetic intensity, and the snapshots were exhibited in Figure 7d. With the increase of the magnetic field, the deflection and bending angle of the sensor tip increased, reaching peak values at 9 mm and 30° when the magnetic field was 110 mT, as shown in Figure 7b,c. As the magnetic field gradually decreased to 0 mT, the deflection and bending angle of the sensor tip returned to 0.3 mm and 1.1°, which was close to the initial value. The small hysteresis effect may be ascribed to the big flexing deformation in excess of the elastic deformation of NiFe2O4/paper, resulting in irreversible deformation.  Figure 8 shows the hysteresis loop of NiFe2O4 nanoparticles, NiFe2O4/paper nanocomposite, and paper-based ME biosensor. Under an external magnetic field, the B increased with the magnetic field until reaching saturation. The static magnetic permeability values corresponding to each material at the Hm were calculated as μ1 = 29.5 H/m, μ2 = 18.8 H/m，and μ3 = 10.5 H/m. Among them,  Figure 8 shows the hysteresis loop of NiFe 2 O 4 nanoparticles, NiFe 2 O 4 /paper nanocomposite, and paper-based ME biosensor. Under an external magnetic field, the B increased with the magnetic field until reaching saturation. The static magnetic permeability values corresponding to each material at the H m were calculated as µ 1 = 29.5 H/m, µ 2 = 18.8 H/m, and µ 3 = 10.5 H/m. Among them, the B s value of NiFe 2 O 4 nanoparticles is the largest. Due to the mixing of non-magnetic paper, the B s of NiFe 2 O 4 /paper decreased, and the corresponding µ value also decreased. After immobilization of the antibody on NiFe 2 O 4 /paper, the µ value continued to decrease (µ 3 < µ 2 < µ 1 ) because the antibody tried to expand on the cellulose fiber and cellulose fiber network, but since it was confined, a compressive stress was generated on the surface. At this time, the magnetization direction turned to the direction perpendicular to the stress, and the magnetization in the direction of the compressive stress was weakened, resulting in a weakened µ value in the direction of the compressive stress. Analysis of the hysteresis loops confirms the successful fabrication of NiFe 2 O 4 /paper and antibody immobilization.  Figure 9a shows the hysteresis loop of the biosensor for HSA detection at concentrations ranging from 0 to 200 µg·mL −1 . The specific binding of the antibody and the biomacromolecule HSA caused an increase in the compressive stress on the biosensor, leading to a decrease in the B s value. The results show that as the HSA concentration increased from 10 µg·mL −1 to 200 µg·mL −1 , the B s value of the biosensor gradually decreased, and the corresponding µ at the H m also decreased from 9.4 H/m to 3.5 H/m. By subtracting the µ of the biosensor detecting different concentrations of HSA from the µ 3 (Figure 8) of the biosensor without HSA detection, the corresponding change in static magnetic permeability (∆µ) was obtained, as shown in Figure 9b. The biosensor was measured in 0 µg·ml −1 HSA as a blank control experiment to confirm that the response was specific to HSA. By calculation, the corresponding µ at the H m was 10.39 H/m and the ∆µ was only 0.11 H/m. This noise response was far less than the ∆µ corresponding to 10 µg·mL −1 HSA, so it could be ignored in the detection process. Figure 9b shows that the ∆µ of the biosensor shows good linear correlation with HSA concentration for a range of 10 µg·mL −1 to 200 µg·mL −1 . The linear equation was y = −0.03285x − 0.55372, and the correlation coefficient was 0.98273, with a detection limit is 0.43 µg·mL −1 . The experimental results show that the NiFe 2 O 4 /paper-based ME biosensor can achieve low-cost, accurate and sensitive detection for HSA. The successful detection of HSA demonstrated the unique properties of NiFe 2 O 4 /paper make it an effective material for magnetoelastic biosensor design.

Characterization of the Preparing and Detecting Process
The surface elements were assessed by energy dispersive spectrometer (SU3500-EDS, Hitachi, Japan). The EDS analysis of non-functionalized NiFe 2 O 4 /paper nanocomposite is shown in Figure 10a. Only C and O elements in cellulose paper and Fe and Ni elements in NiFe 2 O 4 nanoparticles were detected. During the entire experiment, the amount of Fe and Ni elements showed little reduction, which demonstrated that NiFe 2 O 4 nanoparticles were firmly attached to the paper and hardly diminished, as shown in Figure 10. The peak corresponding to Au element was detected after deposition of CYS-AuNPs, and the peak of N element in CYS was detected in the range of 0.250-0.5 keV, as shown in Figure 10b, indicating that CYS-AuNPs were successfully deposited on the NiFe 2 O 4 /paper. Since antibodies and HSA are protein molecules, they both contain the C, O, and N elements. After immobilizing the antibody, the peak value of the C, O, and N elements were greater than the peak value of the C, O, and N elements of the paper-based ME sensor, as shown in Figure 10c, indicating effective anti-HSA antibody immobilization. When the antibody-immobilized sensor was exposed to HSA solution (100 µg·mL −1 ), the peak values of the C, O, and N elements were greatly increased, since the HSA was successfully adsorbed by the specific recognition region of the anti-HSA antibody, as shown in Figure 10d. Comparison of EDS analysis results for the different samples indicates that the fabrication and the detection procedures were successful.

Specificity Measurement
To evaluate the specificity of the paper-based ME biosensor, different potential interferents such as UA, CRE, HGB, BSA, and CEA, with concentrations of 200 µg·mL −1 , were detected. It can be clearly observed from Figure 11 that the amount of ∆µ caused by HSA interaction is almost four times greater than for other biomolecules, indicating that the biosensor shows little response to these other biomolecules. The results implied that the biosensor has high specificity for the determination of HSA.
than the peak value of the C, O, and N elements of the paper-based ME sensor, as shown in Figure  10c, indicating effective anti-HSA antibody immobilization. When the antibody-immobilized sensor was exposed to HSA solution (100 μg•mL −1 ), the peak values of the C, O, and N elements were greatly increased, since the HSA was successfully adsorbed by the specific recognition region of the anti-HSA antibody, as shown in Figure 10d. Comparison of EDS analysis results for the different samples indicates that the fabrication and the detection procedures were successful. Figure 10. EDS analysis of (a) NiFe2O4/paper nanocomposite; (b) paper-based ME sensor (CYS-AuNPs/NiFe2O4/paper); (c) paper-based ME biosensor (Antibody/CYS-AuNPs/NiFe2O4/paper); (d) After HSA detection (HSA/Antibody/CYS-AuNPs/NiFe2O4/paper).

Specificity Measurement
To evaluate the specificity of the paper-based ME biosensor, different potential interferents such as UA, CRE, HGB, BSA, and CEA, with concentrations of 200 μg•mL −1 , were detected. It can be clearly observed from Figure 11 that the amount of Δμ caused by HSA interaction is almost four times greater than for other biomolecules, indicating that the biosensor shows little response to these other biomolecules. The results implied that the biosensor has high specificity for the determination of HSA.

Conclusions
In this study, a NiFe2O4 nanoparticle-impregnated paper-based ME biosensor for HSA detection was developed based on the magnetoelastic effect of NiFe2O4 magnetic nanoparticles and the specific binding of antibodies to antigens. The NiFe2O4/paper was fabricated using a low-cost blending method, and EDS, SEM, and FTIR measurement were used to confirm that the NiFe2O4 nanoparticles infiltration was successful. The specific binding of HSA and anti-HSA antibody increases the compressive stress on the biosensor surface, resulting in a decrease in the static magnetic permeability. The biosensor can detect a wide range of HSA concentrations (10 μg•mL −1 −200 μg•mL −1 ) and exhibits a linear relationship with a correlation coefficient of 0.98273 within this range. The biosensor detection limit is 0.43 μg•mL −1 , meeting the health diagnostic criteria. The NiFe2O4/paper-based ME biosensor shows a low-cost, specific, and sensitive response towards HSA, making this an effective method for early prevention of diabetes and diabetic complications. Furthermore, this work demonstrated a new and simple method for ME biosensor design by exploiting the unique properties of NiFe2O4/paper, and provided a new idea for the

Conclusions
In this study, a NiFe 2 O 4 nanoparticle-impregnated paper-based ME biosensor for HSA detection was developed based on the magnetoelastic effect of NiFe 2 O 4 magnetic nanoparticles and the specific binding of antibodies to antigens. The NiFe 2 O 4 /paper was fabricated using a low-cost blending method, and EDS, SEM, and FTIR measurement were used to confirm that the NiFe 2 O 4 nanoparticles infiltration was successful. The specific binding of HSA and anti-HSA antibody increases the compressive stress on the biosensor surface, resulting in a decrease in the static magnetic permeability. The biosensor can detect a wide range of HSA concentrations (10 µg·mL −1 −200 µg·mL −1 ) and exhibits a linear relationship with a correlation coefficient of 0.98273 within this range. The biosensor detection limit is 0.43 µg·mL −1 , meeting the health diagnostic criteria. The NiFe 2 O 4 /paper-based ME biosensor shows a low-cost, specific, and sensitive response towards HSA, making this an effective method for early prevention of diabetes and diabetic complications. Furthermore, this work demonstrated a new and simple method for ME biosensor design by exploiting the unique properties of NiFe 2 O 4 /paper, and provided a new idea for the application of cellulose paper in the field of ME biosensors.

Conflicts of Interest:
The authors declare no conflict of interest.