Structural characterization
According to the literature [42, 43, 45, 46] and the synthesis processes, the material employed as anodes is composed of ZnO and nickel complexes of hydrazine and ammonia. Figure 2 presents the XRD patterns and the FTIR spectra for the four electrocatalysts evaluated. As can be seen in Fig. 2a) the diffraction peaks for 2\(\theta\) located at 31.74°, 34.44°, 36.26°, 47.51°, and 56.56° correspond to the principal ZnO crystallographic planes: (100), (002), (101), (102) and (110), respectively (ICDD 36-1451), indicating the powders correspond to a polycrystalline hexagonal wurtzite phase. The crystallite size of the wurtzite phase ZnO was calculated for the (0 0 2) crystalline plane in all materials using Scherrer´s formula; the results are presented in Table 1. The crystallite size of ZnO-Ni materials shows a minor decrease compared to ZnO without nickel. This reduction could arise from the incorporation of Ni into the lattice. In the case of the 9% sample, there is a little increase, which could be attributed to the precipitation of the other phases.
The process followed for the precipitation of nickel-based particles in the ZnO indicates that the product obtained is a mixture of nickel complexes, [Ni(N2H4)2]Cl2 and [Ni(NH3)6]Cl2, as reported by Park et al. [43] and Huang et al. [46]. In Fig. 2b), the vibration bands indicate the presence of those hydrazine and ammonia complexes with the ZnO-Ni materials. The three ZnO-Ni (3, 6, 9%) catalysts present the same bands, and only a difference in intensity is observed. The wide band located between 3280–3550 cm-1 could be due to an overlapping of the N-H asymmetric stretching of N-H and O-H stretching signals, the last one indicating water adsorption mainly associated with the ZnO [47], which is also observed in the spectra of the ZnO without nickel. The double band at 3227 and 3290 cm− 1 correspond to the symmetric stretching of N-H (νNH) [43], the bands at 1606 cm− 1 and 1173 cm− 1 correspond to the degenerate deformation and the symmetric deformation of the H-N-H bonds (δHNH) [48, 49]. These bands, νNH and δHNH, are associated with both complexes, hydrazine, and ammonia, while the band at 978 cm− 1 corresponds to N-N stretching [43, 46, 50, 51], which is associated only with the presence of hydrazine complexes. Finally, the bands found between 690 − 613 cm− 1 correspond to rocking vibration and symmetric distortion of NH3, indicating the presence of ammonia complexes [43, 48]. It is observed that while the concentration of Ni-based particles increases, the intensity of the signal does too. In Fig. 2a), it is observed that the XRD pattern of the ZnO-Ni 9% presents some diffraction peaks with low intensity that are not associated with the ZnO, indicating, as the FTIR results, the presence of hydrazine and ammonia complexes. The peaks located at 22.3°, 31.9° and 33.9 ° (2θ) are identified as nickel hydrazine chloride, [Ni(N2H4)2]Cl2 (ICDD: 28–0695)[43, 46], and the minor signal located at 15.2° correspond to the main diffraction peak of the nickel ammine chloride, \(\left[Ni{\left(N{H}_{3}\right)}_{6}\right]C{l}_{2}\) (ICDD: 76-1842) [43]. According to the nature of these nickel complexes and a probably amorphous structure (synthesis process) of the materials with 3 and 6% nickel-based particles, the mentioned diffraction peaks are not observed in XRD patterns; however, as discussed above, the FTIR shows the presence of hydrazine and ammonia, which also plays an important role in the catalyst activity and will be explained later.
The excess of the hydrazine and ammonia complexes could affect the catalytic activity (Figs. 4 and 5), which could also be related to the structure and size of the conglomerates or particles obtained, as observed in the SEM images presented in Fig. 3. The SEM images show the formation of flake-shaped particle clusters of the synthesized materials. Figure 3a) and b) show the ZnO without nickel and the ZnO- Ni 3%, respectively, and they present a cluster structure that allows the formation of pores between clusters meanwhile ZnO-Ni 6% and ZnO-Ni 9% shown in Fig. 3c) and d) respectively, present a less porous conglomerate, probably due to an excessive presence of the nickel complexes that block up the spaces between the ZnO particles.
Table 1
Crystallite size of ZnO and ZnO-Ni synthesized materials
Catalyst | Crystallite size (nm) |
ZnO | 27.0 |
ZnO-Ni 3% | 26.5 |
ZnO-Ni 6% | 23.8 |
ZnO-Ni 9% | 24.6 |
Paper-based µFC performance
The performance of the microfluidic fuel cell is based on the oxidation of urea in the urine and the reduction of the oxygen in the air. Therefore, the electrode materials and electrocatalyst are essential in performance improvement. This work focused on the power obtained when nickel-based particles are incorporated into a ZnO catalyst. The performance of the paper-based µFC was evaluated using human urine with an average composition (healthy man) or diabetic urine (Table 4) as fuels.
In Fig. 4are presented the polarization and power density curves using normal urine as fuel. Figure 4a) shows the performance of the ZnO catalyst, while Fig. 4b) shows the ZnO-Ni with the different nickel concentrations (3, 6, and 9%). The polarization curves presented in Fig. 4 show the representative behavior with the three characteristics zones (reaction, mixed, and diffusion zones); however, when the ZnO and ZnO-Ni materials are compared, the one without nickel (Fig. 4a) presents a more resistive polarization curve than the three ZnO-Ni materials (Fig. 4b).
The results with ZnO (without nickel) are a maximum voltage of 0.88 V, a maximum current of 0.05 mA cm− 2, and a maximum power density of 0.0081 mW cm− 2. Those values are highly overcome when using the ZnO-Ni electrodes (Table 2), meaning there is a significant improvement in the catalytic activity when the nickel is present in the composite. Although the improvement with Ni is evident, it could also be observed that the concentration of Ni is relevant to the µFC performance. As is observed in Table 2 and Fig. 4b, the best results are obtained with the 3% Ni concentration, while the 6 and 9% Ni show a current and power decrease, being the less effective the 9% Ni concentration. Nevertheless, as mentioned above, the values are still higher than those of ZnO without nickel. The open circuit voltage (OCV) reached with the ZnO-Ni 3% is 0.95 V, while the current and power densities are 2.12 mAcm− 2 and 0.237 mW cm− 2, respectively.
The presence of the nickel complexes supported in the ZnO improves urea oxidation; however, the catalytic activity is affected as the nickel concentration increases. This effect could be related to fewer active sites, as the structure seems less porous (Fig. 3), or could also be associated with a different reaction mechanism, resulting in a decrease in the performance of the µFC.
Table 2
Performance parameters of paper-based µFC with ZnO-Ni anodes, using normal urine as fuel
Anode | Voltage (V) | Current density (mA cm− 2) | Power density (mW cm− 2) |
ZnO | 0.88 | 0.05 | 0.0081 |
ZnO-Ni (3%) | 0.95 | 2.12 | 0.237 |
ZnO-Ni (6%) | 0.83 | 1.09 | 0.098 |
ZnO-Ni (9%) | 0.86 | 0.734 | 0.068 |
Several works present the use of urine as fuel in fuel cells; however, its application in microfluidic fuel cells, especially in paper-based µFC, is limited. Table 3 presents the performance parameters of the most recently reported microfluidic fuel cells that use normal urine as fuel. It is observed that the performance obtained with ZnO-Ni 3% and normal urine overcomes the current and power densities of the microfluidic fuel cells that do not use a pump to feed the cell.
Table 3
Microfluidic fuel cells reported in the literature that use urine as fuel
Type of fuel cell | Anode | Cathode | Fuel/oxidant | Voltage (V) | Current density (mA cm− 2) | Power density (mW cm− 2) | Reference |
paper-based microbial FC | Ag ink bio anode | C nanotubes ink bio cathode | Normal human urine/9.78 M H2O2 | 0.5 | 0.56 | 0.129 | [31] |
Photo- microfluidic FC | TiO2 with Ni(OH)2 | Pt/C | Normal human urine/air | 0.7 | 1.7 | 0.09 | [52] |
paper-based microfluidic FC | TiO2-Ni | Pt/C | Normal human urine/air | 0.96 | 1.00 | 0.23 | [32] |
Pump- microfluidic FC | NiAl-LDHs and PdNiO | NA | Normal human urine/0.3 M KOH | 1.0 | 152.54 | 50.19 | [53] |
paper-based microfluidic FC | ZnO-Ni 3% | Pt/C | Normal human urine/air | 0.95 | 2.12 | 0.24 | This work |
paper-based microfluidic FC | ZnO-Ni 3% | Pt/C | Diabetic human urine/air | 0.89 | 1.18 | 0.13 | This work |
FC = fuel cell NA = not available |
The use of paper-based µFC in POC devices as a power source for disease detection implies that the fuel used, in this case, the urine, will not present the same parameters as healthy urine, and the differences could affect the µFC performance. The performance obtained in this work with the ZnO-Ni anode and normal urine set the background for using diabetic urine. The urinalysis of the normal and diabetic urine used as fuel in the paper-based µFC is presented in Table 4. As can be seen, there are some differences in the composition of both samples. As was expected, the first is the presence of glucose in the diabetic urine; physical parameters such as pH and specific gravity also show light changes; another important difference is the appearance of piocytes, urethral epithelial cells, and calcium oxalate and there is a larger count of leukocytes and urethral cellularity in the diabetic urine. The presence of all these cells in the diabetic urine could lead to a poisoning or blocking effect in the catalyst or even a reaction competition (glucose oxidation) affecting the urea oxidation and therefore decreasing the performance of the paper-based µFC; however, it is also expected that the presence of the nickel complexes would favor the oxidation of urea present in the diabetic urine.
Table 4
Urinalysis of a healthy man (normal urine) and a diabetic man (diabetic urine)
| NORMAL URINE | DIABETIC URINE | REFERENCE VALUE |
MICROSCOPIC EXAM |
Color | Yellow | Yellow | |
Appearance | Clear | Clear | |
Sediment | Few | Few | |
CHEMICAL EXAM |
Specific gravity | 1.03 | 1.025 | |
pH | 5.0 | 6.0 | |
Glucose | negative | 2.3 g/dL | negative |
Proteins | negative | negative | negative |
Blood | negative | negative | negative |
Hemoglobin | negative | negative | negative |
Ketones | negative | negative | negative |
Bilirubin | negative | negative | negative |
Urobilinogen | 0.2 E.U./dL | 0.2 E.U./dL | 0.2 E.U./dL |
Nitrites | negative | negative | negative |
SEDIMENT MICROSCOPIC EXAM |
Erythrocytes | None seen | None seen | 0–2 /HPF |
Leukocytes | 0–1 /HPF | 4–5 /HPF | 0–2 /HPF |
Piocytes | None seen | 0–1 /HPF | |
Yeast | None seen | None seen | |
Bacteria | Few | Few | |
Bacterial cytolysis | Non seen | Non seen | |
Urine cast | None seen | Non seen | |
Mucus threads | Few | Few | |
Urethral epithelial cells (basal) | None seen | 0–1 /HPF | |
Urethral cellularity | 0–1 /HPF | 7–9/HPF | 0–2 /HPF |
Urothelium (Urethra) | None seen | None seen | |
Renal (bladder, ureters, renal pelvis) | None seen | None seen | |
Crystalluria (kidney tubules) | None seen | Moderate calcium oxalate dehydrates | |
lithogenic risk | | Absent | |
*Urinalysis was performed in “Precyso laboratorios clínicos” at San Juan del Río Querétaro México *Both samples were collected in a sterile vessel at the first time of the day |
The performance of the paper-based µFC is presented in Fig. 5. The polarization and power density curves using diabetic urine as fuel using the ZnO electrode are shown in Fig. 5a), while the results using ZnO-Ni (3,6 and 9%) are presented in Fig. 5b). The behavior is similar to the one described with normal urine. There is a notable difference between the ZnO and the ZnO-Ni electrodes; the performance with ZnO is lower than the observed with ZnO-Ni. The performance parameters using ZnO are presented in Table 5; the maximum voltage is 0.91 V, while the maximum current and power densities are .034 mA cm− 2 and .007 mW cm− 2, respectively. In this case, the polarization curve of ZnO (Fig. 5a) is also observed to be more resistive than when normal urine was used, which is probably due to the molecules present in this type of urine, which increment the internal current losses and that could be associated to blocking of the catalyst's active sites. Meanwhile, the polarization curves of the ZnO-Ni materials (Fig. 5b) do not show a different or more resistive behavior, thus indicating a significant tolerance to poisoning and blocking. The results obtained with ZnO-Ni and diabetic urine show again that the best performance is achieved with a 3% Ni concentration and that the current decreases while the Ni concentration increases; the ZnO-Ni (3%) presented a maximum voltage of 0.89 V, a maximum current density of 1.18 mA cm− 2 and a maximum power density of 1.26 mW cm− 2. It is observed that the maximum voltage obtained with normal and diabetic urine is quite similar, thus indicating that the ZnO-Ni is performing the urea oxidation even in the presence of glucose (diabetic urine). Therefore, the selectivity of the anodic catalyst could be highlighted. In this sense, the general decrease in current density when diabetic urine is employed may be due to the blockage of active sites caused by the presence of substances not observed in normal urine, such as glucose, calcium oxalate, or the change in the concentrations of leukocytes or urethral cells (Table 4), is important to consider that some other components non reported in the urinalysis could also affect the performance such as salts (phosphates, sulfate, etc.) and organic compounds as creatinine or uric acid [16]. Although the results do not exceed those obtained with normal urine as fuel (Table 3), this performance could be considered more accurate, considering that the purpose is the implementation of the µFC in portable devices for disease detection; therefore, the physiological fluid sample would not meet the normal standards.
Table 5
Performance parameters of paper-based µFC with ZnO-Ni anodes, using diabetic urine as fuel
Anode | Voltage (V) | Current density (µA cm− 2) | Power density (µW cm− 2) |
ZnO | 0.91 | .034 | 0.007 |
ZnO-Ni (3%) | 0.89 | 1.18 | 0.126 |
ZnO-Ni (6%) | 0.85 | 0.88 | 0.090 |
ZnO-Ni (9%) | 0.83 | 0.70 | 0.062 |