Introduction

Metal nanoparticles (M-NPs) have become essential functional materials in modern nanotechnology due to their size-dependent optical, magnetic, chemical, electronic, and mechanical properties that differ significantly from their bulk counterparts1. These characteristics enable applications in biomedicine, catalysis, electronics, environmental remediation, drug delivery, and energy systems2,3. Among transition metal oxides, nickel oxide nanoparticles (NiO NPs) are particularly attractive because of their high chemical stability, capacitance, electron-transfer efficiency, and electrocatalytic activity, supporting their use in sensors, photocatalysis, battery cathodes, fuel cells, and antimicrobial systems4,5,6,7,8. Despite these advantages, conventional physical and chemical synthesis routes often rely on hazardous reducing agents,high energy inputs, and costly instrumentation, generating toxic by-products and residual contaminants1,9,10,11. Such approaches conflict with the principles of sustainable production and resource efficiency emphasized in the United Nations Sustainable Development Goal 12 (Responsible Consumption and Production), which advocates minimizing waste generation and reducing reliance on hazardous substances. Consequently, there is growing interest in environmentally benign synthesis strategies that integrate green chemistry principles into nanomaterial fabrication. Plant-mediated synthesis represents a sustainable alternative, utilizing naturally occurring phytochemicals such as polyphenols and flavonoids as reducing and stabilizing agents12,13,14,15,16,17,18. This approach reduces chemical toxicity, lowers environmental impact, and promotes renewable resource utilization. However, several plant species reported for NiO NPs synthesis, including Piper betle, okra, ginger, and Calotropis gigantea, gum arabic are either geographically restricted or integral to the human food chain, raising concerns regarding long-term sustainability, supply competition, and ethical resource allocation19,20,21,22,23,24,25. The identification of non-food, widely available, and ecologically resilient plant sources is therefore essential to advancing sustainable nanoparticle production. Sida acuta (SA), an abundant and globally distributed weed, contains flavonoids, alkaloids, and polyphenols suitable for nanoparticle biosynthesis26,27,28,29. Its non-edible and invasive nature makes it an attractive sustainable biomass resource. Although SA extracts have been successfully employed for the green synthesis of Ag, CuO, and ZnO nanoparticles29,30,31, their application in NiO NPs synthesis has not been reported. Furthermore, the performance of NiO NPs is strongly governed by particle size, morphology and polydispersity, which depend on synthesis parameters such as extract concentration, reaction time, and temperature21,23,32. Traditional one-variable-at-a-time optimization strategies are inefficient and fail to capture interaction effects, potentially leading to suboptimal resource utilization15,21,23. Statistical optimization methods, including Taguchi design and Grey Relational Analysis (GRA), offer systematic, resource-efficient frameworks for multi-parameter control, aligning with sustainable production objectives33,34,35,36,37. Accordingly, this study advances SDG 12 principles by developing a plant-mediated, resource-efficient synthesis route for NiO NPs using SA extract and optimizing hydrodynamic diameter (HDD) and polydispersity index (PDI) via a Taguchi–Grey Relational Analysis (TGRA) approach. Extract concentration, reaction time, and temperature were varied according to an L9 orthogonal array, and the resulting nanoparticles were characterized using UV–Vis, FTIR, DLS, XRD, SEM/EDS, and TEM/SAED.

Materials and methods

Materials and apparatus

The chemicals (Ni (NO3)2.6H2O, HCl, NaOH) used in this study are of AR analytical grade supplied by Guangdong Guanghua Sci-Tech. Co. Ltd China. These chemicals were used as received. SA leaves used in this study were collected from Dogarawa Village, Sabon Gari Local Government Area, Kaduna State, Nigeria. The plant was identified and authenticated by Dr. Namadi Sunusi, a taxonomist at the Herbarium Unit, Department of Botany, Ahmadu Bello University, Zaria, Nigeria. A voucher specimen (Voucher No. ABUH090018) was prepared and deposited in the Herbarium Unit of the same department for future reference and verification. Permission for the collection of plant material was obtained from local authorities and landowners, in accordance with institutional and national guidelines. Deionized water obtained from Department of Pharmacognosy and Drug Development, Ahmadu Bello University Zaria, Nigeria was used as the solvent. Apparatus such as Beaker, whatman filter paper no. 1, conical flask and spatula were also employed in the synthesis.

Preparation of bio-reducing agent from SA leaf

Figure 1 shows the procedural sequence for the SA assisted synthesis of NiO NPs. For the preparation of the SA leaf extract: A batch of the matured plant was harvested from an open field near Dogarawa village, Sabon Gari Local Government Area, Kaduna State. The leaves were manually detached from their stems and processed following the methodology outlined by Hazarika et al.38 and Huimei et al.39. The leaves underwent thorough washing with tap water and de-ionized water to eliminate impurities. Subsequently, the cleaned leaves were shade-dried at ambient temperature for 28 days and then coarsely pulverized using a Sonik (1500 W/2.5 L) mechanical blender. The extraction process involved macerating 100 g of SA leaf powder in 1500 ml of deionized water for 24 h, with periodic stirring. After this period, the residue was separated by filtration using a sieve and Whatman No. 1 filter paper. The filtrate was then concentrated using a rotary evaporator (Model BUCHI, Model Number R110, England) and further dried in a water bath (Model BATH, Model Number B11, China) at 60 °C and finally in an oven for 24 h at 80 °C. The resulting extract residue was collected in an air tight container, and preserved in a freezer for subsequent use.

Green synthesis of NiO NPs

NiO NPs was synthesized according to the method prescribed by Uddin et al.40 but with slight adjustments in the concentration of the nickel salt and the reaction time. Using this method, 12.5 ml of aqueous SA leaf extract (12.8 mg/ml) was introduced into 20mM of Ni(NO3)26H2O solution and its pH adjusted to the value of 10 using HCl and NaOH in a dropwise manner. Typically, the mixture was stirred using magnetic stirrer (ST150SA-1500 rpm/100 °C) for 60 min at 60 °C. Change in the colour of the solution indicated the formation of NiO NPs. The mixture was centrifuged (Sorvall/Dupont centrifuge/ GLC-2B/ 6000 rpm) at 3000 rpm for 20 min, the residue was washed three times with deionized water and oven dried for 12 h at 80 °C and then calcined at 300 °C for 3 h.

Fig. 1
Fig. 1The alternative text for this image may have been generated using AI.
Full size image

Graphical illustration of SA-assisted NiO NPs production process.

Taguchi-GRA optimization of the NiO NPs production process

Taguchi orthogonal array

The HDD and the PDI of the biosynthesized NiO NPs were optimized considering factors such as reaction temperature (T), extract concentration (E) and reaction time (t), at three levels (1,2,3) each as shown in Table 1, using Taguchi-GRA. The reaction time and temperature levels were determined based on a preliminary study guided by the works of previous researches. Notably, the works of Kale et al.41, Javad et al.42 and Uddin et al.40 who suggested that plant-mediated synthesis of NiO NPs can be successfully carried out at temperatures ranging from 70 to 120 °C, with reaction times between 2 and 6 h.

Based on the three selected factors, each with three levels (as shown in Table 1), an L9 Taguchi Orthogonal Array was used. The experimental array, which includes the combinations of factor levels for each run, along with the corresponding performance characteristics (responses) for the produced NiO NPs, is presented in Table 2.

Table 1 Parameters and levels assigned for the NiO NPs optimization experiments.
Full size table
Table 2 Taguchi L9 orthogonal array for the NiO NPs production.
Full size table

Procedure for GRA

In GRA, multiple quality characteristics were transformed into a single response, usually referred to as grey relational grade (GRG). The GRGs were determined by separately normalizing each of the responses using the appropriate objective function. The obtained GRGs were then used as basis to optimize the process parameters. Similar to the Taguchi method, GRA involves selecting an objective function that suits the target goal of each response during the pre-processing of the experimental data. In this study, both responses have the same objective function, which is “the smaller, the better” as expressed in Eq. 1.

$${x}_{i}^{\pi}\left(k\right)=\frac{max.{x}_{i}^{\left(O\right)}\left(k\right)-{x}_{i}^{\left(O\right)}\left(k\right)}{max.{x}_{i}^{\left(O\right)}\left(k\right)-min.{x}_{i}^{\left(O\right)}\left(k\right)}$$
(1)

where \({x}_{i}^{\pi}\left(k\right)\) = Grey relational generated value, \(min.{x}_{i}^{\left(O\right)}\left(k\right)\) = Least value for \({x}_{i}^{\left(O\right)}\left(k\right)\) in the kth response, \(max.{x}_{i}^{\left(O\right)}\left(k\right)\) = Highest value for \({x}_{i}^{\left(O\right)}\left(k\right)\) in the kth response.

The normalized experiment data was used to calculate the grey relational coefficient (GRC) using Eq. (2). The calculated GRC, \({\epsilon}_{i}\left(k\right)\) expresses the relationship between the desired and the actual experimental data.

$$\varepsilon _{i} \left( k \right) = \frac{{\Delta \min - \zeta \times \Delta x\max }}{{\Delta _{{0i}} \left( k \right) - \zeta \times \Delta x\max }}$$
(2)

where \(\Delta _{{0i}} \left( k \right) = \left| {x_{0}^{*} \left( k \right) - x_{i}^{*} \left( k \right)} \right|~\)is the absolute value of the difference between the reference sequence \(x_{0}^{*} \left( k \right)\) and the comparable sequence, \(x_{i}^{*} \left( k \right)\); \(\zeta\)= Distinguishing coefficient (0.5); \(\Delta \min\) = Smallest value of \(\Delta _{{0i}} \left( k \right)\); \(\Delta \max\) = Largest value of \(\Delta _{{0i}} \left( k \right)\).

It is worth noting that in this study, both the HDD and PDI were given equal importance, therefore, The GRG was obtained as the average of the GRCs of the individual responses, using Eq. (3).

$${Y}_{i}=\frac{1}{n}\sum_{k=1}^{n}{\varepsilon}_{i}\left(k\right)$$
(3)

where: Yi is the GRGs for the ith experiment.

The values for the GRGs were used to analyse the optimal processing parameters that yield the best quality characteristics. The values of the GRGs were ranked from highest to lowest, and the highest value corresponds to the experimental parameters that provide the quality characteristics that is close to the desired quality. The optimum factor levels that gave the desired quality characteristics were determined by mean grey relational grades (MGRGs). This involves averaging the GRGs corresponding to each of the factors at each of their levels. The levels that have the highest MGRGs are selected as the optimum process parameters. Upon identifying the optimum process parameters, the GRGs of the measured HDD and PDI of the NiO NPs synthesized using the optimum synthesis parameters was predicted using Eq. 4 and was compared with the GRG computed from the NiO NPs response obtained experimentally.

$$\widehat{\gamma}={\gamma}_{m}+\sum_{i=1}^{k}\left({\stackrel{-}{\gamma}}_{i}-{\gamma}_{m}\right)$$
(4)

Where \(\widehat{\gamma}\) = the estimated GRGs computed using the optimal level for each of the factors, γm = The average value of the GRGs for all the experimental runs, \(\bar{\gamma }_{i}\) = The MGRGs at the optimal level for each factor and k = The number of factors that have considerable effect on the performance characteristics.

The quantitative percentage contributions and the significance of each of the production parameters on the target characteristics were analysed using ANOVA. The experimental design and optimization process were hasten using free version Minitab software 16.

The relationship between the GRGs and the selected process parameters (extract concentration, reaction time and temperature) were modelled using linear regression model. The model was fitted to the experimental data to predict GRGs values based on the selected factors. The goodness-of-fit was evaluated using the coefficient of determination (R2) and adjusted R2. The GRGs predicted using the model and GRGs computed from experimental data were compared, to determine the extend with which the model can accurately predict GRGs within the experimental domain.

Characterization of the SA-NiO NPs produced

The formation of NiO NPs was confirmed using a UV-2500PC (v2.30) spectrophotometer, with absorbance measured in the range of 200–600 nm. The optical band gap energy of the SA-mediated NiO NPs was estimated from UV–Vis spectroscopy data using a Tauc plot, as described by Eqs. (5–7).

$$\left( {\alpha hv} \right) = C\left( {hv - E_{g} } \right)^{n}$$
(5)
$$hv=\raisebox{1ex}{$1240$}\!\left/\!\raisebox{-1ex}{$\lambda$}\right.$$
(6)
$$\alpha=2.3A$$
(7)

where C is a constant usually taken as one, λ is the absorption wavelength, Eg is the average band gap energy of the material, α is the reflection coefficient, A is the absorption, n is a value whose magnitude depend on the transition. As for direct allowed transition, n is taken as 1/2 and the average band gap Eg is obtained by interpolating the linear portion of the (αhv)2 vs. hv plot.

FTIR spectroscopy was conducted using a Shimadzu FTIR-8400 S to identify the functional groups involved in the reduction and stabilization of the NPs. The prepared sample was scanned in the range of 4000–650 cm−1 after 15 min of chamber evacuation. Particle size and distribution were measured using a NanoS dynamic light scattering (DLS) instrument. X-ray diffraction (XRD) analysis was performed using a Miniflex 300/600 + diffractometer with Cu Kα₁ radiation (λ = 1.540 Å) over a 2θ range of 5°–80° and a step size of 0.02°. Crystallite size and interplanar spacing were calculated using Scherrer’s and Bragg’s equations, respectively (Eqs. 8 and 9).

$$D=\raisebox{1ex}{$K\lambda$}\!\left/\!\raisebox{-1ex}{$Lcos\theta$}\right.$$
(8)

And the interplanar spacing in the crystal structures was calculated using Bragg’s formula given in Eq. 943.

$$d=\raisebox{1ex}{$n\lambda$}\!\left/\!\raisebox{-1ex}{$2\text{sin}\theta$}\right.$$
(9)

where D is the average crystallite size, K is a constant whose value depends on the shape of the particles, λ is the wavelength of the X-ray (1.5409), L is the full width at half maximum in radians, d is the interplanar spacing, \(\theta\) is the Bragg angle in radian and n is a positive integer depicting the order of the x-ray.

Morphological and elemental analyses were carried out using a Phenom Pharos G2 scanning electron microscope (SEM) equipped with energy-dispersive X-ray spectroscopy (EDS). Further structural analysis was performed using transmission electron microscopy (TEM) and selected area electron diffraction (SAED). The TEM and SAED Imaging and diffraction were carried out at an accelerating voltage of 200 kV. ImageJ software was used to analyze particle size, size distribution, and lattice characteristics.

Results and discussion

Production of SA-NiO NPs

Figure 2 shows the aqueous mixture of Ni(NO3)2·6 H2O and SA leaf extract (a) before and (b) after 60 min of stirring at 70 °C.

Fig. 2
Fig. 2The alternative text for this image may have been generated using AI.
Full size image

Solution mixture of SA leaf extract with (a) Ni(NO3)26H2O before reaction (b) NPs NPs after reaction.

Figure 2 shows the aqueous mixture of Ni(NO3)2·6 H2O and SA extract before and after 60 min of stirring at 70 °C. The colour change from brown to green suggests the reduction of Ni2+ ions to Ni0 atoms by SA phytochemicals, which then served as nucleation sites for the formation of Ni(OH)2 or related intermediates. These intermediates subsequently decomposed during calcination to yield NiO NPs. Although Ni does not display strong SPR features like noble metals, the colour shift likely reflects changes in electronic structure or particle size that influence light absorption44. Similar green colour transition has been reported in plant-mediated syntheses of NiO NPs, including those using Clitoria ternatea45, Abutilon indicum46, and other botanical extracts25,41,47. This consistent observation across studies supports the use of colour change as an initial qualitative indicator of Ni-based nanoparticle formation in phytochemical-assisted syntheses. However, such visual cues remain preliminary and must be corroborated by spectroscopic and structural characterization to confirm the composition and phase of the resulting NPs.

Characterization of the NiO NPs synthesized with SA-reducing agent

UV–visible spectrophotometric analysis

Figure 3 shows the UV–Vis spectra of the SA leaf extract and the synthesized SA-NiO NPs. The SA extract exhibits a broad absorption peak at 270 nm, consistent with the presence of hydroxyl-rich metabolites such as flavonoids and polyphenols20,46. After reaction and calcination, the NiO-containing sample displays a broad absorption feature centered at 284 nm, which falls within the reported 250–385 nm range for NiO NPs45. Comparable absorption maxima have been documented for plant-mediated NiO NPs, including those synthesized using Avicennia marina (297 nm), guava leaf (266 nm), Bauhinia tomentosa (280 nm) and Croton macrastachyus extracts (230–300 nm)48,49,50,51. These variations are typically associated with electronic transitions from the valence band to the conduction band in NiO and are influenced by particle size, surface states, and the band-gap energy. Smaller particles often exhibit blue-shifted absorption due to quantum confinement, while synthesis conditions can also alter optical features. Imran et al.21 highlighted that synthesis parameters strongly affect these transitions, leading to differences in optical behavior across studies. Additionally, residual SA phytochemicals acting as capping agents may contribute to absorption in the UV region due to their intrinsic chromophore activity. The optical band gap of the SA-NiO NPs, calculated from the Tauc plot (Fig. 3), was found to be 4.8 eV. This value is higher than the commonly reported 3.2–3.4 eV range for biogenic NiO NPs, such as those synthesized using Limonia acidissima fruit extract (3.26–3.38 eV)23 or Phytolacca dodecandra leaf extract (3.19 eV)15. However, similarly elevated band-gap values (e.g., 4.78 eV) have been reported for NiO NPs produced with Gymnema sylvestre leaf extract52. Higher band gaps may result from quantum confinement, defect states, phytochemical binding, or differences in synthesis and annealing conditions23,45,52. The relatively large band gap observed here suggests that the SA-NiO NPs may exhibit enhanced performance in UV-absorbing coatings, optoelectronic devices, and photocatalytic applications requiring wide-band-gap semiconductors.

Fig. 3
Fig. 3The alternative text for this image may have been generated using AI.
Full size image

UV–Vis spectra of aqueous solutions of SA leaf (blue) and NiO NPs (red).

Functional group on the NiO NPs

FTIR analysis was performed on both the SA leaf extract and the synthesized NiO NPs to identify the functional groups involved in the reduction, capping, and stabilization of the NPs. Spectra were recorded in the 4000–650 cm−1 range. As shown in Fig. 4, the FTIR spectrum of the SA leaf extract (i) displays characteristic absorption peaks at approximately 3250, 2922.2, 1580, 1390, 1032, and 890 cm−1. The broad band between 3500 and 3350 cm−1 corresponds to O–H stretching vibrations of hydroxyl groups, which are typically associated with flavonoids, phenolics, and other polyhydroxylated metabolites present in the extract20,46. The weak band observed at 2895 cm−1 is assigned to the C–H stretching of methyl groups, while peaks around 2100 cm−1, 1580 cm−1, and 1390 cm−1 correspond to the C= C stretch of alkynes, the C =C stretching of aromatic compounds, and the N–O stretching of nitro functional groups, respectively. A peak at 1032 cm−1 appears in the fingerprint region and is attributed to C–F stretching vibrations of alkyl or aryl halides. These spectral features are consistent with previously of reported FTIR analyses of SA leaf extracts29,31,53, which highlight the abundance of phytochemicals capable of reducing metal ions and subsequently stabilizing the resulting NPs. The FTIR spectrum of the NiO NPs synthesized using the SA extract shows prominent absorption peaks at approximately 3374, 2324, and 1036.2 cm−1, corresponding to O–H, C≡ N, and C–F functional groups, respectively. A comparison of the FTIR spectra of the SA extract and the NiO NPs reveals noticeable changes: several peaks either disappear, diminish in intensity, or exhibit slight shifts from their original wavenumbers. Specifically, the disappearance of C–H and C=C functional groups, along with reduced intensity in the O–H, C=C, and N–O peaks, suggests their participation in the reduction of Ni2+ ions and their subsequent interaction with the nanoparticle surface. The shift in the O–H and N–O absorption bands indicates the possible formation of metal–oxygen bonds or coordination complexes between nickel species and phytochemical constituents during nanoparticle nucleation and growth. These functional groups, particularly O–H, C=C, C=C, and N–O, are known to donate electrons, facilitating the reduction Ni2+ to Ni0, which serves as an intermediate prior to NiO formation. The decrease in N–O stretch intensity may reflect reduction of nitro-containing compounds or their transformation during the synthesis process, further supporting their active involvement in nanoparticle formation. These results are consistent with previous studies by Ullah et al.54 and Shaik et al.32, who reported similar functional group modifications and spectral shifts during plant-mediated NiO NPs synthesis. The broadening and shifting of peaks in the NiO NPs spectrum confirm that SA-derived phytochemicals act as effective capping agents, contributing to nanoparticle stability and preventing agglomeration by forming protective coatings around the particle surfaces55. Based on the FTIR analysis, a plausible reaction mechanism for the synthesis of NiO NPs using SA leaf extract is proposed and illustrated in Fig. 5, highlighting the roles of key functional groups in reduction, stabilization, and nanoparticle formation.

Fig. 4
Fig. 4The alternative text for this image may have been generated using AI.
Full size image

FTIR spectra of SA leaf extract and NiO NPs.

SA-NiO NPs production mechanism

Phytochemical analysis of SA extracts has confirmed the presence of alkaloids, polyphenols, flavonoids, saponins, tannins, and phenolics, even in aqueous extracts suitable for green synthesis28,30. Key compounds such as chrysin, cryptolepine, kaempferol, xanthyletin, N-feruloyltyramine, amnosmonin, cryptolepinone, luteolin, and acacetin contain electron-rich functional groups capable of reducing metal ions56. Upon introducing SA extract into aqueous nickel nitrate (Fig. 5), Ni2+ ions interact primarily with flavonoids and polyphenols, which donate electrons via hydroxyl and carbonyl groups, forming Ni(OH)2 intermediates. These subsequently nucleate and convert to NiO NPs. Concurrently, phytochemicals adsorb onto the particle surface, providing capping and steric stabilization that regulate particle size and prevent agglomeration57. This dual reduction–capping function aligns with plant-mediated mechanisms reported for other biosynthesized NiO NPs18.

Fig. 5
Fig. 5The alternative text for this image may have been generated using AI.
Full size image

proposed mechanism for the synthesis of NiO NPs using SA leaf extract as reducing.

Agent.

Density Functional Theory (DFT) calculations employed using Avogadro and Gaussian 9 W software further support this mechanism by revealing the electronic properties of the phytochemicals and their interactions with Ni2+. In their pristine state (Table S1), HOMO energies range from − 5.31 to − 6.01 eV, LUMO energies from − 2.09 to − 2.87 eV, and energy gaps (ΔE) from 2.62 to 3.77 eV. Cryptolepine (ΔE = 2.62 eV) and Amnosmonin (ΔE = 2.69 eV) show smaller gaps, indicating higher chemical softness and greater charge-transfer potential, consistent with their ability to donate electrons readily58. In contrast, Luteolin (3.77 eV) and Chrysin (3.69 eV) display larger gaps, reflecting greater stability and lower spontaneous reactivity. After interaction with aqueous Ni(NO3)2 (Table S2), most compounds exhibit less negative HOMO energies and reduced ΔE, indicating enhanced electron-donating ability and coordination potential. For example, Chrysin’s ΔE decreases from 3.69 to 3.22 eV, while Kaempferol (3.66 → 3.34 eV) and Acacetin (3.51 → 2.95 eV) show similar reductions, reflecting effective orbital overlap and charge delocalization with Ni2+59. Cryptolepine and Cryptolepinone exhibit minimal ΔE changes, suggesting weaker electronic perturbation or non-covalent interactions. Interestingly, Amnosmonin shows a slight ΔE increase, indicating stabilization rather than enhanced softness. Overall, the DFT results confirm that the reduction in HOMO–LUMO gaps for most phytochemicals enhances electronic softness and metal-binding ability, providing strong evidence that these compounds facilitate Ni2+ reduction, nucleation, and stabilization during NiO NP formation.

HDD and PDI of the produced NiO NPs

The HDD and the PDI of the SA-NiO NPs produced was analysed using Dynamic Light Scattering spectroscopy (Malvern Nano-S zetasizer equipment).

Fig. 6
Fig. 6The alternative text for this image may have been generated using AI.
Full size image

DLS analysis of NiO NPs synthesized using SA leaf extract.

As shown in Fig. 6, the HDD and PDI of the produced NiO NPs were 106.1 nm and 0.425 respectively. The PDI value (see Fig. 6) suggests broad particle size variation of the HDD. The value was close to the PDI value of 0.401 reported by Sudhasree et al.60, who synthesized Ni NPs with an HDD of 2695 nm using Desmodium gangeticum DC root extract as a reducing and capping agent. However, the larger HDD reported in their study demonstrated the absence of much smaller particles or high aggregation of particles in the sample. In a separate study, Prabhu et al.45 and Iqbal et al.61, reported a PDI of 1.0 for biosynthesized NiO NPs having an HDD of 174 and 119 nm, respectively. Generally, samples with a PDI of \(<\)0.05 are classified as monodispersed while samples with a PDI of \(>\)0.7 implies particles having broad size distribution (i.e. highly polydispersed) which affects their performance62. Smaller particle size offers a high surface area to volume ratio, which in turn modifies their resulting physical, mechanical, and chemical properties63,64,65. For these reasons, the HDD and PDI of the SA leaf extract mediated synthesized NiO NPs was optimized by adjusting process parameters, such as extract concentration, reaction temperature, and reaction time, using the Taguchi L9 orthogonal array in conjunction with Grey Relational Analysis.

Particle size and PDI optimization

The Taguchi orthogonal array was adopted for the experimental design, while the responses were analysed using GRA as the Taguchi method does not accurately optimize more than one performance characteristics37. Figure 7 presents experimental parameters and the performance characteristics (responses) which are the HDD and the PDI of the NPs in aqueous solutions as obtained from the DLS analysis. The nine experiments conducted are referred to by their serial number (see Table 2). As can be seen in Fig. 7, the least PDI and the highest HDD were observed with experiment number one (1). While, experiment number eight (8) gave the least HDD and the highest PDI value for the produced SA-NiO NPs. The least PDI (0.204) at the HDD of 77.96 nm for the produced NiO NPs was observed with experimental parameters corresponding to experiment number nine (9). GRA was used to combine the effect of these responses into a single value (GRGs) based on the desired targeted characteristics. This single value (see Fig. 8) was used as a basis for the optimization.

Fig. 7
Fig. 7The alternative text for this image may have been generated using AI.
Full size image

Variation of HDD and PDI of the NiO NPs produced with experimental runs.

Fig. 8
Fig. 8The alternative text for this image may have been generated using AI.
Full size image

Variation of the calculated GRGs with experimental runs.

Based on the GRGs for the biosynthesized NiO NPs, as seen in Fig. 8, experiment number nine (9) has the best ranking with the highest GRGs value of 0.736671 indicating that the process parameters corresponding to experiment nine provide NiO NPs with the best combination of balanced characteristics. However, the optimal combination of experimental conditions may not necessarily be included in the Taguchi orthogonal arrays used for the experimentation. In such a case, the experiment settings that have the best rankings are only closer to the optimum parameters that would offer the desired characteristics66. Mean grey relation grades (MGRGs) generated from the data in Fig. 8, as shown in Table 3, provide the optimum combination of experimental settings with potentials for best performance characteristics. Factor levels with the highest MGRGs (see bold values) correspond to the optimum levels for those factors.

Table 3 MGRGs for NiO NPs production.
Full size table

From this analysis that considered only the main effect of individual factors, the combination of experiment parameters that would produce NiO NPs with optimum HDD and PDI using SA extract as reducing agent (see bold values in Table 3) were found to be extract concentration at level 3, reaction temperature at level 1 and reaction time at level 3 with the process parameter values corresponding to 60 mg/ml, 70 °C and 120 min respectively.

Fig. 9
Fig. 9The alternative text for this image may have been generated using AI.
Full size image

Contour plot of MGRGs vs. (a) temperature and extract (b) temperature and time (c) time and extract for the SA assisted production of NiO NPs.

The effect of interactions between the independent process parameters on the responses were examined using contour plots, as shown in Fig. 9. These interaction contour plots indicate that NiO NPs with higher GRGs can be synthesized by adjusting the extract concentration, reaction temperature, and reaction time within the ranges of 50–60 mg/ml, 70–75 °C, and 115–120 min, respectively. This observation aligns with the optimal process parameters predicted by the main effect data presented in Table 3. These results suggest that the interaction between the parameters does not significantly influence the characteristics of the produced NPs as affirmed by the mathematical model developed. However, it is important to note that these specific combinations of factor levels for SA-NiO NPs production do not correspond to any of the experimental runs shown in Fig. 2, highlighting the need for a confirmatory experiment.

Confirmatory experiment

The optimum process parameters identified in Table 3 were validated through a confirmatory experiment. Prior to experimentation, the GRG for the optimal factor combination was predicted using Eq. 4 and subsequently compared with the experimentally obtained GRG under identical conditions. As presented in Table 4, the predicted GRG value (0.788095) was in close agreement with the experimental value (0.752202), yielding a relative error of 4.55%, thereby confirming the reliability of the optimization model. Under the optimal conditions (E3, t3, T1), the HDD and PDI values were reduced to 62.72 nm and 0.232, respectively.

Table 4 Comparative results of NiO NPs produced with initial, best experimental and the optimize conditions.
Full size table

Compared with the initial experimental condition (E1, T1, t1), the HDD decreased from 106.1 nm to 62.72 nm (40.89% reduction), while the PDI decreased by 45.17%, indicating enhanced particle size uniformity and dispersion stability. The significant improvement in nanoparticle characteristics can be attributed to the high concentration of SA leaf extract at 70 °C. At this temperature, a favorable balance between nucleation and growth is achieved. The elevated phytochemical concentration accelerates reduction kinetics, promoting rapid nucleation of Ni atoms65,67. Simultaneously, non-reactive phytoconstituents act as stabilizing and capping agents, inhibiting excessive particle growth and preventing agglomeration68. In contrast, at higher temperatures (e.g., 80 °C), although nucleation rates increase, the reduced viscosity of the reaction medium may weaken the capping efficiency of the extract. This can slightly enhance the growth rate constant, leading to marginal increases in particle size, as also reported by Liu et al.67 during silver nanoparticle biosynthesis. While prolonged reaction times generally favor particle growth, the strong capping effect at high extract concentration effectively suppressed excessive growth under the optimized condition.

Regression analysis

Linear regression analysis was performed on the experimental data using Minitab version 16 software. This analysis enabled the development of a mathematical model capable of predicting performance characteristics both within and beyond the experimental domain. The significance and contribution of each experimental parameter to the measured responses were assessed through analysis of variance (ANOVA) at a 95% confidence level. As shown in Table 5, all factors (extract concentration, reaction temperature, and reaction time), along with the interaction terms, were statistically significant, with corresponding p-values below 5%. Among the factors, temperature exhibited the highest contribution to response variability at 33.10%, followed by extract concentration (27.81%), reaction time (11.60%), and the interaction effects (8.94% and 18.54%). These results confirm the dominant roles of thermal energy and phytochemical concentration in governing nanoparticle nucleation and growth processes. The predictive performance of the regression model (Eq. 7) was evaluated using the coefficient of determination (R2). The model demonstrated strong predictive capability, yielding high values for R2 (99.39%), adjusted R2 (98.36%), and predicted R2 (93.76%). The difference between the adjusted and predicted R2 values was 4.68%, which, according to Dan-Asabe et al.69, is well within the acceptable threshold of 20%, indicating a reliable and robust model.

Table 5 ANOVA for GRA NiO NPs.
Full size table
$${\text{GRG }}\left( {{\text{NiO NPs}}} \right)0.{\text{78}}0 + 0.0{\text{1281E}} - 0.00{\text{823t}} - 0.00{\text{3992T}} - 0.0000{\text{9}}0{\text{E}}^{{\text{2}}} + 0.0000{\text{58t}}^{{\text{2}}}$$
(10)

The GRG predicted by the model was compared with the GRGs obtained from the experimental data, as shown in Fig. 10. The predicted GRGs closely matched the experimental values, with the percentage error for each run being less than 2.2%. This small deviation confirms the robustness of the model and its suitability for accurately predicting the production characteristics of NiO NPs within the experimental domain70.

Fig. 10
Fig. 10The alternative text for this image may have been generated using AI.
Full size image

Comparison between the GRGs calculated and the GRGs predicted for the SA assisted NiO NPs produced.

Phase and crystallite size analyses of the biosythesized NiO NPs

Figure 11 presents the X-ray diffraction (XRD) pattern of (salicylic acid) SA-assisted biosynthesized NiO NPs after calcination at 300 °C for 2 h.

Fig. 11
Fig. 11The alternative text for this image may have been generated using AI.
Full size image

X-Ray Diffraction pattern of SA assisted biosynthesis of NiO NPs.

The X-ray diffraction (XRD) pattern exhibits distinct peaks at 2θ values of 37.55°, 43.64°, 63.34°, 75.82°, and 78.72°, corresponding to the (111), (200), (220), (311), and (222) crystallographic planes, respectively. These reflections confirm the formation of a face-centered cubic (FCC) crystal structure, in agreement with the standard NiO reference pattern (JCPDS No. 47-1049), which shows 2θ peaks at approximately 37°, 43°, 63°, 75°, and 79°. The average crystallite size of the SA- NiO NPs was estimated to be approximately 6 nm using the Scherrer equation (Eq. 8; see Table 6). At this nanoscale, quantum confinement effects become significant, which may explain the observed increase in band gap energy68. The observed peak positions are in close agreement with previous green synthesis studies of NiO NPs, such as those reported by Berhe et al.71 using Calpurnia Aurea Leaf extract, Kumar et al.23 using Limonia acidissima fruit extract, Prabhu et al.45 with Clitoria ternatea, and Haidar et al.25 using Zingiber officinale and Allium sativum. However, slight shifts in the peak positions relative to the standard pattern were observed. These deviations can be attributed to several factors, including: (i) crystallite size reduction, which induces peak broadening and shifts due to size-dependent lattice contraction; (ii) lattice strain arising from defects or dislocations during synthesis or calcination; and (iii) presence of phytochemicals from the plant extract, potentially leading to surface doping or chemical modification72. The corresponding interplanar spacings (d-values), calculated using Bragg’s law, were 2.3935 Å, 2.068 Å, 1.467 Å, 1.254 Å, and 1.217 Å for the respective diffraction planes. These values further confirm the crystalline nature of the synthesized NiO NPs.

Table 6 Crystallite sizes calculation for the XRD diffraction Peaks of the produced NiO NPs.
Full size table

Surface morphology and elemental compositions of the SA-NiO NPs

Figure 12a presents an SEM micrograph showing the surface morphology of the synthesized SA-NiO NPs at 2000× magnification.

Fig. 12
Fig. 12The alternative text for this image may have been generated using AI.
Full size image

(a) SEM micrograph showing surface morphology and aggregated structure of SA leaf extract-synthesized NiO NPs at 2000× magnification. (b) EDS spectrum revealing elemental composition and confirming the presence of Ni and O, along with minor elements from the plant extract.

The SEM image reveals particles exhibiting crystalline and polyhedral features, with sizes ranging approximately from 2 to 7 μm. It is important to note that these relatively large particle sizes indicate that the observed structures are aggregates formed by the agglomeration of smaller primary NPs, which is common in nanomaterial synthesis due to van der Waals forces or magnetic interactions. The primary NPs size likely falls below the micron scale, but SEM imaging at this magnification mainly reveals the aggregated morphology rather than individual NPs. For more accurate primary particle sizing, complementary techniques such as Transmission Electron Microscopy (TEM) was employed. The irregular polygonal shapes observed align well with similar NiO NPs morphologies reported by Firisa et al.15, Govindasamy et al.73, and Pandian et al.24, further supporting the reliability of the synthesis process. Figure 12b displays the Energy Dispersive X-ray Spectroscopy (EDS) spectrum of the synthesized NiO NPs, revealing the elemental composition. Key elements detected include oxygen, phosphorus, calcium, sodium, chlorine, magnesium, silicon, aluminium, iron, and titanium. The presence of these elements, particularly those aside nickel and oxygen, can be attributed to residual components originating from the SA leaf extract used during synthesis46,53. This suggests that trace amounts of phytochemicals and minerals from the SA extract remained on the nanoparticle surfaces or were incorporated within the aggregates. The EDS analysis indicated a NiO NPs purity of approximately 80 wt%, with about 45 wt% attributed to the Ni element. This value is notably higher than the 43.72 wt% reported by Barzinjy et al.74. Similarly, Prabhu et al.45 reported a comparable overall purity (81 wt%) for NiO NPs synthesized using Clitoria ternatea flower extract, but with a higher Ni content (80.68 wt%). It is worth noting that EDS is a surface-sensitive technique and the purity value reflects the relative elemental composition near the surface; therefore, residual organic or inorganic species from the extract may contribute to the detected impurities which could possibly be the reason for the shift observed in XRD peaks. Further purification or washing steps could potentially improve this purity and thus, its electronic and catalytic activities. The prominent oxygen peak observed in the EDS spectrum corroborates the XRD analysis results, which confirm the successful synthesis of crystalline NiO NPs. The consistency between these characterization techniques supports the conclusion that the material produced is predominantly NiO with minor residual elements from the biological synthesis method.

Particles size, size distribution and crystalline nature of NiO NPs

Figure 13 presents TEM analyses of NiO NPs synthesized using SA leaf extract as a reducing and stabilizing agent. The TEM micrograph in Fig. 13a reveals that the NiO NPs are predominantly spherical and fairly-dispersed, although some degree of agglomeration is evident. The particles exhibit sizes ranging from approximately 3–13 nm. This size distribution, determined by measuring 50 randomly selected particles, is summarized in Fig. 13b, yielding an average particle size of 6.55 nm. Interestingly, the average particle size obtained from TEM is slightly larger than the crystallite size (5.99 nm) calculated from XRD data. This discrepancy is commonly observed in nanomaterials and could be attributed to particle agglomeration or the presence of multi-crystallite particles, where several crystallites form a single nanoparticle. Overlapping and aggregation of particles clearly observed in Fig. 13a, may also contribute to this difference. Similar observations have been reported by Santhi et al.75, who noted that agglomeration is often facilitated by the inherent magnetic interactions among particles, as also supported by Wang et al.76. Additionally, residues from the SA leaf extract appear to be present on or around the NPs as observed in the TEM image. This is consistent with the EDS findings (Fig. 12b), which revealed the presence of various elements associated with the extract, and with XRD data that suggested secondary components. The organic material may have contributed to the observed particle stability and dispersion, despite minor agglomeration. The selected area electron diffraction (SAED) pattern shown in Fig. 13c exhibits a ring structure composed of distinct diffraction pots. This pattern confirms the polycrystalline nature of the NiO NPs, indicating that the NPs may consist of multiple crystalline domains. The diffraction rings correspond to the crystallographic planes of FCC NiO NPs, which agrees with XRD results. Similar SAED patterns have been reported by Kumar et al.23 for polycrystalline NiO structures. These structural and nanoscale features are expected to enhance catalytic activity through increased surface area and active sites, while improving electronic performance via defect-mediated charge transport pathways.

Fig. 13
Fig. 13The alternative text for this image may have been generated using AI.
Full size image

(a) TEM micrograph showing spherical and dispersed SA-NiO NPs with some agglomeration. (b) Particle size distribution histogram showing an average size of 6.55 nm. (c) SAED pattern indicating polycrystalline nature of the NiO NPs.

Conclusion

This study optimized the green synthesis of NiO NPs using SA leaf extract through Taguchi–Grey Relational Analysis. The optimum conditions which are 60 mg/mL extract concentration, 70 °C reaction temperature, and 120 min reaction time, yielded NPs with an HDD of 62.72 nm and a PDI of 0.232. ANOVA identified reaction temperature as the most significant factor, followed by extract concentration and reaction time. Characterization confirmed polycrystalline NiO NPs with FCC structure, an average crystallite size of 5.99 nm, a UV–Vis absorption peak at 285 nm, and a band gap of 4.78 eV. SEM, EDS, and TEM analyses revealed fairly dispersed particles with about 80% elemental purity. The wide band gap and strong UV absorption highlight the potential of these NPs for UV photodetectors, UV-blocking coatings, photocatalysis, dielectric materials, and gas sensors. Overall, SA-mediated synthesis provides an eco-friendly and effective route for producing NiO NPs with desirable physicochemical properties, warranting further application-specific studies.