Electrochemical Deposition of Ferromagnetic Ni Nanoparticles in InP Nanotemplates Fabricated by Anodic Etching Using Environmentally Friendly Electrolyte

Porous InP templates possessing a thickness of up to 100 µm and uniformly distributed porosity were prepared by anodic etching of InP substrates exhibiting different electrical conductivities, involving an environmentally friendly electrolyte. Ni nanoparticles were successfully directly deposited by pulsed electroplating into prefabricated InP templates without any additional deposition of intermediary layers. The parameters of electrodeposition, including the pulse amplitude, pulse width and interval between pulses, were optimized to reach a uniform metal deposition covering the inner surface of the nanopores. The electrochemical dissolution of n-InP single crystals was investigated by measuring the current–voltage dependences, while the Ni-decorated n-InP templates have been characterized by scanning electron microscopy (SEM) and energy-dispersive X-ray spectroscopy (EDX). The proposed technology is expected to be of interest for sensing and photocatalytic applications, as well as for the exploration of their plasmonic and magnetic properties.


Introduction
Various porous templates, especially those with ordered arrangement of pores, have found a series of applications in nanofabrication [1]. Two types of templates are most widely used for nanofabrication purposes nowadays, namely porous alumina templates produced by the anodization of aluminum foils [2,3], and etched ion track membranes based on either inorganic materials or organic polymers [4,5]. In recent years, semiconductor nanotemplates have emerged as a prospective basis for the templated fabrication of nanowires and nanotubes from various materials, as well as of composite nanomaterials [6]. Wide possibilities to control the electrical conductivity of semiconductor templates, e.g., by means of external illumination and applied electric fields, are among their most important advantages, especially when the preparation of metallic nanotubes with a controlled diameter and thickness of the walls is envisaged.
Usually, the pores in semiconductors are introduced via anodization in electrolytes containing acids such as HF, HCl, H 2 SO 4 , HNO 3 , etc., or in alkaline electrolytes [7,8]. In the last decade, to make the process of nanofabrication based on anodic etching broadly accessible and environmentally friendly, research has focused on nanostructuring in neutral electrolytes based on an aqueous solution of NaCl, instead of the commonly used aggressive acids or alkaline electrolytes, for the purpose of the electrochemical nanostructuring of semiconductor substrates. Anodization in neutral electrolytes has proven to be feasible

Materials and Methods
Crystalline 500-µm thick n-InP(100)-oriented substrates with the free electron concentration of 2 × 10 17 cm −3 and 2 × 10 18 cm −3 , supplied by University Wafer, Boston, MA, USA, were used in this study. Anodization of n-InP crystals was carried out in 1.75 M and 3.5 M NaCl aqueous solutions for 180 s, in a potentiostatic mode, under different applied voltages, their selection being based on the I-V curves. The anodization was performed in a common two-electrode cell, where the sample served as a working electrode as described elsewhere [29]. Briefly, an electrical contact with a conductive silver paste was made on the backside of the samples. From the top side, the sample was pressed by an O-ring to leave a surface with an area of 0.2 cm 2 exposed to the electrolyte. A mesh of platinum wire with 0.5 mm diameter was used as the counter electrode. An AUTOLAB Potentiostat/Galvanostat/EIS/ECD (Metrohm Autolab B.V., Utrecht, The Netherlands) was used to record the polarization curves with a scan rate of 10 mV·s −1 before the electrochemical etching, as well as chronoamperometry curves recorded during the anodization process. After the pore growth, the top nucleation layer of porous InP samples was removed by isotropic wet etching, immersing the specimen in 1:1 (vol.) HCl:H 3 PO 4 mixture for 25 s, followed by sonication in acetone for 60 s.
The Ni NPs electroplating in n-InP template was performed by pulsed electrochemical deposition involving the same two-electrode electrochemical cell, where the sample served as the working electrode and the Pt mesh acted as the counter electrode. A source meter, Keithley 2400 (Tektronix, Beaverton, OR, USA) was used for rectangular pulse generation with a voltage amplitude of −20 V. The electrodeposition process was carried out at 45 • C, employing an aqueous electrolyte containing Ni(SO 3 NH 2 ) 2 ·4H 2 O (400 g·L −1 ), NiCl 2 ·6H 2 O (12 g·L −1 ), H 3 BO 3 (40 g·L −1 ), and CH 3 (CH 2 ) 11 OSO 3 Na (0.5 g·L −1 ), with a pH value of 3.5. The morphology of the prepared samples was studied using a Hitachi SU8230 Scanning Electron Microscope (SEM) (Hitachi High-Tech Corp., Tokyo, Japan) equipped with an Energy Dispersive X-Ray Spectroscopy (EDX) detector-analyser (Oxford Instruments PLC, Oxford, UK).

Optimization of the Anodization Process
The electrochemical dissolution behavior of n-InP crystals with different electron concentrations in 3.5 M NaCl electrolyte was characterized by I-V curves, as shown in Figure 1a. The current sharply increased up to an applied potential of 6 V for the sample possessing higher conductivity, with a concentration of 2 × 10 18 cm −3 (red curve in Figure 1a). During the further increase of the applied voltage, a current plateau was noticed. The use of crystals with the free electron concentration of 2 × 10 17 cm −3 determined a different evolution of I-V curves (black curve in Figure 1a). Therefore, a small increase of the current up to 8 V was evidenced, with the etch pit formation at about 5 V. The voltage domain between 12-22 V presents interest for a controllable pore growth. At higher applied voltages (more than 22 V), fluctuations in pore diameters or even electropolishing are expected. The recorded behavior of the current against time shows an exponential decrease, as can be observed in Figure 1b, and can be related to the depletion of the electrolyte in pores and difficulties of electrolyte exchange with the deepness of the produced porous layer.
The morphology of the prepared samples was studied using a Hitachi SU8230 Scanning Electron Microscope (SEM) (Hitachi High-Tech Corp., Tokyo, Japan) equipped with an Energy Dispersive X-Ray Spectroscopy (EDX) detector-analyser (Oxford Instruments PLC, Oxford, UK).

Optimization of the Anodization Process
The electrochemical dissolution behavior of n-InP crystals with different electron concentrations in 3.5 M NaCl electrolyte was characterized by I−V curves, as shown in Figure 1a. The current sharply increased up to an applied potential of 6 V for the sample possessing higher conductivity, with a concentration of 2 × 10 18 cm −3 (red curve in Figure  1a). During the further increase of the applied voltage, a current plateau was noticed. The use of crystals with the free electron concentration of 2 × 10 17 cm −3 determined a different evolution of I−V curves (black curve in Figure 1a). Therefore, a small increase of the current up to 8 V was evidenced, with the etch pit formation at about 5 V. The voltage domain between 12-22 V presents interest for a controllable pore growth. At higher applied voltages (more than 22 V), fluctuations in pore diameters or even electropolishing are expected. The recorded behavior of the current against time shows an exponential decrease, as can be observed in Figure 1b, and can be related to the depletion of the electrolyte in pores and difficulties of electrolyte exchange with the deepness of the produced porous layer. measured with a scan rate of 10 mV·s −1 before the anodization of InP samples with a free carrier concentration of 2 × 10 18 cm −3 (red) and 2 × 10 17 cm −3 (black). (b) Chronoamperograms recorded during InP templates fabrication with a carrier concentration of 2 × 10 18 cm −3 (red) and 2 × 10 17 cm −3 (black) at 6.5 V and 20 V, respectively.
As mentioned above, the higher the carrier concentration in the semiconductor material, the lower the anodization voltage necessary for achieving a uniform and controlled porosification. As shown in Figure 2, an anodization voltage of 6.5 V is enough to produce templates presenting a uniform distribution of pores in InP substrates with a carrier concentration of 2 × 10 18 cm −3 . As mentioned above, the higher the carrier concentration in the semiconductor material, the lower the anodization voltage necessary for achieving a uniform and controlled porosification. As shown in Figure 2, an anodization voltage of 6.5 V is enough to produce templates presenting a uniform distribution of pores in InP substrates with a carrier concentration of 2 × 10 18 cm −3 .
Pore diameters between 80-120 nm have been evidenced when applying this anodization voltage associated with the self-ordered arrangement of pores. Note that self-ordering of pores occurs due to the growth of current line-oriented pores and without any photolithographic means [6]. The thickness of the porous layer is around 100 µm, and the pores maintain their diameter along the pore propagation direction, which is perpendicular to the sample surface. Pore diameters between 80-120 nm have been evidenced when applying this anodization voltage associated with the self-ordered arrangement of pores. Note that selfordering of pores occurs due to the growth of current line-oriented pores and without any photolithographic means [6]. The thickness of the porous layer is around 100 µm, and the pores maintain their diameter along the pore propagation direction, which is perpendicular to the sample surface.
The InP substrates possessing a carrier concentration of 2 × 10 17 cm −3 require anodization voltages higher than 10 V, as can be seen from the I−V curve in Figure 1a. Figure 3 illustrates the optimization of the anodization voltage in these substrates when they are electrochemically etched in 3.5 M NaCl electrolyte. The thickness of the produced porous layer exhibits a strong fluctuation at the applied voltage of 10 V (see Figure 3). The fluctuations are significantly reduced at 12 V (Figure 3b), and the thickness of the porous layers is constant at 15 V and 20 V (Figure 3c,d). The nonuniformity in the thickness of the porous layer at low applied voltages can be explained by the preponderant growth of the crystallographic oriented pores characterized by radial propagation in directions underneath the surface. The InP substrates possessing a carrier concentration of 2 × 10 17 cm −3 require anodization voltages higher than 10 V, as can be seen from the I-V curve in Figure 1a. Figure   The concentration of the electrolyte solution also influences the produced porosi as presented in Figure 4. One can see that a flower-like morphology of the pores is dev oped using the 1.75 M NaCl electrolyte, at an applied anodization voltage of 12 V (Figu 4a,b,c). This behavior is characteristic for anodization in low-concentration electrolyt resulting in nucleation of the pores at surface defects (dislocations) with subsequent rad growth facing away underneath the surface. On the other hand, the pores propagate p pendicularly to the sample surface when anodization is performed at the same voltage the 3.5 M NaCl electrolyte, as illustrated by the SEM image taken at the bottom of t pores ( Figure 4d). In such a case, the dissolution mechanism occurs as follows: the f mation of pore nuclei at surface with a further abundant branching of crystallographica oriented (crysto) pores forming the nucleation layer (see Section 3.2). Around each po a surface-depleted layer (W) is formed with a thickness that can be estimated from t relation (1): where φ0 is the surface potential, ε0εS is the static dielectric constant of the material, a ND + is the concentration of ionized donors. At a certain stage of pore growth in the depths, the branching of crysto pores slo down and the transition from crysto pores to current line-oriented (curro) pores occu The current line-oriented pores start to grow aligned perpendicular to the substrate su face. It should be mentioned that current line-oriented pores cannot intersect with ea other, therefore providing conditions for the self-ordering of pores [6,30]. Taking into a count that around each pore a depletion layer is formed, and the fact that curro por The concentration of the electrolyte solution also influences the produced porosity, as presented in Figure 4. One can see that a flower-like morphology of the pores is developed using the 1.75 M NaCl electrolyte, at an applied anodization voltage of 12 V (Figure 4a-c). This behavior is characteristic for anodization in low-concentration electrolytes, resulting in nucleation of the pores at surface defects (dislocations) with subsequent radial growth facing away underneath the surface. On the other hand, the pores propagate perpendicularly to the sample surface when anodization is performed at the same voltage in the 3.5 M NaCl electrolyte, as illustrated by the SEM image taken at the bottom of the pores (Figure 4d). In such a case, the dissolution mechanism occurs as follows: the formation of pore nuclei at surface with a further abundant branching of crystallographically oriented (crysto) pores forming the nucleation layer (see Section 3.2). Around each pore, a surface-depleted layer (W) is formed with a thickness that can be estimated from the relation (1): where ϕ 0 is the surface potential, ε 0 ε S is the static dielectric constant of the material, and N D + is the concentration of ionized donors. The uniformity of the pore diameters along the propagation directio the anodization voltage increased from 12 V to 20 V, as shown in Figure 5 conclude that the use of the 3.5 M NaCl electrolyte is suitable for the ano substrates possessing a carrier concentration of 2 × 10 17 cm −3 , while the opti tion voltage is 20 V. However, removal of the top nucleation layer is necessa the entrance in pores, which is necessary for templated metal deposition. At a certain stage of pore growth in the depths, the branching of crysto pores slows down and the transition from crysto pores to current line-oriented (curro) pores occurs. The current line-oriented pores start to grow aligned perpendicular to the substrate surface. It should be mentioned that current line-oriented pores cannot intersect with each other, therefore providing conditions for the self-ordering of pores [6,30]. Taking into account that around each pore a depletion layer is formed, and the fact that curro pores cannot intersect, they start to push each other to maintain the two depletion layers 2 W between them [6].
The uniformity of the pore diameters along the propagation direction improved as the anodization voltage increased from 12 V to 20 V, as shown in Figure 5. Thus, one can conclude that the use of the 3.5 M NaCl electrolyte is suitable for the anodization of InP substrates possessing a carrier concentration of 2 × 10 17 cm −3 , while the optimum anodization voltage is 20 V. However, removal of the top nucleation layer is necessary for opening the entrance in pores, which is necessary for templated metal deposition. the anodization voltage increased from 12 V to 20 V, as shown in Fi conclude that the use of the 3.5 M NaCl electrolyte is suitable for th substrates possessing a carrier concentration of 2 × 10 17 cm −3 , while th tion voltage is 20 V. However, removal of the top nucleation layer is n the entrance in pores, which is necessary for templated metal depos

Optimization of the Pore Opening Process
In both samples characterized by different free carrier concentrations, the top nucleation layer formed with a thickness of about 2-5 µm. It should be noted that in the sample with a lower electrical conductivity, the ramification of crystallographic oriented pores occurred slower, and the self-ordering process, characterized by strictly parallel pores with an equal wall thickness, occurred at a deepness of about 15 µm. Consequently, a perfect cross-sectioning of the porous layer is difficult to obtain (see Figure 6a). Generally, for samples with 2 × 10 18 cm −3 carrier concentration, it is enough to use chemical dissolution for 20-25 s. At the same time, the nucleation layer in the sample with a carrier concentration of 2 × 10 17 cm −3 requires a longer duration of wet isotropic chemical etching allowing the dissolution of thick pore walls, in comparison with the high carrier concentration sample. It is obvious that during a longer duration of chemical etching, the electrolyte will act more deeply. Consequently, to avoid the complete removal of the porous layer, a combination of chemical etching for 25 s in an acidic solution followed by sonication in acetone was proposed for removal of the top nucleation layer and opening the entrance of the pores. In such a way, the pore walls are not dissolved but are destroyed by ultrasound. One can see from Figure 6b that the top porous layer is destroyed when the sample is chemically treated in an acid solution for 25 s. The final removal of the nucleation layer and the opening of pores is obtained by sonication during 1 min in acetone (Figure 6c). cation in acetone was proposed for removal of the top nucleation layer and opening entrance of the pores. In such a way, the pore walls are not dissolved but are destroy by ultrasound. One can see from Figure 6b that the top porous layer is destroyed wh the sample is chemically treated in an acid solution for 25 s. The final removal of the cleation layer and the opening of pores is obtained by sonication during 1 min in aceto (Figure 6c).    Figure 7 compares two porous templates with opened pores prepared on InP possessing different carrier concentrations. One can see that the dimensions of the pores increased about 5 times from around 100 nm to around 500 nm as the carrier concentration in the InP substrate decreased from 2 × 10 18 cm −3 to 2 × 10 17 cm −3 . The anodization process is sensitive to the concentration of the free carriers in the used semiconductor substrate. As can be seen from Equation (1), the thickness of the depletion layer decreases as the concentration of electrons increases. As shown in a previous publication [31], the increase of the electron concentration in the ZnSe crystals from 7 × 10 16 cm −3 to (1-2) × 10 18 cm −3 allowed us to reduce the pore diameter from 400-500 nm to 40 nm.
As can be seen from Equation (1), the thickness of the depletion layer decreases as the concentration of electrons increases. As shown in a previous publication [31], the increase of the electron concentration in the ZnSe crystals from 7 × 10 16 cm −3 to (1-2) × 10 18 cm − allowed us to reduce the pore diameter from 400-500 nm to 40 nm.

Ni Deposition inside the Porous InP Templates
Ni NPs and nanowires (NWs) deposition in the InP template was previously re ported [32,33] using an Al2O3 thin layer to cover the walls of the template pores. In thi work, Ni NPs were successfully electrochemically deposited inside the porous InP tem plates without any preliminary functionalization of the pore walls. To the best of ou knowledge, the formation of a direct interface between the walls of porous InP and N NPs is here reported for the first time. The InP template with pore diameters of 500 nm was selected since the penetration of solution deep inside the pores is enhanced. At the same time, the larger wall thickness of the InP template provides better mechanical stabil ity in the final structure.
The electrochemical deposition of Ni NPs was performed by the application of pulse with an amplitude of −20 V. It was found that the optimum pulse width (Ton) was around 8 ms, while the interval between pulses (Toff) was in the range of seconds, in this way providing enough time for refilling the template pores with the electrolyte solution, and also avoiding massive Ni deposition on the sample surface that could block further meta deposition into pores.
Adjusting the pulse width is essential for a uniform deposition of metals in semicon ductor porous templates. An important task consists of avoiding the electrolyte depletion inside pores and refreshing the metal species during the interval between pulses. In ou recent paper [34] we found that for uniform deposition along the pore depth (at a pore diameter less than 100 nm) the pulse width should be set at a value of (10-300) µs fo

Ni Deposition inside the Porous InP Templates
Ni NPs and nanowires (NWs) deposition in the InP template was previously reported [32,33] using an Al 2 O 3 thin layer to cover the walls of the template pores. In this work, Ni NPs were successfully electrochemically deposited inside the porous InP templates without any preliminary functionalization of the pore walls. To the best of our knowledge, the formation of a direct interface between the walls of porous InP and Ni NPs is here reported for the first time. The InP template with pore diameters of 500 nm was selected since the penetration of solution deep inside the pores is enhanced. At the same time, the larger wall thickness of the InP template provides better mechanical stability in the final structure.
The electrochemical deposition of Ni NPs was performed by the application of pulses with an amplitude of −20 V. It was found that the optimum pulse width (T on ) was around 8 ms, while the interval between pulses (T off ) was in the range of seconds, in this way providing enough time for refilling the template pores with the electrolyte solution, and also avoiding massive Ni deposition on the sample surface that could block further metal deposition into pores.
Adjusting the pulse width is essential for a uniform deposition of metals in semiconductor porous templates. An important task consists of avoiding the electrolyte depletion inside pores and refreshing the metal species during the interval between pulses. In our recent paper [34] we found that for uniform deposition along the pore depth (at a pore diameter less than 100 nm) the pulse width should be set at a value of (10-300) µs for enabling the deposition of only 70-80% of the metal ions inside the pores during the applied pulse. This was enough to avoid electrolyte depletion inside pores and the refreshing of the metal species. In the case of InP templates with a pore diameter of 500 nm, the pulse width should be increased to several milliseconds for these purposes. Figure 8 shows the cross-section images of samples with Ni deposition at a constant T off = 1 s and T on values of 4, 8, 10 and 15 ms. As presented in Figure 8a, applying a pulse width of 4 ms led to the formation of Ni NPs inside the pores. As the T on increased to 8 ms, the diameter of nanoparticles became larger and a slight increase in the density of NPs deposited on the pore walls was also observed. Higher values of T on of 10 and 15 ms, respectively (see the insets from Figure 8c,d), determined the electrodeposition of Ni on the surface of the porous layer, thus blocking any further deposition in the depth of the pores. Therefore, an optimum pulse width of 8 ms to provide Ni NPs growth inside the pores was considered. enabling the deposition of only 70-80% of the metal ions inside the pores during the applied pulse. This was enough to avoid electrolyte depletion inside pores and the refreshing of the metal species. In the case of InP templates with a pore diameter of 500 nm, the pulse width should be increased to several milliseconds for these purposes. Figure 8 shows the cross-section images of samples with Ni deposition at a constant Toff = 1 s and Ton values of 4, 8, 10 and 15 ms. As presented in Figure 8a, applying a pulse width of 4 ms led to the formation of Ni NPs inside the pores. As the Ton increased to 8 ms, the diameter of nanoparticles became larger and a slight increase in the density of NPs deposited on the pore walls was also observed. Higher values of Ton of 10 and 15 ms, respectively (see the insets from Figure 8c,d), determined the electrodeposition of Ni on the surface of the porous layer, thus blocking any further deposition in the depth of the pores. Therefore, an optimum pulse width of 8 ms to provide Ni NPs growth inside the pores was considered.  It was observed that the Ni electroplating in the n-InP template occurs in agreement with the early reported "hopping electrodeposition" of Au [35]. As can be seen from Figure 8, the deposition of individual Ni nanodots is produced during the electroplating process. In accordance with the proposed model in [35], the nanodots grow up to a critical threshold diameter governed by the Schottky contact height. The deposition is continued  The experimental demonstration of the Schottky contact formation is evidenced in Figure 9 by the bright color of the deposited Ni nanodots as compared to the less conductive porous skeleton. As the electroplating process continues, the dots tend to overlap, forming metal nanotube walls. The electrodeposition in this mode results in a gradient-like deposition with maximum metal deposition close to the sample surface and decreasing deposition in the depth of pores, as illustrated in Figure 10. It was observed that the Ni electroplating in the n-InP template occurs in agreement with the early reported "hopping electrodeposition" of Au [35]. As can be seen from Figure 8, the deposition of individual Ni nanodots is produced during the electroplating process. In accordance with the proposed model in [35], the nanodots grow up to a critical threshold diameter governed by the Schottky contact height. The deposition is continued by the generation of new nanodots until all internal surfaces are covered by a monolayer of dots.
The type of the contact, i.e., ohmic or Schottky, can be estimated from the analysis of the difference between the work function of the metal (ϕ m ) and the value of the electron affinity of InP (χ s ). One can expect the formation of a Schottky contact on an n-type material when ϕ m − χ s > 0. Among metals (such as Mg, Zn, Al, Cr, Ni, Pt), Pt exhibits the highest value of ϕ m − χ s which equals 1.3 eV, followed by Ni (0.81 eV) as seen in Table 1. Table 1. The values of the work function for several metals [36] and calculated difference between the work function of the metal and the value of the electron affinity of InP.

Metal
Work Function ϕ m (eV) The experimental demonstration of the Schottky contact formation is evidenced in Figure 9 by the bright color of the deposited Ni nanodots as compared to the less conductive porous skeleton. As the electroplating process continues, the dots tend to overlap, forming metal nanotube walls.
The electrodeposition in this mode results in a gradient-like deposition with maximum metal deposition close to the sample surface and decreasing deposition in the depth of pores, as illustrated in Figure 10.
atomic percentage of Ni decreases from around 17 at.% at the entrance of pores to 2 at.% at the bottom of pores. Such a preferential deposition of Ni NPs in pores close to the surface of the InP template is beneficial for potential applications when interaction with the surrounding environment is desirable.   The gradient-like deposition is confirmed by the EDX analysis performed in depth of the porous layer, as shown in Figure 11 and Table 2. As one can see from this analysis, the atomic percentage of Ni decreases from around 17 at.% at the entrance of pores to 2 at.% at the bottom of pores. Such a preferential deposition of Ni NPs in pores close to the surface of the InP template is beneficial for potential applications when interaction with the surrounding environment is desirable.
The histograms of Ni NPs diameter distribution, as determined from SEM analysis for deposition processes with T off = 1.5 s and T off = 2 s, are shown in Figure 12. This analysis demonstrates that the average Ni NPs diameter is 85 nm in the case of T off = 1.5 s and 55 nm for T off = 2 s. The longer interval between pulses provides better replenishment of the electrolytes inside of pores, thus assuring the nucleation of new nanodots instead of their growing in diameter. It should be noted that the deposition of the size-saturated 20 nm monolayer of gold nanodots electroplated on the InP and GaP porous semiconductor substrates was reported using 10 µs pulse width [35]. These results are in the same range as the data reported in [32], and with diameters of ferromagnetic NPs forming nanotubular structures on GaAs nanowires [23][24][25]. The histograms of Ni NPs diameter distribution, as determined from SEM analysis for deposition processes with Toff = 1.5 s and Toff = 2 s, are shown in Figure 12. This analysis demonstrates that the average Ni NPs diameter is 85 nm in the case of Toff = 1.5 s and 55 nm for Toff = 2 s. The longer interval between pulses provides better replenishment of the electrolytes inside of pores, thus assuring the nucleation of new nanodots instead of their growing in diameter. It should be noted that the deposition of the size-saturated 20 nm monolayer of gold nanodots electroplated on the InP and GaP porous semiconductor substrates was reported using 10 µs pulse width [35]. These results are in the same range as the data reported in [32], and with diameters of ferromagnetic NPs forming nanotubular structures on GaAs nanowires [23][24][25]. Ni-based nanotubes are promising for magnetic applications, including applications in high-density data storage [20,21]. The technology of ferromagnetic nanotube fabrication is complementary to the previously proposed technology based on GaAs nanowires [24,25]. At the same time, it has the advantages of widening the range of nanotube Figure 11. Cross-sectional SEM micrograph denoting the zones for EDX analysis, proving the gradient-like deposition of Ni NPs.  Figure 11. Cross-sectional SEM micrograph denoting the zones for EDX analysis, proving the gradient-like deposition of Ni NPs.
The histograms of Ni NPs diameter distribution, as determined from SEM analysis for deposition processes with Toff = 1.5 s and Toff = 2 s, are shown in Figure 12. This analysis demonstrates that the average Ni NPs diameter is 85 nm in the case of Toff = 1.5 s and 55 nm for Toff = 2 s. The longer interval between pulses provides better replenishment of the electrolytes inside of pores, thus assuring the nucleation of new nanodots instead of their growing in diameter. It should be noted that the deposition of the size-saturated 20 nm monolayer of gold nanodots electroplated on the InP and GaP porous semiconductor substrates was reported using 10 µs pulse width [35]. These results are in the same range as the data reported in [32], and with diameters of ferromagnetic NPs forming nanotubular structures on GaAs nanowires [23][24][25]. Ni-based nanotubes are promising for magnetic applications, including applications in high-density data storage [20,21]. The technology of ferromagnetic nanotube fabrication is complementary to the previously proposed technology based on GaAs nanowires [24,25]. At the same time, it has the advantages of widening the range of nanotube Ni-based nanotubes are promising for magnetic applications, including applications in high-density data storage [20,21]. The technology of ferromagnetic nanotube fabrication is complementary to the previously proposed technology based on GaAs nanowires [24,25]. At the same time, it has the advantages of widening the range of nanotube diameters and diminishing the wall thickness to lower values. Note that nanopores with diameters smaller than 100 nm can be produced in InP templates, while the wall thickness can be reduced thanks to the possibility of depositing a monolayer of metallic nanodots in a controlled fashion. The fabrication of ferromagnetic nanotubes with smaller geometrical parameters is important from the point of view of producing structures with larger magnetic anisotropy. It was previously shown that nanotubes with diameters smaller than 100 nm and thin walls are needed to reach a high magnetic anisotropy [28].

Conclusions
As a result of the performed investigations, the preparation of uniform porous InP templates presenting the depth of pores up to 100 µm and pore diameter in the range 100-500 nm, depending on the electrical conductivity of the InP substrate, involving an anodization process in an environmentally friendly electrolyte, has been proposed. The optimum composition of the electrolyte was found to be 3.5 M NaCl in water, while the optimum anodization voltage was 6.5 V for substrates possessing a carrier concentration of 2 × 10 18 cm −3 and 20 V for those with a carrier concentration of 2 × 10 17 cm −3 . Ni NPs were successfully deposited into prefabricated InP templates without any preliminary surface functionalization or deposition of intermediary layers before pulsed electrodeposition of Ni NPs. The optimum pulse width was found to be around 8 ms, while the interval between pulses was 1.5-2 s. Ni NPs with diameters of 85 nm-55 nm were formed. A gradient-like Ni deposition was demonstrated with predominant deposition in the upper region of the porous template, which is expected to be of interest for interactions with the surrounding environment, i.e., in chemical and biological sensors. The proposed technology could be also prospective for the exploration of magnetic properties and for photocatalytic/photoelectrocatalytic applications.

Data Availability Statement:
The data presented in this study are available on request from the corresponding authors.