# Nanostructured Silicon as Host Material

Authored by: P. Granitzer , K. Rumpf

# 21 Century Nanoscience – A Handbook

Print publication date:  April  2020
Online publication date:  April  2020

Print ISBN: 9780815357087
eBook ISBN: 9780429351594

10.1201/9780429351594-3

#### Abstract

Nanostructuring of materials is a basic topic in today’s research, such as material science, physics, chemistry, or biology. Low-dimensional systems are not only of interest due to miniaturization and the utilization in integrated devices but also because of their arising novel properties. Nanostructuring of semiconductors and especially silicon can lead to on-chip applications that are appropriate for today’s microtechnology not only because of the small dimensions but also because of the changed physical properties of materials compared with their bulk behavior. Three-dimensional nanostructured systems also enable their utilization as templates for the incorporation of various materials to achieve new nanocomposites with specific properties. This chapter reviews the use of nanostructured semiconductors as host substrate, especially for the deposition of various materials. The incorporated or attached materials can be metals, magnetic materials, polymers, and also molecules for biological sensing applications. The incorporation of magnetic materials into nanostructured silicon results in a semiconducting/magnetic system offering the properties of both materials which are determined by their nanoscale sizes. Nanostructuring of a semiconductor can be achieved by patterning methods such as lithography or ion beam irradiation as well as by self-organized processes, especially chemical etching. In the following, wet etching, namely anodization, will be mainly discussed.

#### 3.1  Introduction

In today’s semiconductor technology, silicon is the most dominant material that is used for microelectronic devices. The desire to develop integrable devices with decreasing dimensions results in downscaling by nanostructuring of the base material. Crystalline silicon in its bulk appearance is commonly not taken into consideration as optical, magnetic, or biomedical material, but nanostructuring, leading to a dramatic change of bulk properties (e.g. light emitting, biodegradable), is a method to enhance the functionality of silicon in nanotechnology. Also other semiconductor materials such as Ge, GaAs, or InP are under intense investigation due to the production of low-dimensional structures.

In general, due to the novel obtained physical properties, nanostructured and low-dimensional materials play a decisive role in today’s basic research as well as in applications like integrated circuits at nanometric sizes [1], optoelectronic [2] and magneto-optical devices [3], perpendicular media for high-density data storage [4], and nanostructures as functionalized sensors in nanobiology [5]. The fabrication of nanopatterned materials is widely spread in physics, chemistry, and also in biology. A popular technique to produce nanometric structures is lithography used for top-down strategies [6] or bottom-up growth mechanisms [7]. Self-assembled and self-organized structures are of great interest due to the elementary fabrication processes. Quite common are nanoparticles grown on a substrate by self-assembly [8]. But also three-dimensional arrays of nanowires [9] or nanotubes have been produced without prestructuring, whereas porous alumina templates, growing in a hexagonal arrangement, are the widely used matrices [10]. Magnetic properties of metal-filled membranes (e.g. porous alumina, polycarbonate) are under extensive investigation [11,12]. Magnetostatic interaction mechanisms between deposited metal structures arranged in an array [13], exchange coupling between nanostructures composed of two different metals [14], spin transport phenomena such as magnetoresistance in spin valves [15], and also the investigation of magneto-optical effects [16] are of great interest. Nanostructured materials offer drastic but different physical properties compared with their bulk materials, especially regarding the nanostructuring of silicon by electrochemical etching, or the growth of Si-nanowires leads to a complete new behavior of the resulting materials. Electrochemically treated silicon exhibiting a porous structure (porous silicon) shows properties that cannot be achieved by bulk silicon itself (e.g. electrical isolating properties, light emission in the visible due to quantum confinement (QC) effects or biocompatibility, and biodegradability).

Not only it’s big surface area but also the controllable morphology renders the system applicable as template for the incorporation of various materials. In the early 1990s, porous silicon has been filled with metals to improve the electrical contact for electroluminescence investigations [17]. Cu has been filled into macroporous silicon to examine the deposition mechanism in detail [18] and also for applications in microelectronics such as heat sinks in high power density electronic cooling [19]. The pore filling with magnetic materials leads to a nanocomposite with semiconducting as well as magnetic properties. Thus the system is a good candidate for magnetic applications in integrated devices.

#### 3.1.1  Nanostructured Semiconductors

Nanostructuring of silicon, which will be mainly discussed in this chapter, can be achieved by wet or dry etching methods, such as anodization [20], stain etching [21], metal-assisted etching (MAE) [22], galvanic etching [23], or reactive ion etching, which is quite common in microelectronic processes. These methods are often used as self-organizing processes, but they can also be used together with a lithographic mask to fabricate well-ordered pore arrangements [24]. The self-organization is beneficial because of the low cost and less time consumption. To produce ordered macro-pores, e.g. photonic crystal applications, prestructuring is essential, but quasi-regular pore arrangements can also be achieved by self-organization [25].

Nanostructuring of germanium can also be carried out by etching in alkaline or acidic solutions, whereas the latter one leads to better anisotropic pit formation [26]. Although the anisotropic formation process is weaker than in silicon, it can be used for micromachining of samples. A further structuring method by self-organization is ion implantation to produce structures at a nanoscale [27].

The anodization of InP is also intensely investigated, such as n-InP in aqueous KOH [28] or in HCl, HBr, and hydrofluoric acid (HF) solutions, with and without illumination of the sample [29]. The anodization of InP can lead to high aspect ratio pores, which have been investigated in detail with respect to the growth mechanism [30].

#### 3.1.2  Methods of Nanostructuring by Patterning and Self-Organization

Lithography (e-beam, optical), which is a standard method to achieve desired well-ordered patterns, is often used to fabricate ordered arrangements of nanostructures (particles, rods, wires) on a substrate, but it is also used to produce a mask. In the case of macroporous silicon, such a mask is employed for alkaline etching, resulting in patterns of pits that are subsequently etched, resulting in the formation of regular arranged macropores, e.g. photonic crystal applications [31].

Generally the formation of porous silicon is performed by self-organization, which is cheaper and, especially, less time consuming. Porous silicon can be produced with different techniques and in a great variety of morphologies. The most prominent wet-etching methods are stain etching, MAE, galvanic etching, and anodization.

In the case of stain etching, the silicon dissolution is initiated by electronic hole injection from an oxidant. The advantage of this porosification method, which is carried out in a solution containing HF and nitric acid, is the electroless process, but the disadvantage is that generally the tuning of the pore diameter is diffcult and also only limited porous layer thicknesses can be achieved [32]. The process has been refined using HF and FeCl · 6H2O or V2O5 solutions to increase the porous layer thickness [33].

MAE, a method where the metal acts as a catalyst for the hole injection from the oxidant, can be used to produce pores or silicon nanowires, depending on the structure of the metal mask. A typically employed electrolyte is an HF/H2O2 solution. Either discrete metal nanoparticles or a metal layer with openings is deposited on a silicon wafer (Figure 3.1). Usually MAE is performed by depositing Au or Ag on the wafer that inhibits pore formation [34].

A further metal containing method is galvanic etching, which is usually performed using an acidic fluoride electrolyte. In this case a metal layer is deposited on the backside of the wafer, which provides the holes for the pore formation in the silicon substrate [36]. A sketch of a galvanic etching setup is shown in Figure 3.2. Metals such as Au and Pt act as hole supply for the silicon dissolution process [36].

A method that allows to produce porous silicon with a great variety of morphologies and to tune pore diameters and pore distances is anodization. In this case the pore arrangements are mainly determined by the doping of the silicon wafer, the electrolyte composition, and the applied current density. Usually an aqueous HF solution, often with the addition of an oxidizing agent, is used. By varying the current density the pore diameter can be adjusted quite accurately. In Figure 3.3 examples of five different morphologies obtained by applying different current densities are shown. The electrolyte composition is the same for all samples and consists of HF:H2O:ethanol in the ratio 1:1:2. The anodization time is 8 min, and the current densities (j) are varied between 25 and 125 mA/cm2.

Figure 3.1   (a) Sketch of the process steps of MAE resulting in a porous silicon layer. The deposited metal (Ag, Au) particles initiate the pore formation process [after 35]. (b) Sketch of MAE resulting in silicon wires.

Figure 3.2   Illustration of a galvanic etching process with a Pt layer on the backside of the silicon wafer.

Figure 3.3   Top-view scanning electron micrographs (SEM) of porous silicon prepared with different current densities. (a) j = 125 mA/cm2 leading to an average pore-diameter of 95 nm and an average pore-distance of 45 nm; (b) j = 100 mA/cm2, average pore-diameter 60 nm and average pore-distance 50 nm; (c) j = 75 mA/cm2, average pore-diameter 45 nm and average pore-distance 55 nm, (d) j = 50 mA/cm2, average pore-diameter 25 nm and average pore-distance 60 nm; (e) j = 25 mA/cm2 leading to a random pore distribution. For all samples, a highly n-doped silicon wafer as substrate material has been used.

Detailed descriptions of the formation process of the various morphologies can be found in numerous publications [31,37–39].

#### 3.1.3  Deposition Techniques

Electrodeposition is a common technique to produce metal films and metal structures on substrates. The dynamic equilibrium between the metal and its ions in the solution is reached after an exchange of metal ions between the two phases:

3.1 $M z + + ze ↔ M$

M . . . metal

e . . . electron

z . . . integer

A specific common deposition technique is pulsed electrodeposition, which modifies the diffusion layer [40]. This method is often used to improve the distribution of the deposit, the leveling, and the brightness of the deposit [40]. Furthermore it is also employed in the case of filling high aspect ratio structures to suppress the exhaustion of the electrolyte. The mostly used pulses are rectangular, periodic-reverse, or sinusoidal pulses. The main growth mechanisms in using electrodeposition are (i) nucleation– coalescence growth, which happens first by nuclei formation, followed by the coalescence of three-dimensional crystallites, and finally the formation of a continuous deposit and (ii) layer growth that is formed by crystal enlargement due to spreading of discrete layers [40].

Electroless deposition is a plating method without the application of electricity. In this case the electrons are supplied by the reducing agent of the solution. The reaction is described as follows by reduction and oxidation [40]:

3.2 $M sol z + + Red sol → catalytic surface M lat + Ox sol$

Red . . . reducing agent

Ox . . . oxidation product

sol . . . solution

lat . . . lattice

Also the fabrication of low-dimensional structures can be achieved by deposition techniques. Besides the preparation of thin films, in-plane wires, and particles, three dimensional arrays of nanostructures can be produced by using a template material such as porous alumina (AAO) [41,42] or porous silicon [43,44]. Besides Ni, Co, and Fe, segmented structures consisting of a combination of Ni x Co y or Ni x Co y /Cu are used, e.g. as three-dimensional magnetic data storage system [45]. The system can be tuned by the variation of the Ni x Co y composition from hard to soft magnetic. The deposited Cu segments give rise to well-defined NiCo rods with pinning-free domain wall propagation [45]. Three-dimensional arranged segmented rods within a template can be used to control magnetization reversal and domain wall motion.

#### 3.2.1  Nanostructured Silicon Fabrication

The morphology of porous silicon can be varied in a broad range, whereas the most important structure affecting parameters are the type and doping level of the silicon wafer, the crystalline orientation, the electrolyte composition, the HF concentration, the bath temperature, and the applied current density. By varying these parameters, the pore arrangement can be tuned from pores of a few nanometers to pores of 10 μm. With respect to the International Union of Pure and Applied Chemistry (IUPAC) notation the morphology is classified into microporous with structure sizes between 1 and 5 nm, mesoporous (pore diameters between 5 and 50 nm), and macroporous silicon (pore diameters greater than 50 nm).

The most common electrolytic anodization cells are either cells consisting of one tank for the electrolyte and a backside contact, whereas often a metal is evaporated on the back of the substrate or double tank cells (Figure 3.4). In the latter case the electrolyte acts as a backside contact, and thus no metal layer is necessary, and furthermore such cells are used for backside illumination, which is necessary, e.g. in the case of n-doped silicon, to generate the required holes for the silicon dissolution process. Furthermore these cells can be used to produce double-sided porous silicon [46]. In the case of using thinned silicon wafers, double-sided samples are produced to fabricate magnetic metal filled porous layers that magnetically interact [46].

Generally for the dissolution process of silicon, electronic holes are required, and these holes are depleted in the remaining silicon skeleton. If the applied current density j is greater than the critical current density j PS, holes are accumulated at the silicon surface and a concomitant depletion of HF takes place, which results in electropolishing [47]. In the case of j < j PS, holes are depleted at the silicon electrode and HF accumulates at its surface. This causes an initial pore growth at the existing depressions and pits because the electric field lines within the space charge region are focused so that the local current density is enhanced. If the current density is equal to the critical current density (j = j PS), charge transfer and ionic transport fulfill a steady-state condition, meaning that the current density at the pore tips has to be equal to the critical current density to achieve pore growth [48].

Figure 3.4   Illustration of a double cell arrangement for the formation of porous silicon.

Explaining the formation of various morphologies, different models are assumed [37,31]. The formation of microporous silicon (pore diameter in the range of 2 nm) is due to QC effects [49]. Mesopores are formed by highly anodizing n- or p-doped silicon with applied high current densities due to avalanche breakthrough [50]. The formation of macropores can be explained by charge transfer across the silicon/electrolyte Schottky barrier if the nonplanar pore interface is considered [51].

In keeping all parameters constant and only modifying the applied current density, the pore diameter and the concomitant pore distance can be modified in a certain regime. Considering pore diameters between 20 and 100 nm the pore distance is concomitantly formed between 70 and 40 nm. Pores in this size regime are oriented and separated from each other. To assure a separation of pores the thickness of the remaining silicon between the pores should not exceed twice the thickness of the space charge region [31].With decreasing pore diameter the pore distance increases, and in addition, the dendritic growth of the pores is enhanced. A sophisticated method to decrease the strong dendritic pore growth of mesopores is magnetic field assisted etching, which has been developed at the Tokyo University of Agriculture and Technology [52]. Due to the applied magnetic field being perpendicular to the sample surface, the motion of the holes which is responsible for the silicon dissolution is controlled and restricted to the pore tip region, which results in less dendritic pores. Furthermore the electrolyte is kept at T = 0 ° C, which also reduces the mobility of the holes. The filling of such pore arrangements with a ferromagnetic metal results in specific magnetic properties of the nanocomposite material, which are mainly due to the shape, size, and spatial arrangement of metal deposits. The magnetic behavior is also strongly influenced by the morphology of the template material, especially the distance between the pores, which determines the magnetic coupling between metal structures of adjacent pores. Furthermore dendritic pore growth enhances the magnetic coupling between the pores due to a reduced average pore distance [53]. Therefore magnetic field assisted etched porous silicon filled with a magnetic metal (Ni) offers higher coercivities and higher magnetic anisotropies compared with conventionally etched samples because of less magnetic interactions between deposits of neighboring pores [53]. Figure 3.5 shows a comparison of porous silicon prepared by conventional anodization and magnetic field assisted etching.

Figure 3.5   SEM of a cross-sectional region of (a) conventional etched and (b) magnetic field assisted anodized porous silicon by applying a magnetic field of 8 T parallel to the pores during the anodization process [53].

#### Deposition of Ferromagnetic Metals

Already in the 1990s, microporous silicon has been filled with Fe, Al, and In to improve the electrical contact for electroluminescence experiments [54,55]. Furthermore Au, Cu, and Ni have been deposited to study the electroless as well as the cathodic electrodeposition processes. It has been shown that during the electroless deposition an oxide layer has been formed on the pore walls, whereas during electrode-position the filling of the pores occurred without oxidation of the pore walls [56]. Considering metal deposition within mesoporous silicon the filling has shown that mass transfer is an important factor that controls the metal deposition in the pores [57]. To facilitate the deposition into meso-pores, especially in the case of high aspect ratio pores, often pulsed electrodeposition under cathodic conditions is used to suppress the exhaustion of the electrolyte within the pores [58,59]. The mechanism responsible for the metal deposition is proposed to be particle nucleation and diffusion-controlled growth [59]. As electrolyte, an adequate metal solution is employed. In the case of Ni deposition, either an aqueous NiCl2 solution or the so-called Watts electrolyte consisting of NiCl2 and NiSO4 is used. In both cases H3BO3 can be used as a buffer [60]. For depositing Co within porous silicon a CoSO4 electrolyte is employed [61], and to achieve NiCo alloy, a combination of CoSO4 and Watts electrolyte in the ratio 1:1 is used [62]. Fe deposition within porous silicon can be carried out by employing FeSO4, ammonium lauryl sulfate, saccharine, and acetic acid solution [63]. A current density in the milliampere regime is applied for a duration of few seconds to few minutes under galvanostatic conditions. The iron formation is investigated with respect to the applied current density, deposition duration, and the pH of the electrolyte [63].

In using ferromagnetic metals for the deposition process it is of interest to achieve nanocomposite systems with specific magnetic properties, which can be obtained by modifying mainly the pulse duration and the applied current density. Also the electrolyte composition and concentration as well as the temperature of the solution influence the deposition process. For a given electrolyte a decrease of the pulse duration from 40 to 5 s leads to an increase of the deposited Ni structures from sphere-like particles (~60 nm) to wires up to 2.5 μm [64], which is shown in Figure 3.6.

Besides elongated structures, fine particles of a few nanometer in size can also be deposited on the inner pore walls (Figure 3.7), forming a tube-like arrangement. By increasing the current density from 20 to 50 mA/cm2, such densely packed fine particles forming quasi-nanotubes around the pore walls can be achieved [65].

In the case of filling a microporous layer with a magnetic metal before the electrodeposition process, the sample is impregnated into the used electrolyte under sonication for 15 min to facilitate the penetration of the solution into the pores. The deposition time is also a critical parameter to avoid the formation of a metal layer on the surface. This fact is important because this kind of porous material offers luminescence in the visible range, which is investigated with respect to the amount of metal filling [66]. The growth mechanism of Pt and Ag within hydrophobic and hydrophilic microporous silicon has been investigated and explained by the effect of the overpotential and the displacement deposition rate [67].

Figure 3.6   Deposition of Ni within porous silicon with varying pulse duration from 5 to 40 s. With decreasing pulse duration the elongation of Ni structures increases from about 100 nm to about 2.5 μm. The applied current density for all samples was 25 mA/cm2 and the total deposition time was 20 min.

Figure 3.7   (a) Cross-sectional transmission electron microscopy (TEM) image showing Ni particles of about 3 nm deposited on the pore wall. (b) Corresponding top-view image showing the tube-like arrangement of the deposited fine Ni particles on the inner pore wall [65].

Figure 3.8   Thinned n+ silicon wafer of about 60 μm thickness offering a porous layer on each side. The left side is filled with Ni, and the right one with Co [68].

The filling of double-sided porous silicon samples is performed by using a double tank cell, whereas the metal deposition can be performed by using the same electrolyte in each tank or using different electrolytes. For the deposition of Ni on one side and Co on the other one, the deposition parameters (pulse duration, current density) have to be adjusted on each side with respect to the used metal salt solution [68]. Figure 3.8 shows an ultrathin wafer with double-sided porous silicon layers, one side filled with Ni and the other one with Co.

#### 3.3.1  Adjusting of Magnetic Behavior by the Shape and Size of Incorporated Metal Deposits

Mesoporous silicon with deposited magnetic nanostructures can be used to produce a semiconducting/magnetic system with desired magnetic properties. Magnetic properties of such nanocomposites strongly depend on the size and shape of the deposited magnetic metal structures. Generally the diameter of the deposits correlates with the pore diameter. With decreasing pore diameter the mechanism of the magnetization reversal [69] of deposited wires changes from the vortex mode to transverse mode. Ellipsoidal particles show either coherent or incoherent rotation in the case of deviations from the ideal ellipsoidal geometry [70]. A further aspect that influences the magnetic behavior is the distance between the pores, which also indicates the distance between metal structures of adjacent pores. With decreasing distance between the pores, the magnetic coupling between the structures increases. Considering magnetic field dependent measurements one can say that the hysteresis of isolated particles generally offers a higher coercivity than of wires [69]. The deposited metal particles within porous silicon can be tuned in their shape (spherical-like, elongated, wires) and also in their packing density, which determines the magnetic coupling between structures. In the case of deposited Ni particles the packing density strongly influences the magnetic properties. Figure 3.9 shows densely packed and less densely packed Ni particles within porous silicon of equal morphology, offering an average pore diameter of 60 nm and a mean distance between the pores of 50 nm.

In the case of wires, which offer an aspect ratio higher than 100, deposited within the template, the packing density within one pore is negligible, and the magnetic properties are determined by the elongated shape, especially the magnetic anisotropy. A typical sample containing Ni wires can be seen in Figure 3.10.

Figure 3.9   Cross-sectional SEM image of porous silicon with (a) densely packed and (b) less densely packed Ni particles. The porous morphology of both samples is equal.

Figure 3.10   SEM image of a cross section showing mainly Ni wires deposited within porous silicon with an average pore diameter of 80 nm.

#### Dipolar Coupling of Metal Structures of Adjacent Pores

Due to the dendritic pore growth of the porous silicon template in the investigated diameter range the magnetic interactions between adjacent pores are enhanced. By producing porous silicon with a reduced branched morphology, which can be carried out by magnetic field assisted etching [53], the average pore distance is reduced and thus leads to weaker magnetic coupling between Ni structures of neighboring pores. In Figure 3.11 a comparison of magnetic field dependent measurements of Ni wires deposited within conventional etched and magnetic field assisted etched templates is shown. One sees that the coercivity of the magnetic field assisted etched sample is about double compared with the conventional etched one because of weaker magnetic coupling between wires of adjacent pores due to the reduction of stray fields concomitant to the reduced dendritic pore growth.

#### Dipolar Coupling of Deposits within the Pores

Considering samples with deposited particles (sphere-like) the magnetic behavior strongly depends on the spatial distribution of deposits within individual pores. Samples containing more densely packed particles offer a lower coercivity due to stronger magnetic interactions between the metal particles within the oriented pores. In the case of Ni deposits within the pores with a much greater distance than their size, the observed coercivities are higher because of weaker magnetic coupling. With increasing distance between the particles the magnetic behavior approximates to the behavior of isolated particles.

Figure 3.11   Magnetic field dependent magnetization curves of two porous silicon samples containing Ni wires. One has been prepared without (full line) and the other one with (dotted line) applied magnetic field. In the case of magnetic field anodization an increase of coercivity and remanence to about twice the value compared with the conventional one has been observed [53].

Different geometries of the incorporated metal structures offer various anisotropic behaviors between the two magnetization directions, with an applied field perpendicular and parallel to the sample surface, respectively. Particles that are deposited in a dense distribution, which means that the distance between particles is in the range of their size, magnetically interact within oriented individual pores, and thus the nanocomposite offers a higher magnetic anisotropy between the two magnetization directions than do particles that are less densely packed [60]. Particles that are deposited with a distance much greater than their size offer only a weak dipolar coupling and therefore no significant magnetic anisotropy can be observed (Figure 3.12).

#### Exchange Coupling of Bimetal Nanostructures

To extend the range of tunable magnetic properties two different metals have been deposited within one sample. In using a hard and soft magnetic phase, the resulting properties can be adjusted by the ratio between the hard and the soft one. Choosing the proper ratio between the two materials a maximum energy product can be achieved [71].

The energy product is determined by the saturation magnetization and the crystalline magnetic anisotropy. In the case of combining hard/soft magnetic materials, exchange coupling between them is exploited to increase the energy product, whereas the hard phase offers a high coercivity and the soft one a high magnetization. Furthermore nanostructuring and producing single domain structures lead to an increase of coercivity. It has been shown theoretically that the diameter of the soft phase structures should not exceed twice the domain wall thickness of the hard phase material [72]. Special attention has to be paid to the volume fraction of two magnetic materials. If the volume fraction of the soft phase becomes bigger, the coercivity decreases and thus the energy product as well. Generally an energy product as high as possible of a bimetallic structure consisting of a hard and a soft magnetic phase can be achieved by choosing a small size of the hard phase but big enough such that the anisotropy does not decrease with the size of the nanostructure. By decreasing a certain size, the structures turn from a stable single domain behavior with high coercivity to a superparamagnetic behavior with negligible coercivity.

In depositing Ni and Co within porous silicon simultaneously, the nanocomposite shows exchange coupling between the two different metal deposits, and one sees that the modification of the volume ratio between the two metals influences the magnetic behavior. In the case of porous silicon containing nanostructures of only one metal (Ni or Co), magnetization measurements show that the coercivity and remanence are determined by the shape and size of the metal deposits, and they are lower compared with samples containing nanostructures composed of both metals.

Figure 3.12   Hysteresis curve of a specimen with Ni particles offering a distance much greater than their average size. The magnetic field has been applied perpendicular (full line) and parallel (dotted line) to the surface, respectively. Due to big distances between the particles and therefore weak magnetic coupling, no significant anisotropy between the two magnetization directions is observed. Inset: Cross-sectional SEM image showing loosely packed Ni particles within the pores [60].

Figure 3.13   Field-dependent magnetization measurements performed with a magnetic field applied parallel and perpendicular to the sample surface, respectively. The hysteresis loops show a high magnetic anisotropy with the easy axis for an applied field parallel to the surface.

#### Magnetic Behavior of Metal-Filled Microporous Silicon

The metal filling of microporous silicon is carried out to influence the optical properties of the samples. The metal deposits offer the same branched morphology as that of the porous silicon template. Due to the size of the pores (2–5 nm) the metal structures should offer superparamagnetic behavior above the transition temperature (which is far below the room temperature), but because of their interconnections, ferromagnetic properties are observed. Field-dependent magnetization measurements have been performed with a magnetic field applied perpendicular and parallel to the sample surface showing a high magnetic anisotropy. Due to the interconnected Ni structures the samples offer a film-like behavior with the easy axis parallel to the surface. Figure 3.13 shows the film-like behavior of a Ni-filled sample. The deposition time of the metal determines the amount of metal inside the porous layer. With increasing amount, the number of interconnections increases, and thus the magnetic anisotropy between the two magnetization directions increases. Furthermore the optical properties, especially the photoluminescence, are modified by metal deposition and also depend on the deposition time. The modification of the photoluminescence is on the one hand a blueshift of the luminescence peak and on the other hand a change in the intensity of the luminescence, and depending on whether there is a spacer (e.g. SiO2) between the metal structures and the silicon skeleton or not, the light emission is enhanced or decreased or quenched, respectively.

#### 3.4  Summary

Porous silicon is a versatile material used in various research fields, such as optics, biomedicine, sensor technology, and also magnetism. Due to the tunable morphology from micro- to macroporous structures, it is appropriate as a template for, e.g. magnetic materials. The arising magnetic properties are tunable by the electrochemical deposition process, which allows the fabrication of spherical, ellipsoidal particles or wires within the porous silicon matrices. Furthermore the packing density of these structures within the pores is adjustable. The magnetic properties are mainly determined by the size and shape of the deposits and by the magnetic coupling between them. So the spatial distribution, especially of particles, strongly influences the magnetic interactions between the particles within the pores. The morphology of the template material, especially the distance between the pores, determines the magnetic coupling between structures within adjacent pores. A further parameter to broaden the palette of the magnetic behavior is the deposition of segmented nanostructures consisting of two different materials. In choosing a hard and a soft magnetic phase, the choice of a proper volume ratio of the two metals leads to a high energy product, which is desirable for various applications such as magnetic data storage. In the case of metal filling within microporous silicon, which offers photoluminescence, the optical properties can be modified by the amount of metal deposition. Since the substrate material is silicon, the presented systems are promising candidates for on-chip applications in today’s microtechnology processes.

#### Acknowledgments

The authors would like to thank the Institute for Electron Microscopy at the University of Technology Graz, Austria, especially Dr P. Poelt, for SEM investigations and the Institute for Solid State Physics at the Vienna University of Technology for making available the magnetometers (SQUID and VSM) for magnetic measurements.

#### References

1
R.P. Cowburn , Science, 311, 183–184, 2006.
2
M. Di Ventra , S. Evoy , J. R. Heflin Jr . (Eds) Introduction to Nanoscale Science and Technology, Series: Nanostructure Science and Technology. Springer: Berlin, 2004.
3
D.A. Allwood , G. Xiong , M.D. Cooke , R.P. Cowburn , J. Phys. D: Appl. Phys., 36, 2175, 2003.
4
G.C. Han , B.Y. Zong , Y.H. Wu , IEEE Trans. Magn., 38, 2562, 2002.
5
H. Ouyang, Ph . M. Fauchet , Biosensing using porous silicon photonic bandgap structures , Proceedings of SPIE: The International Society for Optical Engineering, Boston, MA, 6005, 2005.
6
M. Lundstrom , Science 299, 210, 2003.
7
P. Yang , Nature 425, 243, 2003.
8
E.V. Shevchenko , D.V. Talapin , N.A. Kotov , S. O’Brien , C.B. Murray , Nature 439, 55–59, 2006.
9
B.E. Alaca , H. Sehitoglu , T. Saif , Appl. Phys. Lett. 84, 4669–4671, 2004.
10
H. Masuda , H. Yamada , M. Satoh , H. Asoh , M. Nakao , T. Tamamura , Appl. Phys. Lett. 71, 2770–2772, 1997.
11
G. Fornasieri , A. Bordage , A. Bleuzen , Eur. J. Inorg. Chem. 2018, 259–271, 2018.
12
A. Zhang , J. Zhou , P. Das , Y. Xiao , F. Gong , F. Li , L. Wang , L. Zhang , Y. Cao , H. Duan , J. Electrochem. Soc. 165, D129, 2018.
13
L. Clime , P. Ciureanu , A. Yelon , J. Mag. Mag. Mat. 297, 60–70, 2006.
14
F. Nasirpouri , P. Southern , M. Ghorbani , A. Irajizad , W. Schwarzacher , J. Mag. Mag. Mat. 308, 35–39, 2007.
15
J. Nogues , J. Sort , V. Langlais , V. Skumryev , S. Surinach , J.S. Munoz , M.D. Baro , Phys. Rep. 422, 65–117, 2005.
16
G. Armelles , A. Cebollada , A. Garcia-Martin , M.U. Gonzalez , Adv. Opt. Mater. 1, 10, 2013.
17
R. Herino , Impregnation of porous silicon. In: Properties of Porous Silicon, L. Canham , (Ed). INSPEC: London, 66–76, 1997.
18
C. Fang , E. Foca , S. Xu , J. Carstensen , H. Föll , JES 154, D45, 2007.
19
F. Zacharatos , A.G. Nassiopoulou , Phys. Stat. Sol. (a) 205, 2513, 2008.
20
X.G. Zhang , J. Electrochem. Soc. 151, C69, 2004.
21
K. Kolasinski , Porous silicon formation by galvanic etching . In: Handbook of Porous Silicon, 2nd Edition, L. Canham (Ed). Zug, Heidelberg, New York, Dordrecht, London: Springer International Publishing, 25–37, 2018.
22
S. Yae , Y. Morii , N. Fukumuro , H. Matsuda , Nanoscale Res. Lett. 7, 352, 2012.
23
K.W. Kolasinski , Nanoscale Res. Lett. 9, 432, 2014.
24
E.V. Astrova , T.N. Borovinskaya , A.V. Tkachenko , S. Balakrishnan , T.S. Perova , A. Rafferty , Y.K. Gun’ko , J. Micromech. Microeng. 14, 1022, 2004.
25
K. Rumpf , P. Granitzer , M. Reissner , P. Poelt , M. Albu , ECS Trans. 41, 59, 2012.
26
R. Leancu , N. Moldovan , L. Csepregi , W. Lang , Sens. Actuators A 46–47, 35, 1995.
27
L. Romano , G. Impellizzeri , M.V. Tomasello , F. Giannazzo , C. Spinella , M.G. Grimaldi , JAP 107, 0843.4, 2010.
28
C. O’Dwyer , D.N. Buckley , D. Sutton , S.B. Newcomb , JES 153, G1039, 2006.
29
P. Schmuki , L. Santinacci , T. Djenizian , D.J. Lockwood , Phys. Stat. Sol (a) 182, 51, 2000.
30
M. Leisner , J. Carstensen , A. Cojocaru , H. Föll , Phys. Stat. Sol. (c) 6, 1566, 2009.
31
V. Lehmann , Electrochemistry of Silicon: Instrumentation, Science, Materials and Applications. Wiley-VCH: Weinheim, 2002.
32
E. Vazsonyi , E. Szilagyi , P. Petrik , Z.E. Horvath , T. Lohner , M. Fried , G. Jalsovszky , Thin Solid Films 388, 295, 2001.
33
K.W. Kolasinski , J. Phys. Chem. C 114, 2209., 2010.
34
H. Han , Z. Huang , W. Lee , Nano Today 9, 271, 2014.
35
C. Levy-Clement , Porous silicon formation by metal nanoparticle-assisted etching . In: Handbook of Porous Silicon, 2nd Edition, L. Canham (Ed). Zug, Heidelberg, New York, Dordrecht, London: Springer International Publishing, 61–78, 2018.
36
K.W. Kolasinski , Handbook of Porous Silicon, 1–1 1. Springer International Publishing: Switzerland, 2014.
37
H. Föll , M. Christophersen , J. Carstensen , G. Hasse , Mater. Sci. Eng. R39, 93, 2002.
38
V. Lehmann , Thin Solid Films 255, 1, 1995.
39
M.J. Sailor , Porous Silicon in Practice: Preparation, Characterization and Applications. Wiley-VCH: Weinheim, 2012.
40
M. Schlesinger , M. Paunovic , Modern Electroplating, 1st Edition, The Electrochemical Society. Wiley: New York, 2010.
41
R. Kotha , D. Strickland , A.A. Ayon , Open J. Inorg. Non-Met. Mater. 5, 41, 2015.
42
H. Masuda , M. Yotsuya , M. Ishida , Jpn. J. Appl. Phys. 37, L1090–L1092, 1998.
43
K. Fukami , Y. Tanaka , M.L. Chourou , T. Sakka , Y.H. Ogata , Electrochim. Acta 54, 2197, 2009.
44
P. Granitzer , K. Rumpf , Front. Mater. doi: 3389/fmats.2015.00004.
45
S. Bochmann , A. Fernandez-Pacheco , M. Mackovic , A. Neff , K.R. Siefermann , E. Spiecker , R.P. Cowburn , J. Bachmann , RCS Adv. 7, 37627, 2017.
46
K. Rumpf , P. Granitzer , P. Poelt , Nanoscale Res. Lett. 5, 379, 2010.
47
R.L. Smith , S.F. Chuang , S.D. Collins , Sens. Actuators A21–A23, 825, 1990.
48
J.N. Chazalviel , F. Ozanam , N. Gabouze , S. Fellah , R.B. Wehrspohn , J. Electrochem. Soc. 149, 511, 2002.
49
A.J. Read , R.J. Needs , K.J. Nash , L.T. Canham , P. D. J. Calcott , A. Queish , Phys. Rev. Lett. 69, 1232, 1992.
50
A. Janshoff , K. P. S. Dancil , C. Steinem , D.P. Greiner , V. S. Y. Lin , C. Gurtner , K. Motesharei , M.J. Sailor , J. Am. Chem. Soc. 120, 12108, 1998.
51
V. Lehmann , S. Rönnebeck , J. Electrochem. Soc. 146, 2968–2975, 1999.
52
D. Hippo , Y. Nakamine , K. Urakawa , Y. Tsuchiya , H. Mizuta , N. Koshida , S. Oda , Jpn. J. Appl. Phys. 47, 7398, 2008.
53
P. Granitzer , K. Rumpf , T. Ohta , N. Koshida , M. Reissner , P. Poelt , Appl. Phys. Lett. 101, 033110, 2012.
54
F. Ronkel , J.W. Schultze , R. Arens-Fischer , Thin Solid Films 276, 40, 1996.
55
P. Steiner , F. Kozlowski , W. Lang , Thin Solid Films 255, 49, 1995.
56
M. Jeske , J.W. Schultze , M. Thönissen , H. Münder , Thin Solid Films 255, 63, 1995.
57
K. Fukami , Y. Tanaka , M.L. Chourou , T. Sakka , Y.H. Ogata , Electrochim. Acta 54, 2197, 2009.
58
P. Granitzer , K. Rumpf , P. Poelt , M. Albu , B. Chernev , Phys. Stat. Sol. (c) 6, 2222, 2009.
59
E. Michelakaki , K. Valalaki , A.G. Nassiopouolou , J. Nanopart. Res. 15, 1499, 2013.
60
K. Rumpf , P. Granitzer , G. Hilscher , M. Albu , P. Poelt , Microelectron. Eng. 90, 83, 2012.
61
K. Rumpf , P. Granitzer , H. Michor , P. Pölt , ECS Trans. 75, 57, 2016.
62
K. Rumpf , P. Granitzer , P. Pölt , M. Albu , Phys. Stat. Sol. (c) 8, 1808, 2011.
63
B. Bardet , T. Defforge , B. Negulecu , D. Valente , J. Billoue , P. Poveda , G. Gautier , Mater. Chem. Front. 1, 190, 2017.
64
K. Rumpf , P. Granitzer , M. Albu , P. Poelt , Electrochem. Solid State Lett. 13, K15, 2010.
65
K. Rumpf , P. Granitzer , P. Pölt , M. Albu , K. Ali , M. Reissner , ECS Trans. 33, 203, 2011.
66
P. Granitzer , K. Rumpf , P. Poelt , M. Reissner , Front. Chem. 7, 41, 2019.
67
R. Koda , K. Fukami , T. Sakka , Y.H. Ogata , Nanoscale Res. Lett. 7, 330, 2012.
68
K. Rumpf , P. Granitzer , H. Michor , Nanoscale Res. Lett. 11, 398, 2016.
69
R. Skomsky , J. M. D. Coey , Permanent Magnetism. Institute of Physics Publishing: Bristol, PA, 1999.
70
B.D. Cullity , C.D. Graham , Introduction to Magnetic Materials, 2nd Edition. IEEE Press; Wiley: Chichester; Hoboken, NJ, 2009.
71
R. Skomski , J. M. D. Coey , Phys. Rev. B 48, 15812, 1993.
72
Y. Liu , D.J. Sellmyer , AIP Adv. 6, 056010, 2016.

## Use of cookies on this website

We are using cookies to provide statistics that help us give you the best experience of our site. You can find out more in our Privacy Policy. By continuing to use the site you are agreeing to our use of cookies.