diameter determination using fluid flow measurements...

8
10ο ΠΑΝΕΛΛΗΝΙΟ ΕΠΙΣΤΗΜΟΝΙΚΟ ΣΥΝΕΔΡΙΟ ΧΗΜΙΚΗΣ ΜΗΧΑΝΙΚΗΣ, ΠΑΤΡΑ, 4-6 ΙΟΥΝΙΟΥ, 2015. Fabrication of free-standing porous silicon and alumina membranes and mean pore diameter determination using fluid flow measurements. M. Savranakis, P. Kontou, M. Dimitropoulos, A. Christoulaki, N. Spiliopoulos, A. A. Vradis, D. L. Anastassopoulos Solid State Laboratory, Department of Physics, University of Patras, GR 26504, Greece Research Area: Materials and Nanotechnology ABSTRACT The purpose of this study is to investigate the dependence of the geometrical characteristics of free-standing porous alumina and silicon membranes by fluid flow experiments. More specifically for porous alumina, anodization of an aluminum foil is carried out in sulfuric acid 0.3M and oxalic acid 0.3M under applied voltages of 20V and 40V respectively in a two-step anodization process. For porous silicon anodization of a n-type silicon wafer is carried out in HF(40%):ethanol electrolyte solution. Electrical contact is achieved with an electrolytic backside contact. Finally, the samples were analyzed with Scanning Electron Microscopy (SEM). 1.INTRODUCTION Porous materials are of scientific and technological importance due to the presence of controllable dimension voids at the nanometer scale. Research efforts in this field have been driven by the rapidly emerging applications such as biosensors, drug delivery, gas separation, energy storage and fuel cell technology [1-3]. The research in this field offers exciting new opportunities for developing new strategies and techniques for the synthesis and applications of these materials. Good control over the structure of porous materials is of fundamental importance in order to tailor and control their properties. Aluminum and silicon present on their surface a thin native oxide layer, 1-10 nm thick. This is the result of material oxidation from the oxygen of the atmosphere and acts as a protective layer against further oxidation of the material. Porous Alumina (PA) : Anodic aluminum oxide (AAO) was reported for the first time fifty years ago. Two different types of AAO exist : the nonporous barrier oxide and the porous oxide. Porous anodic alumina (PAA) is formed via an electrochemical process (anodization). The anodization of aluminum in acidic media results in self-organization of hexagonal array of pores with an interpore distance controlled by the voltage. The porous structure provides a large effective surface area on a small footprint and can, amongst others, be used as templates for the fabrication of nanodots, nanowires and nanotubes [4,5]. To obtain perfectly straight pores the anodization is typically done in a two step process [6,7]. A one-step anodizing process was commonly used for manufacturing nanohole arrays on aluminum until the two- step anodizing process showed more ordered nanopore arrays. This was achieved by stripping away the oxide layer obtained by a first anodization step and subsequently re-anodizing it. The depth of these pores is closely related to the duration of anodizing, and extended anodizing times improve the pore regularity. Porous Silicon (PSi) : Over the last few decades porous silicon (PSi) has attracted remarkable interest due to the high degree of flexibility given by electrochemical etching in order to tune its optical, electrical and mechanical properties. PSi is still actively investigated for the development of sensors [8], filters [9], drug-release materials [10,11], photovoltaic (PV) [12] and thermoelectric systems [13-15] and also finds applications in microelectronics. Despite its long history, the control over PSi porous structure is still under research, particularly for pore sizes between a few tens and a few hundred nanometers. Pore size heavily depends on the doping levels of the silicon substrate and its type (p or n). Several articles were published about the fabrication of free-standing membranes (FSMs) in p-type silicon [16-18], but the limited anisotropy of the etching in this substrate is limited by two factors: (1) the pores always have a conical shape with the top part of the pores significantly larger than the bottom part and (2) such high porosity superficial layers are often damaged due to the surface tension of the liquid, which makes the pore walls collapse [19]. N-type silicon is less investigated because light-assisted etching is not effective at moderate to heavy doping levels and the porosification is performed on a reversed biased electrochemical junction [20]. This means that small variations either in doping concentration or current density translate into a large variation of the porous structure. Despite these difficulties, moderately doped n-type Silicon (0.010.1 Ω cm) is an ideal candidate for the fabrication of porous membranes with well controlled and tailored properties and it overcomes most of the limits found in the p-type material. Membrane detachment is a self-limited process that involves the formation of a thin transitional layer at the bottom of the porous region and, thus, no high burst of current have to be applied to detach them [21]. On the other hand, the doping level, poses a limit to the maximum membrane thickness achievable. This is a striking difference that characterizes moderately doped n-type substrates from both lightly and heavily doped p- and n-type wafers.

Transcript of diameter determination using fluid flow measurements...

Page 1: diameter determination using fluid flow measurements ...pesxm10.chemeng.upatras.gr/sites/default/files/papers/P01... · moderately doped n-type Silicon (0.01–0.1 Ω cm) is an ideal

10ο ΠΑΝΕΛΛΗΝΙΟ ΕΠΙΣΤΗΜΟΝΙΚΟ ΣΥΝΕΔΡΙΟ ΧΗΜΙΚΗΣ ΜΗΧΑΝΙΚΗΣ, ΠΑΤΡΑ, 4-6 ΙΟΥΝΙΟΥ, 2015.

Fabrication of free-standing porous silicon and alumina membranes and mean pore

diameter determination using fluid flow measurements.

M. Savranakis, P. Kontou, M. Dimitropoulos, A. Christoulaki,

N. Spiliopoulos, A. A. Vradis, D. L. Anastassopoulos

Solid State Laboratory, Department of Physics, University of Patras, GR 26504, Greece

Research Area: Materials and Nanotechnology

ABSTRACT

The purpose of this study is to investigate the dependence of the geometrical characteristics of free-standing porous

alumina and silicon membranes by fluid flow experiments. More specifically for porous alumina, anodization of an

aluminum foil is carried out in sulfuric acid 0.3M and oxalic acid 0.3M under applied voltages of 20V and 40V

respectively in a two-step anodization process. For porous silicon anodization of a n-type silicon wafer is carried out in

HF(40%):ethanol electrolyte solution. Electrical contact is achieved with an electrolytic backside contact. Finally, the

samples were analyzed with Scanning Electron Microscopy (SEM).

1.INTRODUCTION

Porous materials are of scientific and technological importance due to the presence of controllable dimension voids

at the nanometer scale. Research efforts in this field have been driven by the rapidly emerging applications such as

biosensors, drug delivery, gas separation, energy storage and fuel cell technology [1-3]. The research in this field offers

exciting new opportunities for developing new strategies and techniques for the synthesis and applications of these

materials. Good control over the structure of porous materials is of fundamental importance in order to tailor and

control their properties.

Aluminum and silicon present on their surface a thin native oxide layer, 1-10 nm thick. This is the result of material

oxidation from the oxygen of the atmosphere and acts as a protective layer against further oxidation of the material.

Porous Alumina (PA): Anodic aluminum oxide (AAO) was reported for the first time fifty years ago. Two different

types of AAO exist : the nonporous barrier oxide and the porous oxide. Porous anodic alumina (PAA) is formed via an

electrochemical process (anodization). The anodization of aluminum in acidic media results in self-organization of

hexagonal array of pores with an interpore distance controlled by the voltage. The porous structure provides a large

effective surface area on a small footprint and can, amongst others, be used as templates for the fabrication of nanodots,

nanowires and nanotubes [4,5]. To obtain perfectly straight pores the anodization is typically done in a two step process

[6,7]. A one-step anodizing process was commonly used for manufacturing nanohole arrays on aluminum until the two-

step anodizing process showed more ordered nanopore arrays. This was achieved by stripping away the oxide layer

obtained by a first anodization step and subsequently re-anodizing it. The depth of these pores is closely related to the

duration of anodizing, and extended anodizing times improve the pore regularity.

Porous Silicon (PSi): Over the last few decades porous silicon (PSi) has attracted remarkable interest due to the

high degree of flexibility given by electrochemical etching in order to tune its optical, electrical and mechanical

properties. PSi is still actively investigated for the development of sensors [8], filters [9], drug-release materials [10,11],

photovoltaic (PV) [12] and thermoelectric systems [13-15] and also finds applications in microelectronics. Despite its

long history, the control over PSi porous structure is still under research, particularly for pore sizes between a few tens

and a few hundred nanometers. Pore size heavily depends on the doping levels of the silicon substrate and its type (p or

n). Several articles were published about the fabrication of free-standing membranes (FSMs) in p-type silicon [16-18],

but the limited anisotropy of the etching in this substrate is limited by two factors: (1) the pores always have a conical

shape with the top part of the pores significantly larger than the bottom part and (2) such high porosity superficial layers

are often damaged due to the surface tension of the liquid, which makes the pore walls collapse [19]. N-type silicon is

less investigated because light-assisted etching is not effective at moderate to heavy doping levels and the porosification

is performed on a reversed biased electrochemical junction [20]. This means that small variations either in doping

concentration or current density translate into a large variation of the porous structure. Despite these difficulties,

moderately doped n-type Silicon (0.01–0.1 Ω cm) is an ideal candidate for the fabrication of porous membranes with

well controlled and tailored properties and it overcomes most of the limits found in the p-type material. Membrane

detachment is a self-limited process that involves the formation of a thin transitional layer at the bottom of the porous

region and, thus, no high burst of current have to be applied to detach them [21]. On the other hand, the doping level,

poses a limit to the maximum membrane thickness achievable. This is a striking difference that characterizes

moderately doped n-type substrates from both lightly and heavily doped p- and n-type wafers.

Page 2: diameter determination using fluid flow measurements ...pesxm10.chemeng.upatras.gr/sites/default/files/papers/P01... · moderately doped n-type Silicon (0.01–0.1 Ω cm) is an ideal

10ο ΠΑΝΕΛΛΗΝΙΟ ΕΠΙΣΤΗΜΟΝΙΚΟ ΣΥΝΕΔΡΙΟ ΧΗΜΙΚΗΣ ΜΗΧΑΝΙΚΗΣ, ΠΑΤΡΑ, 4-6 ΙΟΥΝΙΟΥ, 2015.

2.EXPERIMENTAL

2.1 Experimental Setup

2.1.1 Porous Alumina

Figure 1 shows a schematic illustration of the electrochemical cell used for the PA anodization. As shown in the

figure, a Ni electrode is used as cathode (1) and an aluminum foil as anode (2). The electrodes are connected via a

spring loaded electric contact (3). The electrolyte (4) is either oxalic acid or sulfuric acid. A constant power supply (5)

was used for the applied potential. In order to reduce the bubbles of the electrolyte a stirrer (6) was employed. For the

support of the cell and the alternation of temperature a Cu plate (7) was attached. The temperature is controlled with a

cold plate (8) (peltier device) and the temperature measurements are taken by a mercury thermometer (9).

Figure 2 shows a schematic illustration of the double tank cell for the dissolution of PA barrier layer. As shown, two

Ni electrodes are connected with a power supply providing constant voltage. Each tank is filled with a different mixture

(KCl & H3PO4). The PA membrane is placed between the two tanks. The PA barrier layer is exposed to the tank filled

with H3PO4.

2.1.2 Porous Silicon

Figure 3 shows a schematic illustration of the double tank cell used for the PSi electrochemical etching. As shown, a

Pt mesh was used as the anode and a Ni foil as the cathode. The tanks were filled with a solvent of hydrofluoric acid

and ethanol. A silicon wafer is placed between the two tanks.

2.1.3 Fluid Flow Experiment

Figure 4 presents the experimental setup for the determination of the flow rate through the membranes. It’s

composed of a cylindrical tube with inner diameter Dt=6.82 mm and height h=158 cm. The tube is connected to the

PTFE electrochemical cell, through a conical conjuction in order to avoid bubbles’ formation . The free standing

membrane is located inside the PTFE cell. The tube is engraved per 1 cm in order to monitor the levels of the flowing

liquid.

Figure 1. Electrochemical cell for PA anodization.

(1) Ni electrode

(2) aluminum foil

(3) spring loaded

electric contact

(4) electrolyte

(5) power supply

(6) stirrer

(7) Cu plate

(8) cold plate

(9) mercury

thermometer

Page 3: diameter determination using fluid flow measurements ...pesxm10.chemeng.upatras.gr/sites/default/files/papers/P01... · moderately doped n-type Silicon (0.01–0.1 Ω cm) is an ideal

10ο ΠΑΝΕΛΛΗΝΙΟ ΕΠΙΣΤΗΜΟΝΙΚΟ ΣΥΝΕΔΡΙΟ ΧΗΜΙΚΗΣ ΜΗΧΑΝΙΚΗΣ, ΠΑΤΡΑ, 4-6 ΙΟΥΝΙΟΥ, 2015.

Figure 2. Double tank cell for PA barrier layer dissolution.

Figure 3. Double tank cell for PSi electrochemical etching.

3,0

h0

Figure 4. Experimental setup for the determination of the flow rate through the free-standing membranes.

Page 4: diameter determination using fluid flow measurements ...pesxm10.chemeng.upatras.gr/sites/default/files/papers/P01... · moderately doped n-type Silicon (0.01–0.1 Ω cm) is an ideal

10ο ΠΑΝΕΛΛΗΝΙΟ ΕΠΙΣΤΗΜΟΝΙΚΟ ΣΥΝΕΔΡΙΟ ΧΗΜΙΚΗΣ ΜΗΧΑΝΙΚΗΣ, ΠΑΤΡΑ, 4-6 ΙΟΥΝΙΟΥ, 2015.

2.2 Fabrication of the Samples

2.2.1 Porous Alumina

Thin aluminum foils (100 μm thick, 99.99% purity, (33mm x 33mm) were used. A two step anodization was conducted

at a constant potential of 20V and 40V in 0.3M sulfuric and 0.3M oxalic acid electrolytes respectively, into an

electrochemical cell (fig. 1). The temperature of the electrolyte was kept constant using a cold plate (Peltier device).For

samples prepared in oxalic acid, the duration of the first anodization was 2 hours and the second one was 16 hours in a

steady temperature of 15oC. For samples prepared in sulfuric acid, the duration of the first anodization was 1.5 hours

and the second one was 11 hours in a steady temperature of 10oC. Between the two steps the oxide layer was removed

by immersing the aluminum foils for 3 hours in a mixture of phosphoric acid (6 wt%) and chromic acid (1.8 wt%) at

50°C. After the anodizations the aluminum layer was removed with the chemical procedure of wet-etching in a solution

of HCl and CuCl2 [22]. Subsequently, the membranes’ barrier layers were dissolved in the double tank cell (fig 2) using

phosphoric acid 5%wt and 0.2M KCl, in order to obtain the free-standing membranes [23].

2.2.2 Porous Silicon

Samples were fabricated using n-type silicon wafer (0.15-0.5 Ω cm), polished on one side and oriented along the [100]

crystalline direction. Anodization was performed at room temperature in the absence of light and a computer-controlled

power supply was used to tune the applied current density. Etching solutions were composed of aqueous HF 40% and

ethanol and the electrochemical etching took place in a double tank cell (fig. 3). First, a 16% HF solution was used, with

an applied current density of 60 mA/cm2 for 30 minutes in order to create the PS layer. Subsequently, when the PSi

layer on the sample is formed, a second etching takes place which is necessary to detach the PS layer from the Si

substrate to form the membrane with an applied current density of 120 mA/cm2. The second etching process led to a

double porosity layer of PSi. As a result, the first PS layer is completely detached from its substrate. After fabrication of

the membrane, the sample was rinsed gently with ethanol. Then a final rinse was carried out with hexane in order to

minimize the possibility of shattering of the membrane due to strong capillary forces and thermal stresses exerted when

ethanol evaporates from the pores. The membrane was dried in a N2 flow.

2.2.3. Fluid Flow Experiment

Thickness, pore density and pore diameter of the membrane were measured after imaging with Scanning Electron

Microscopy (SEM). A second method for the estimation of PA and PSi membranes pore diameter was by using them in

a fluid flow experiment with toluene as flowing liquid (fig.4). The hydrostatic pressure imposed by the fluid column

above the membrane is at the order of 10kPa and acts as the driving force for the flow. We measured the time it takes

for the solvent level to drop through the engraved marks.

Poiseuille’s law gives the flow rate of a fluid through a cylindrical channel due to a pressure difference by:

(1)

where ΔP is the pressure difference, R the radius of the tube, η the dynamic viscosity of the solvent and l the tube

length.

From a macroscopic point of view, the rate of flow can be described as the volume of solvent passing through the glass

tube in a period of time:

(2)

where St is the surface area of the glass tube, dh/dt the rate of change in the height level of the solvent. By rearranging

equations (1) and (2) one can obtain:

(3)

where B is the slope of the graph plot :

= f(t).

For the pore radius from (3) we obtain:

(4)

Page 5: diameter determination using fluid flow measurements ...pesxm10.chemeng.upatras.gr/sites/default/files/papers/P01... · moderately doped n-type Silicon (0.01–0.1 Ω cm) is an ideal

10ο ΠΑΝΕΛΛΗΝΙΟ ΕΠΙΣΤΗΜΟΝΙΚΟ ΣΥΝΕΔΡΙΟ ΧΗΜΙΚΗΣ ΜΗΧΑΝΙΚΗΣ, ΠΑΤΡΑ, 4-6 ΙΟΥΝΙΟΥ, 2015.

3. RESULTS AND DISCUSSION

3.1 Fabrication of PA membranes

Figure 5 shows the electrical current as a function of time during aluminum anodization in potensiostatic mode for

constant applied voltage of 40 V and 20V respectively. During the first step a layer of dense aluminum oxide is created

and therefore the current decreased. In the next step creation of the first pores starts and an increase in the current is

observed. The steady state of the current indicates the development of the pores in a steady growing rate. In figure 6 we

observe that for the creation of the 20 nm pores the current stabilizes in a bigger value because the electrolyte (sulfuric

acid) has greater conductivity than the electrolyte used for the 40 nm pores (oxalic acid).

Figure 7 shows the electrical current as a function of time during the dissolution of barrier layer of PA. During the

first step we observe a minus electrical conductivity which is steady until the barrier layer starts the dissolution process.

The increase of the current indicates a rapid rate of barrier layer destruction. The steady state in the last step shows a

higher electrical conductivity which means the total destruction of the barrier layer. The procedure is the same for both

20V and 40V. During the PA anodization processes, the pore diameter is related to the applied voltage and the

membrane’s thickness depends on the duration of the second anodization.

Figure 5. Current vs anodization time for applied

voltage of 40V.

4000 6000 8000

0,000

0,002

0,004

0,006

0,008

0,010

I(A)

t(s)

1

2

3

4 5

Figure 7. Current vs time for the dissolution of

barrier layer of PA.

0 100 200 300

0,01

0,02

i(A

)

t(s)

Figure 6. Current vs anodization time for

applied voltage of 20V.

.

0 10000 20000

0,00

0,02

0,04

P.A 40nm SLOPE=9,8

P.A 20nm SLOPE=1,1

-ln

(h(t)/h0)

t(s)

Figure 8. Height level of fluid as a function of

time.

3.2 Fluid Flow Experiment of PA membranes

In figure 8 we observe the level of liquid column as a function of time for PA (40 V and 20V) fluid flow experiment.

As the flowing liquid we used toluene. Toluene’s density (ρ), gravity acceleration (g), toluene’s viscosity (η), surface

effective area ( ) and tube inner diameter (Dt) were kept constant. The pore areal density and the

thickness in each membrane depend on the applied voltage and were measured by SEM.

Page 6: diameter determination using fluid flow measurements ...pesxm10.chemeng.upatras.gr/sites/default/files/papers/P01... · moderately doped n-type Silicon (0.01–0.1 Ω cm) is an ideal

10ο ΠΑΝΕΛΛΗΝΙΟ ΕΠΙΣΤΗΜΟΝΙΚΟ ΣΥΝΕΔΡΙΟ ΧΗΜΙΚΗΣ ΜΗΧΑΝΙΚΗΣ, ΠΑΤΡΑ, 4-6 ΙΟΥΝΙΟΥ, 2015.

According to Poiseuille’s law, the radius of the pores for applied voltage of 40V was calculated from (4) and was

found 19 nm. For applied voltage of 20V the radius was found 10 nm.

3.3 Scanning Electron Microscopy (SEM)

Figures 8,9,10 are images of PA taken by Scanning Electron Microscopy for constant applied voltage of 40V. To

measure pore diameter we used ImageJ. The porous alumina membranes can be considered as porous medium with a

number of parallel cylindrical nanotubes penetrating throughout its whole thickness. The average pore diameter was

found 44 nm, the pore density 1,2x1010 and the membrane’s thickness 42,61μm. Figures 11,12,13 are images of PA for

constant applied voltage 20V.The average pore diameter was found 27 nm ,the pore density 2,4*1010 cm2 and the

membrane thickness 80,6 μm.

Figure 9. Top view of PA 40nm.

Figure 11. Cross section of PA 40nm. The

membrane thickness is found 42.61 μm.

Figure 13. Cross section of PA 20nm.

Figure 10. Bottom View of PA 40nm.

Figure 12. Top view of PA 20nm.

Figure 14. Cross section of PA 20nm. The membrane

thickness is found 80.6 μ

Page 7: diameter determination using fluid flow measurements ...pesxm10.chemeng.upatras.gr/sites/default/files/papers/P01... · moderately doped n-type Silicon (0.01–0.1 Ω cm) is an ideal

10ο ΠΑΝΕΛΛΗΝΙΟ ΕΠΙΣΤΗΜΟΝΙΚΟ ΣΥΝΕΔΡΙΟ ΧΗΜΙΚΗΣ ΜΗΧΑΝΙΚΗΣ, ΠΑΤΡΑ, 4-6 ΙΟΥΝΙΟΥ, 2015.

3.4 Porous Silicon

Figure 15 presents the anodizing voltage as a function of time during silicon anodization in galvanostatic mode under

constant current density of 60 mA/cm2 . During the first seconds of the process the voltage is increasing until the pore

creation is initiated. The peak in the voltage corresponds to the pore opening.

50 100 150 200 250 300

20

25

30

35

V (

V)

t (s)

Figure 15. Anodization voltage as a function of time for constant current density of 60 mA/cm2

There is a high variety in porous structures that can be created through electrochemical etching on PSi, with altering the

conditions of the process such as current density, HF concentration, wafer doping, illumination (n-type silicon). The

porous media varies from straight and smooth pores to 3D and branched with pores sizes varying from a few tens of

nanometers to micrometers, sometimes on the same membrane. We tried to simulate porous media similar to porous

structures for oil and mineral extractions. Some of the resulting samples are shown in figures 16-19 taken by Scanning

Electron Microscopy (SEM):

Figure 16. Top View of PSi etched at 30

mA/cm2.

Figure 18. Free Standing PSi membrane etched at

60 mA/cm2

Figure 17.Cross Section of PSi etched at 30 mA/cm2.

Figure 19. Layer below free standing PSi

membrane etched at 60 mA/cm2

Page 8: diameter determination using fluid flow measurements ...pesxm10.chemeng.upatras.gr/sites/default/files/papers/P01... · moderately doped n-type Silicon (0.01–0.1 Ω cm) is an ideal

10ο ΠΑΝΕΛΛΗΝΙΟ ΕΠΙΣΤΗΜΟΝΙΚΟ ΣΥΝΕΔΡΙΟ ΧΗΜΙΚΗΣ ΜΗΧΑΝΙΚΗΣ, ΠΑΤΡΑ, 4-6 ΙΟΥΝΙΟΥ, 2015.

Figure 20. Cross Section of PSi membrane

etched at 60 mA/cm2. The membrane thickness is

found 41.14 μm

Figure 21.Cross section of PSi membrane etched

at 60 mA/cm

4 CONCLUSIONS

The temperature of the electrolyte affects the speed of the anodization procedure. A higher conductivity is

observed due to a different type of electrolyte. The results for the pore diameter calculated by the Poisseuille's law, have

a standard error of 10% in comparison to the results we got from the SEM images, where a regularity of the pore’s

development is observed.

REFERENCES

[1]. A.M. Abd-Elnaiem and A. Gaber, Int. J. Electrochem. Sci., 8 (2013) 9741.

[2]. Y.-H.Pai, C.-W.Tseng, and G.-R.Lin, J. Electrochem. Soc., 159(2012) E99.

[3]. A. O. Araoyinbo, A. Rahmat, M. N. Derman, and K. R. Ahmad , Adv. Mat. Lett., 3(2012) 273.

[4]. V. Hardev Singh, Dig. J. Nanomater. Bios., 5( 2010) 593.

[5]. A. Ghahremaninezhad, and A. Dolati, J. Alloy. Compd., 480 (2009) 275.

[6]. H. Masuda, and K. Fukuda, Science, 268(1995) 1466.

[7]. TY Kima, and SH Jeong, Korean J. Chem. Eng., 25(2008) 609.

[8]. J. Alvarez, P. Bettotti, I. Suarez, N. Kumar, D. Hill, V. Chirvony, L. Pavesi, Opt. Express 19, 26106 (2011)

[9]. L.M. Bonanno, L.A. DeLouise, Biosens. Bioelectron. 23, 444 (2007)

[10]. E.C. Wu, J.S. Andrews, L.Y. Chen, W.R. Freeman, L. Pearson, M.J. Sailor, Biomaterials 32, 1957 (2011)

[11]. E. Gultepe, D. Nagesha, S. Sridhar, M. Amiji, Adv. Drug. Deliver. Rev. 62, 305 (2010)

[12]. A. Wolf, B. Terheiden, R. Brendel, Prog. Photovoltaics 15, 199 (2007)

[13].G. Yuan, R. Mitdank, A. Mogilatenko, S.F. Fischer, J. Phys. Chem. C 116, 13767 (2012)

[14]. J. de Boor, D.S. Kim, X. Ao, D. Hagen, A. Cojocaru, H. Foll, V. Schmidt, EPL 96, 16001 (2011)

[15]. J. Tang, H.-T. Wang, D.H. Lee, M. Fardy, Z. Huo, T.P. Russell, P. Yang, Nano Lett. 10, 4279 (2010)

[16]. J.H. Peterman, D. Zielke, J. Schmidt, F. Haase, E.G. Rojas, R. Brendel, Prog. Phtovol. Res. Appl. 20, 1 (2012)

[17]. M. Ghulinyan, C.J. Oton, G. Bonetti, Z. Gaburro, L. Pavesi, J. Appl. Phys. 93, 9274 (2003)

[18]. Michael J. Sailor, Porous Silicon in Practice: Preparation, Characterization and Applications (2012)

[19]. V. Lehmann, R. Stengl, A. Luigart, Mater. Sci. Eng. B 69(70), 11–22 (2000)

[20]. F.A. Harraz, S.M. El-Sheikh, T. Sakka, Y.H. Ogata, Electrochim. Acta 53, 6444 (2008)

[21].N. Kumar, S. Gennaro, P. Vallachira, W. Sasikumar, G.D. Soraru. P. Bettotti Appl. Phys. A (2014) 116:251–257

[22]. A.Christoulaki et al., J.Appl.Electrochem (2014) 44, 701.

[23]. M. Lillo et al., Journal of Membrane Science (2009) 327, 11.