**UV-Vis Photodetector with Silicon Nanoparticles**

J.A. Luna-López1, M. Aceves-Mijares2, J. Carrillo-López1,

A. Morales-Sánchez3, F. Flores-Gracia1 and D.E. Vázquez Valerdi1 *1Science Institute-Research Center for Semiconductor Devices-Autonomous Benemérita University of Puebla 2Department of Electronics, National Institute of Astrophysics, Optics and Electronics INAOE 3Centro de Investigación en Materiales Avanzados S. C., Unidad Monterrey-PIIT, Apodaca, Nuevo León, México* 

### **1. Introduction**

32 Photodetectors

Zhang, B., Wakatsuki, K., Binh, N., Usami, N., & Segawa, Y. (2004). Effects of growth

Zhang, J., Kang, J., Hu, P., & Meng, Q. (2007). Surface modification of poly(propylene carbonate) by oxygen ion implantation. *Appl. Surf. Sci*., Vol.253, pp. 5436-5441

chemical vapor deposition. *Thin Solid Films*, Vol.449, pp. 12-19

temperature on the characteristics of ZnO epitaxial films deposited by metalorganic

Nowadays, photodetector devices are important components for optoelectronic integration. In the past, various photodetector structures have been developed from pn junction, pin diode, bipolar transistor, avalanche photodiodes (APD), and metal-semiconductor-metal (MSM) structures (Ashkan et al., 2008; Hwang Lin, 2005; DiMaria et al., 1984a, 1984b; Sabnis et al., 2005; Foster et al., 2006). In these structures, different semiconductors such as Si, III-V and II-VI compounds have been used, depending on the wavelength range to be detected. Nevertheless, silicon is the most common and important semiconductor in the integrated circuit technology, but it has an indirect band gap inhibiting optical functions. Silicon sensors are usually used in the visible-to near infrared (VIS-NIR) range. However, some commercial Si sensors have been enhanced to detect in the ultraviolet (UV) range, but they have one or more of the following weaknesses: very expensive, reduced responsivity in the VIS-NIR range, lack of compatibility with IC processes and a complex technology. So, most of the available materials for UV detection are not silicon but compound semiconductors (Hwang Lin, 2005). Then, many works have been done to study Si-based optoelectronics materials to overcome the drawback of silicon. Silicon rich oxide (SRO) is one of such materials. SRO is a variation of silicon oxide, in which the content of silicon is changed. The main characteristic to enhance the UV detection in this material is the formation of silicon nanostructures. Among the different techniques to synthesize Si-nps in SRO films, the low pressure chemical vapour deposition (LPCVD) technique offers the films with the best luminescence properties, compared with those obtained by plasma enhanced CVD and silicon implantation into SiO2 films (Morales et al., 2007).

The optical and electrical properties of SRO-LPCVD films have been extensively studied as a function of the silicon excess, temperature and time of thermal annealing (Morales et al., 2007, Zhenrui et al., 2008). Silicon nanoparticles (Si-nps) with different sizes in SRO-LPCVD films have been observed by transmission electron microscope (TEM) technique (Luna et al., 2009). Depending on the excess of Si content, SRO possesses some special properties such as charge trapping, carrier conduction and luminescence. Some novel devices have been

UV-Vis Photodetector with Silicon Nanoparticles 35

In the second part of the experimental procedure SRO films were deposited on n-type Silicon (100) substrates with resistivity of 2000-5000 Ω-cm and with a N+ implanted region on the back side. SRO layers were deposited in a horizontal LPCVD hot wall reactor using SiH4 (silane) and N2O (nitrous oxide) as reactive gases at 700 ºC. The gas flow ratio Ro was used to control the amount of silicon excess in the SRO films. Ro = 10, 20, and 30, corresponding to a silicon excess from 12 to 5 %, were used for this experiment. After deposition, the samples were thermally annealed at 1000° C in N2 atmosphere for 30 minutes. Aluminium grids with area of A = 0.073 cm2 were patterned on the SRO surface by evaporation and standard photo-lithography. Figures 1(a) and 1(b) show a cross-section view scheme, with an approximation at only 3 grid electrodes and an image of the surface of the fabricated devices. The surface Al grid electrodes have several fingers with 40 m width and distanced at 205 m, as observed in figure 1(b). It is important to remark that the zones

(a) (b)

Ellipsometric measurements were made with a Gaertner L117 ellipsometer to obtain the thickness and refractive index of the SRO films before annealing, whose values are shown in

> Ro Refractive index Thickness (Å) 10 1.78 0.01 720 28 20 1.55 0.03 755 25 30 1.46 0.01 591 3

Current versus voltage (I-V) measurements were performed at room temperature, illuminated with UV or white light and under dark conditions, using a computer controlled Keithley 6517A electrometer. The voltage sweep was done with a rate of 0.1 V/s. Illumination was performed with an UV lamp (UVG-54, 5 to 6 eV approximated range) and a white light lamp (1.7 to 4 eV approximated range) with an output power of 4.3 mW/cm2 and 3.2 mW/cm2, respectively. The power of lamps was measured using a radiometer (International Light, USA. IL1 400A). On the other hand, other current versus voltage (I-V)

Fig. 1. Scheme of the Al/SRO/Si MOS-like device, (a) a cross-section view scheme of the fabricated devices with an approximation at only 3 grid electrodes and (b) dimensions of a

between fingers are covered by the SRO film.

device and of the grids electrodes.

Table 1. Refractive index and thickness of SRO films.

Table 1.

proposed using these properties (Zhenrui et al., 2008, Luna et al., 2006, Berman et al., 2008). In Luna et al., 2006, a photodetector using SRO and an induced PN junction were presented, and it was shown that even when the active area is covered by opaque aluminium the photocurrent is considerable. However, it is not clear the reason of such a high response. Also, Aceves et al. proposed (Aceves et al., 1999) that an induced pn junction could be used as a photodetector and that if high resistivity silicon is used a high photoresponse could be expected. However, this idea has not been corroborated.

In this chapter, optical properties as absorption coefficient and optical band gap of SRO with Si-nps and different silicon excess were also studied in order to be able to apply them in optical photodetectors. A photodetector made of a simple Al/SRO/Si MOS-like grid structure is reported. The devices fabricated on high resistivity substrate possess high responsivity in photodetection from UV to VIS-NIR radiation. This structure requires few fabrication process steps and is completely compatible with the complementary metaloxide-semiconductor (CMOS) technology. The possible mechanisms involved in the light detection and the extended photosensitivity of this structure are analysed and discussed.

### **2. Experimental procedure**

The experimental procedure is divided in two parts. The first part is the following: SRO films were deposited on silicon substrates in a horizontal LPCVD hot-wall reactor using nitrous oxide (N2O) and silane (SiH4) as the reactant gases at 700ºC. The partial pressure ratios Ro = [N2O]/[SiH4], were Ro = 10, 20, and 30. The thicknesses of the SRO films were about 500 nm. Silicon Excess in these films are about 12.7, 8.0 and 5.5 at.% for Ro = 10, 20 and 30, respectively (Luna et al., 2007). After deposition, the films were densified at 1000 ºC for 30 minutes in nitrogen atmosphere, and then some of them were thermally annealed at 1100ºC for 180 minutes in nitrogen atmosphere.

The XRD measurements were done using a Bruker AXS D8 Discover diffractometer with Cu K radiation. The surface morphology of SRO films was studied using a nanosurf easy Scan AFM system version 2.3, operated in contact mode. A 4 × 4 μm2 scanned area was used for each topographic image, and a 450-µm-long single-crystal Si cantilever operated at 12 kHz (type vista probes CL contact mode AFM probes) was used. Five different scans were done for each sample, showing good reproducibility. AFM images were analysed using scanning probe image processor (SPIP) software (Jogensen, 2002).

For TEM measurements, an energy-filtered TEM (EFTEM) was used. The images were obtained using an electronic microscope (JEOL JEM 2010F); all the EFTEM images were measured using a Si plasmon of 17 eV and cross-section views. EFTEM enables us to detect amorphous Si nanoclusters embedded in an oxide matrix (Spinella et al, 2005). Also, an FEI Tecnai F30 high-resolution TEM (HRTEM) with acceleration voltage of 300 kV and line resolution of 0.2 nm was used to study the microstructure of the Si nanocrystals.

Room-temperature photoluminescence (PL) of the SRO films was measured using a Perkin-Elmer spectrometer LS-50B model with a xenon source and a monochromator. The samples were excited using 250-nm radiation, and the emission signal was collected from 400 to 900 nm with a resolution of 2.5 nm. A cut-off filter above 430 nm was used to block the light scattered from the source.

Optical transmittance measurements were made in the UV range to near infrared at room temperature, using a spectrophotometer Perkin-Elmer LMBD 3B UV/VIS.

proposed using these properties (Zhenrui et al., 2008, Luna et al., 2006, Berman et al., 2008). In Luna et al., 2006, a photodetector using SRO and an induced PN junction were presented, and it was shown that even when the active area is covered by opaque aluminium the photocurrent is considerable. However, it is not clear the reason of such a high response. Also, Aceves et al. proposed (Aceves et al., 1999) that an induced pn junction could be used as a photodetector and that if high resistivity silicon is used a high photoresponse could be

In this chapter, optical properties as absorption coefficient and optical band gap of SRO with Si-nps and different silicon excess were also studied in order to be able to apply them in optical photodetectors. A photodetector made of a simple Al/SRO/Si MOS-like grid structure is reported. The devices fabricated on high resistivity substrate possess high responsivity in photodetection from UV to VIS-NIR radiation. This structure requires few fabrication process steps and is completely compatible with the complementary metaloxide-semiconductor (CMOS) technology. The possible mechanisms involved in the light detection and the extended photosensitivity of this structure are analysed and discussed.

The experimental procedure is divided in two parts. The first part is the following: SRO films were deposited on silicon substrates in a horizontal LPCVD hot-wall reactor using nitrous oxide (N2O) and silane (SiH4) as the reactant gases at 700ºC. The partial pressure ratios Ro = [N2O]/[SiH4], were Ro = 10, 20, and 30. The thicknesses of the SRO films were about 500 nm. Silicon Excess in these films are about 12.7, 8.0 and 5.5 at.% for Ro = 10, 20 and 30, respectively (Luna et al., 2007). After deposition, the films were densified at 1000 ºC for 30 minutes in nitrogen atmosphere, and then some of them were thermally annealed at

The XRD measurements were done using a Bruker AXS D8 Discover diffractometer with Cu K radiation. The surface morphology of SRO films was studied using a nanosurf easy Scan AFM system version 2.3, operated in contact mode. A 4 × 4 μm2 scanned area was used for each topographic image, and a 450-µm-long single-crystal Si cantilever operated at 12 kHz (type vista probes CL contact mode AFM probes) was used. Five different scans were done for each sample, showing good reproducibility. AFM images were analysed using scanning

For TEM measurements, an energy-filtered TEM (EFTEM) was used. The images were obtained using an electronic microscope (JEOL JEM 2010F); all the EFTEM images were measured using a Si plasmon of 17 eV and cross-section views. EFTEM enables us to detect amorphous Si nanoclusters embedded in an oxide matrix (Spinella et al, 2005). Also, an FEI Tecnai F30 high-resolution TEM (HRTEM) with acceleration voltage of 300 kV and line

Room-temperature photoluminescence (PL) of the SRO films was measured using a Perkin-Elmer spectrometer LS-50B model with a xenon source and a monochromator. The samples were excited using 250-nm radiation, and the emission signal was collected from 400 to 900 nm with a resolution of 2.5 nm. A cut-off filter above 430 nm was used to block the light

Optical transmittance measurements were made in the UV range to near infrared at room

resolution of 0.2 nm was used to study the microstructure of the Si nanocrystals.

temperature, using a spectrophotometer Perkin-Elmer LMBD 3B UV/VIS.

expected. However, this idea has not been corroborated.

**2. Experimental procedure** 

scattered from the source.

1100ºC for 180 minutes in nitrogen atmosphere.

probe image processor (SPIP) software (Jogensen, 2002).

In the second part of the experimental procedure SRO films were deposited on n-type Silicon (100) substrates with resistivity of 2000-5000 Ω-cm and with a N+ implanted region on the back side. SRO layers were deposited in a horizontal LPCVD hot wall reactor using SiH4 (silane) and N2O (nitrous oxide) as reactive gases at 700 ºC. The gas flow ratio Ro was used to control the amount of silicon excess in the SRO films. Ro = 10, 20, and 30, corresponding to a silicon excess from 12 to 5 %, were used for this experiment. After deposition, the samples were thermally annealed at 1000° C in N2 atmosphere for 30 minutes. Aluminium grids with area of A = 0.073 cm2 were patterned on the SRO surface by evaporation and standard photo-lithography. Figures 1(a) and 1(b) show a cross-section view scheme, with an approximation at only 3 grid electrodes and an image of the surface of the fabricated devices. The surface Al grid electrodes have several fingers with 40 m width and distanced at 205 m, as observed in figure 1(b). It is important to remark that the zones between fingers are covered by the SRO film.

Fig. 1. Scheme of the Al/SRO/Si MOS-like device, (a) a cross-section view scheme of the fabricated devices with an approximation at only 3 grid electrodes and (b) dimensions of a device and of the grids electrodes.

Ellipsometric measurements were made with a Gaertner L117 ellipsometer to obtain the thickness and refractive index of the SRO films before annealing, whose values are shown in Table 1.



Current versus voltage (I-V) measurements were performed at room temperature, illuminated with UV or white light and under dark conditions, using a computer controlled Keithley 6517A electrometer. The voltage sweep was done with a rate of 0.1 V/s. Illumination was performed with an UV lamp (UVG-54, 5 to 6 eV approximated range) and a white light lamp (1.7 to 4 eV approximated range) with an output power of 4.3 mW/cm2 and 3.2 mW/cm2, respectively. The power of lamps was measured using a radiometer (International Light, USA. IL1 400A). On the other hand, other current versus voltage (I-V)

UV-Vis Photodetector with Silicon Nanoparticles 37

Ro = 10 **a)** Ro = 30

Ro = 10 **b)** Ro = 30

Fig. 3. 3D AFM images of a) as-deposited and b) thermally annealed at 1100 ºC SRO10 and

Figure 4 shows a comparison of the average roughness <Sa> for the SRO films with different

Silicon

O/SiH4

 As Deposited Densified Annealed

10 20 30

Flow ratio, Ro = N2

Fig. 4. Average roughness <Sa> as a function of the flow ratio (Ro) for as-deposited, densified, and thermally annealed SRO films deposited on silicon (lines are plotted as an

SRO30 films on silicon. Scanned area: 4 × 4 m2.

Si excess for as-deposited, densified and annealed films.

Average roughness <Sa> (nm)

eye-guide). Scanned area: 4 × 4 m2.

measurements were performed at room temperature using a computer controlled Keithley 236 source-meter, under illuminations at specific wavelengths of 400, 650 and 1000 nm conditions, with a Sciencetech 9040 monochromator. The measured power were 83.6 W/cm2, 238.9 W/cm2 and 11.5 W/cm2 for 400, 650, 1000 nm wavelengths, respectively.

### **3. Results**

#### **3.1 Structural, optics and photoelectric results**

Figure 2 shows the XRD patterns of SRO films with Ro = 10 and 20, annealed at 1100 ºC for 180 minutes on silicon substrate. Peaks at 21, 28.5 and 47.4 are observed for SRO films with Ro = 10 and they are ascribed to SiO2, Si (111) and (220), respectively. However, those peaks at 28.5 and 47.4 disappeared for Ro = 20. As Ro becomes higher (Ro = 30), the diffraction peaks were not observed (not shown here).

Fig. 2. XRD patterns of SRO films with Ro = 10 and 20 annealed at 1100 ºC for 180 minutes on silicon substrate.

The average crystal size (D), dislocation density () and microtension () of the silicon nanocrystals (Si-ncs) for Ro = 10 were estimated from the Si (111) peak using the Scherrer formula (Karunagaran, B., et al., 2002, Comedi,D., et al., 2006,) and are listed in Table 2. and decreased when the annealing time increases for SRO10 films, which is related to a reduction on the defect concentration and imperfections of the nanocrystal cell (Yu et al, 2006).


Table 2. Structural parameters of the SRO Films obtained from the XRD measurements for Si (111) peak.

measurements were performed at room temperature using a computer controlled Keithley 236 source-meter, under illuminations at specific wavelengths of 400, 650 and 1000 nm conditions, with a Sciencetech 9040 monochromator. The measured power were 83.6 W/cm2, 238.9 W/cm2 and 11.5 W/cm2 for 400, 650, 1000 nm wavelengths, respectively.

Figure 2 shows the XRD patterns of SRO films with Ro = 10 and 20, annealed at 1100 ºC for 180 minutes on silicon substrate. Peaks at 21, 28.5 and 47.4 are observed for SRO films with Ro = 10 and they are ascribed to SiO2, Si (111) and (220), respectively. However, those peaks at 28.5 and 47.4 disappeared for Ro = 20. As Ro becomes higher (Ro = 30), the diffraction

> **28.5 Si (111)**

**21 a-SiO2**

15 20 25 30 35 40 45 50

2 (degree)

Fig. 2. XRD patterns of SRO films with Ro = 10 and 20 annealed at 1100 ºC for 180 minutes

The average crystal size (D), dislocation density () and microtension () of the silicon nanocrystals (Si-ncs) for Ro = 10 were estimated from the Si (111) peak using the Scherrer formula (Karunagaran, B., et al., 2002, Comedi,D., et al., 2006,) and are listed in Table 2. and decreased when the annealing time increases for SRO10 films, which is related to a reduction on the defect concentration and imperfections of the nanocrystal cell (Yu et al,

Samples of SRO on silicon SRO films 2 (°) [hkl] (D) (nm) (1012 cm-2) (10-3) 10TT30 28.65 Si [111] 4.6 4.6 8.0 10TT180 28.55 Si [111] 5.4 3.4 7.0 20TT180 29.1 Si [111] 2.8 12.2 13.0 Table 2. Structural parameters of the SRO Films obtained from the XRD measurements for Si

**47.3 Si(220)**

**Ro = 10 As-Deposited Ro = 10 for 30 min. Ro = 10 for 180 min. Ro = 20 for 180 min.**

**3. Results** 

**3.1 Structural, optics and photoelectric results** 

peaks were not observed (not shown here).

CPS (a. u.)

on silicon substrate.

2006).

(111) peak.

Fig. 3. 3D AFM images of a) as-deposited and b) thermally annealed at 1100 ºC SRO10 and SRO30 films on silicon. Scanned area: 4 × 4 m2.

Figure 4 shows a comparison of the average roughness <Sa> for the SRO films with different Si excess for as-deposited, densified and annealed films.

Fig. 4. Average roughness <Sa> as a function of the flow ratio (Ro) for as-deposited, densified, and thermally annealed SRO films deposited on silicon (lines are plotted as an eye-guide). Scanned area: 4 × 4 m2.

UV-Vis Photodetector with Silicon Nanoparticles 39

Figure 3 displays three-dimensional (3D) AFM images of as-deposited and thermally annealed SRO films with different excess silicon. All images exhibit a rough surface. For asdeposited SRO films, the roughness decreases with increasing Ro, as shown in Fig. 3 (a). After thermal annealing, roughness increased for SRO films with Ro = 10, as depicted in Fig.

Figure 5 shows the HRTEM and EFTEM images of the annealed SRO films. For Ro=10, HRTEM and EFTEM images clearly show Si nanocrystals (Si-ncs) with a mean size of 5.7 nm, as seen in Figures 5 (a) and 5 (b). Si-ncs were not observed for Ro = 20 and 30. However, bright zones associated with amorphous silicon clusters were observed by EFTEM on SRO films with Ro = 20 and 30, as shown in Fig. 5 (b). Si clusters with a mean size of 2.7 nm were observed for Ro = 20, whereas for Ro = 30, because of low contrast and sharpness, the

400 500 600 700 800 900

A band

A band

Ro =

 10 20 30

5

c)

b)

a)

Wavelength (nm)

Fig. 6. PL spectra of SRO films with Ro = 10, 20 and 30 and a) as- deposited, b) densified at 1000 ºC for 30 min, and c) thermally annealed at 1100º C for 180 minutes deposited on

silicon substrate.

PL intensity (a.u.)

Ro =

Ro =

5

 10 20 30

B band

 10 20 30

3 (b); for Ro = 20 and 30, no significant influence is visible from 3D images.

shape was not very clear, but Si clusters with size 1 nm could be identified.

Fig. 5. Silicon nanocrystals in SRO10 films and silicon clusters in SRO10, SRO20, and SRO30 films were observed by a) HRTEM and b) EFTEM, respectively, after thermal annealing at 1100 °C.

a) HRTEM. Thermally annealed SRO10 film.

b) EFTEM. Thermally annealed SRO10, SRO20, and SRO30 films.

**EFTEM Si Plasmon 17 eV**

**EFTEM Si Plasmon 17 eV** 

1100 °C.

Ro = 10

**Number of Si-Cluster (%)**

Fig. 5. Silicon nanocrystals in SRO10 films and silicon clusters in SRO10, SRO20, and SRO30 films were observed by a) HRTEM and b) EFTEM, respectively, after thermal annealing at

**Number of Si-Nanocrystal (%)**

**Number of Si-Nanocrystal (%)**

**1 2 3 4 5 6 7 8 9 10 11 12**

0123456789

**Diameter (nm)**

Ro = 20 EFTEM dmean = 2.71 nm

**2.0 2.5 3.0 3.5 4.0**

**Diameter (nm)**

Ro = 10 HRTEM

dmean= 5.7 nm

EFTEM

dmean= 4.1 nm

**Diameter (nm)**

Figure 3 displays three-dimensional (3D) AFM images of as-deposited and thermally annealed SRO films with different excess silicon. All images exhibit a rough surface. For asdeposited SRO films, the roughness decreases with increasing Ro, as shown in Fig. 3 (a). After thermal annealing, roughness increased for SRO films with Ro = 10, as depicted in Fig. 3 (b); for Ro = 20 and 30, no significant influence is visible from 3D images.

Figure 5 shows the HRTEM and EFTEM images of the annealed SRO films. For Ro=10, HRTEM and EFTEM images clearly show Si nanocrystals (Si-ncs) with a mean size of 5.7 nm, as seen in Figures 5 (a) and 5 (b). Si-ncs were not observed for Ro = 20 and 30. However, bright zones associated with amorphous silicon clusters were observed by EFTEM on SRO films with Ro = 20 and 30, as shown in Fig. 5 (b). Si clusters with a mean size of 2.7 nm were observed for Ro = 20, whereas for Ro = 30, because of low contrast and sharpness, the shape was not very clear, but Si clusters with size 1 nm could be identified.

Fig. 6. PL spectra of SRO films with Ro = 10, 20 and 30 and a) as- deposited, b) densified at 1000 ºC for 30 min, and c) thermally annealed at 1100º C for 180 minutes deposited on silicon substrate.

UV-Vis Photodetector with Silicon Nanoparticles 41

The fundamental absorption edge in most semiconductors follows an exponential law (Pankove, 1975). The absorption coefficient is correlated with the transmittance T and the

<sup>2</sup> 1 *<sup>d</sup> T Re*

 ln*<sup>T</sup> d*

1/2

Graphs of (ħ)1/2 versus photon energy (ħ) and the corresponding optical band gap are

  (

h)1/2 (eV/cm)1/2

(1)

*Eg* (3)

1234567

Photon Energy (eV)

Ro = 30

Ro = 20

b)

(2)

Ro = 10

30 min 60 min 180 min

reflectance R of a sample with thickness d through the relation:

The absorption coefficient shown in Figure 8 was estimated as:

Ro = 20

123456

Photon Energy (eV)

different Ro values and times of thermal treatment.

Ro = 10

From the Tauc law:

shown in Figure 9.

(

Ro

h)1/2 (eV/cm)1/2

 

a)

Ro = 30

Ro increases. The values obtained for the optical band gap are listed in Table 3.

Densified (1000ºC) Annealed at 1100ºC

different Ro. Straight lines are used to estimate the optical band gap.

Fig. 9. (ħv)1/2 vs. photon energy a) densified and b) thermally annealed SRO films with

It is observed that the position of the absorption edge moves towards higher energy when

10 2.280.04 2.40.04 2.40.02 2.430.04 20 3.520.03 3.570.03 3.6 0.06 3.69 0.03 30 3.76 0.07 3.73 0.06 3.860.04 3.890.04 Table 3. Energy of the optical bang gap obtained from the Tauc's method for SRO films with

The optical band gap increases with Ro. In this case, as the silicon excess reduces the absorption coefficient also reduces, as shown in Figure 8. As Ro moves towards low values,

Band gap (Eg)

Figures 6 (a)–6 (c) show the PL spectra of the as-deposited, densified, and annealed SRO films, respectively. Two PL bands can be observed: B band (400–600 nm) and A band (600– 850 nm). For as-deposited SRO films, the intensity of the B band increases with Ro, as shown in Fig. 6 (a). After densification, the B band disappears and the A band becomes visible [see Fig. 6 (b)]; prolonged annealing at higher temperature of 1100 °C for 180 min enhanced its intensity, as shown in Fig. 6 (c). SRO films with large Ro (low excess Si) always show higher PL intensity. On the other hand, the PL peak position did not show any significant shift with Ro.

Transmittance spectra were measured to obtain the optical band gap of SRO films. Transmittance spectra for Ro = 10, 20 and 30 films deposited on sapphire are shown in Figure 7. As can be observed, the transmittance of all these films is relatively high (> 80 %) between 600 and 900 nm, and reduces to zero for wavelengths below 600 nm. The annealing time does not produce a clear variation; however, the Ro causes a clear shift of the curves towards lower wavelengths. These samples have different behaviours, in agreement with the Ro value and annealing time.

Fig. 7. UV/Vis transmittance spectra for Ro = 10, 20, and 30 on sapphires substrates. a) densified at 1000 °C and b) annealed at 1100 °C.

The absorption coefficients were determined from transmission spectra of Figure 7, and they are shown in Figure 8. The absorption coefficients show different behaviour with the Ro values.

Fig. 8. Absorption coefficients in the range of the U-V to NIR for a) densified and b) thermally annealed SRO films.

The fundamental absorption edge in most semiconductors follows an exponential law (Pankove, 1975). The absorption coefficient is correlated with the transmittance T and the reflectance R of a sample with thickness d through the relation:

$$T \approx \left(1 - R\right)^2 e^{-\alpha d} \tag{1}$$

The absorption coefficient shown in Figure 8 was estimated as:

$$\alpha(\hbar\nu) = \frac{-\ln\left(T(\hbar\nu)\right)}{d} \tag{2}$$

From the Tauc law:

40 Photodetectors

Figures 6 (a)–6 (c) show the PL spectra of the as-deposited, densified, and annealed SRO films, respectively. Two PL bands can be observed: B band (400–600 nm) and A band (600– 850 nm). For as-deposited SRO films, the intensity of the B band increases with Ro, as shown in Fig. 6 (a). After densification, the B band disappears and the A band becomes visible [see Fig. 6 (b)]; prolonged annealing at higher temperature of 1100 °C for 180 min enhanced its intensity, as shown in Fig. 6 (c). SRO films with large Ro (low excess Si) always show higher PL intensity. On the other hand, the PL peak position did not show

Transmittance spectra were measured to obtain the optical band gap of SRO films. Transmittance spectra for Ro = 10, 20 and 30 films deposited on sapphire are shown in Figure 7. As can be observed, the transmittance of all these films is relatively high (> 80 %) between 600 and 900 nm, and reduces to zero for wavelengths below 600 nm. The annealing time does not produce a clear variation; however, the Ro causes a clear shift of the curves towards lower wavelengths. These samples have different behaviours, in agreement with

any significant shift with Ro.

the Ro value and annealing time.

0.0

10<sup>1</sup>

10<sup>2</sup>

10<sup>3</sup>

Absorption coefficient (cm-1)

10<sup>4</sup>

10<sup>5</sup>

0.2

0.4

0.6

Transmittance (a. u.)

0.8

1.0

200 300 400 500 600 700 800 900

Wavelenght (nm)

densified at 1000 °C and b) annealed at 1100 °C.

Energy (eV)

 Ro = 10 Ro = 20 Ro = 30

65 4 3 2

thermally annealed SRO films.

200 300 400 500 600 700 800 900

Wavelength (nm)

 Ro = 10 Ro = 20 Ro = 30

a)

**a**)

Fig. 8. Absorption coefficients in the range of the U-V to NIR for a) densified and b)

Fig. 7. UV/Vis transmittance spectra for Ro = 10, 20, and 30 on sapphires substrates. a)

The absorption coefficients were determined from transmission spectra of Figure 7, and they are shown in Figure 8. The absorption coefficients show different behaviour with the Ro values.

0.0

101

102

103

Absorption coefficient (cm-1)

104

105

0.2

0.4

0.6

Transmittance (a. u.)

0.8

1.0

200 300 400 500 600 700 800 900

200 300 400 500 600 700 800 900

 Ro = 10 Ro = 20 Ro = 30

Wavelength (nm)

 Ro = 10 Ro = 20 Ro = 30

b)

**b**)

Wavelenght (nm)

Energy (eV)

65 4 3 2

$$\left(a\hbar\nu\right)^{1/2} \propto \left(\hbar\nu - E\_{\text{g}}\right) \tag{3}$$

Graphs of (ħ)1/2 versus photon energy (ħ) and the corresponding optical band gap are shown in Figure 9.

Fig. 9. (ħv)1/2 vs. photon energy a) densified and b) thermally annealed SRO films with different Ro. Straight lines are used to estimate the optical band gap.

It is observed that the position of the absorption edge moves towards higher energy when Ro increases. The values obtained for the optical band gap are listed in Table 3.


Table 3. Energy of the optical bang gap obtained from the Tauc's method for SRO films with different Ro values and times of thermal treatment.

The optical band gap increases with Ro. In this case, as the silicon excess reduces the absorption coefficient also reduces, as shown in Figure 8. As Ro moves towards low values,

UV-Vis Photodetector with Silicon Nanoparticles 43

MOS-like structures were fabricated and measured. Figures 10(a) and 10(b) show, respectively, the typical I-V characteristics of devices with Ro = 10 and 30 at the surface inversion condition (negative voltage at the gate respect to the substrate), in dark and under UV and white light illumination. The dark current is on the order of 10-10 A, indicating a low leakage current in these structures. However, a large photocurrent was obtained when the structures were illuminated, this fact indicative of a high optical sensitivity in this simple structure. Similar photo-response was obtained with structures

I-V characteristics of devices with Ro = 10, 20 and 30 at the surface inversion condition and under illumination at specific wavelengths are shown in Figures 11(a), 11(b) and 11(c), respectively. The order of the dark current is 10-8 A indicating a low leakage current in these structures. This current value changed from that obtained in Figure 10 because of the fact that different devices were measured and these ones have a bigger dark current. Nevertheless, at both cases a large photocurrent was obtained when the structures were illuminated. Therefore, these simple structures show high optical sensitivity. Different photocurrents were obtained for the different Ro's in each wavelength, being the SRO

The results show that SRO films properties depend on the silicon excess and the annealing conditions. XRD results showed that in SRO films with thermal annealing, diffusion of the silicon excess took place creating silicon nanoparticles indicated mainly by the peaks, and their size depends upon the silicon excess in the SRO films. Diffractograms on Figure 2 show crystalline orientations (111), (220) and (311) of the nc-Si. It is clear that, when the silicon excess is higher, the agglomerated Si is present and with thermal annealing the generation of nc-Si is obtained, the diffusion coefficient of Si in the SRO at 1100 ºC was calculated as 1 x 10-16 cm2s-1. Diffraction peaks indicate the formation of nc-Si and it should be interpreted as a phase separation between nc-Si and SiO2 induced by thermal

Si-ncs sizes obtained by XRD are similar to those observed by HRTEM and EFTEM. For Ro = 10, HRTEM images clearly show Si-ncs with a mean size of 5.7 nm but Si-ncs were not observed for Ro = 20 and 30. However, bright zones associated with amorphous silicon clusters were observed by EFTEM on SRO films with Ro = 20 and 30. Si clusters with a mean size of 2.7 nm were observed for Ro = 20, whereas for Ro = 30 the shape was not very clear, because of low contrast and sharpness, but Si clusters with 1 nm size could

AFM images of the SRO films on silicon substrates show that the shape and size of the roughness change with Ro. For SRO films with Ro = 10, the roughness increased after thermal annealing. However, for SRO with Ro = 20 and 30, the roughness reduced, as shown in Figure 4. For Ro = 10 the roughness can be as high as a maximum of 24 nm. However, for Ro = 30 the surface looks smooth and the maximum height is about 5 nm. Then, for Ro = 10, the roughness observed through AFM can be associated to Si-ncs. Nevertheless, for SRO films with Ro = 20 and 30 the roughness observed can be ascribed to

structures with Ro = 30 which have the highest photo-response.

made with SRO of Ro = 20.

**4. Analysis and discussion** 

annealing.

be identified.

the band gap moves towards that of the bulk silicon; consequently, it is expected that the properties of an indirect semiconductor be kept.

Fig. 10. I-V curves of the Al/SRO/Si MOS-like grid structure for a) SRO10 and b) SRO30 under dark and illumination conditions.

Fig. 11. Current-voltage characteristics of the Al/SRO/n-Si MOS-like structures with a) SRO10, b) SRO20, and c) SRO30 films under dark and illuminated with 250, 400, 650, 1000 nm wavelengths.

the band gap moves towards that of the bulk silicon; consequently, it is expected that the

10-11 10-10 10-9 10-8 10-7 10-6 10-5 10-4 10-3

10-11 10-10 10-9 10-8 10-7 10-6 10-5 10-4 10-3

Wavelenght Dark Current 400 nm 650 nm 1000 nm

Current (A)

0 2 4 6 8 10 12 14 16 18 20

Negative voltage (V)

Fig. 11. Current-voltage characteristics of the Al/SRO/n-Si MOS-like structures with a) SRO10, b) SRO20, and c) SRO30 films under dark and illuminated with 250, 400, 650, 1000

Current (A)

Al/SRO30/Si

0 1 2 3 4 5 6 7 8 9 10

Vsub (Volts)

0 2 4 6 8 10 12 14 16 18 20

c)

Negative voltage (V)

Wavelenght Dark Current 400 nm 650nm 1000nm

b) **SRO20**

Dark

White light (3.2 mW/cm<sup>2</sup>

UV light (4.3 mW/cm<sup>2</sup>

)

**b)**

)

properties of an indirect semiconductor be kept.

Al/SRO **a)** <sup>10</sup>/Si

0 1 2 3 4 5 6 7 8 9 10

Vsub (Volts)

0 2 4 6 8 10 12 14 16 18 20

Negative voltage (V)

Current (A)

10-11 10-10 10-9 10-8 10-7 10-6 10-5 10-4 10-3 10-2

Wavelenght Dark Current 400 nm 650 nm 1000 nm

**SRO30**

under dark and illumination conditions.

SRO10

Dark

White light (3.2 mW/cm<sup>2</sup>

UV light (4.3 mW/cm<sup>2</sup>

)

Fig. 10. I-V curves of the Al/SRO/Si MOS-like grid structure for a) SRO10 and b) SRO30

a)

)

10-11 10-10 10-9 10-8 10-7 10-6 10-5 10-4 10-3

> 10-11 10-10 10-9 10-8 10-7 10-6 10-5 10-4 10-3

nm wavelengths.

Current (A)

Current (A)

MOS-like structures were fabricated and measured. Figures 10(a) and 10(b) show, respectively, the typical I-V characteristics of devices with Ro = 10 and 30 at the surface inversion condition (negative voltage at the gate respect to the substrate), in dark and under UV and white light illumination. The dark current is on the order of 10-10 A, indicating a low leakage current in these structures. However, a large photocurrent was obtained when the structures were illuminated, this fact indicative of a high optical sensitivity in this simple structure. Similar photo-response was obtained with structures made with SRO of Ro = 20.

I-V characteristics of devices with Ro = 10, 20 and 30 at the surface inversion condition and under illumination at specific wavelengths are shown in Figures 11(a), 11(b) and 11(c), respectively. The order of the dark current is 10-8 A indicating a low leakage current in these structures. This current value changed from that obtained in Figure 10 because of the fact that different devices were measured and these ones have a bigger dark current. Nevertheless, at both cases a large photocurrent was obtained when the structures were illuminated. Therefore, these simple structures show high optical sensitivity. Different photocurrents were obtained for the different Ro's in each wavelength, being the SRO structures with Ro = 30 which have the highest photo-response.

### **4. Analysis and discussion**

The results show that SRO films properties depend on the silicon excess and the annealing conditions. XRD results showed that in SRO films with thermal annealing, diffusion of the silicon excess took place creating silicon nanoparticles indicated mainly by the peaks, and their size depends upon the silicon excess in the SRO films. Diffractograms on Figure 2 show crystalline orientations (111), (220) and (311) of the nc-Si. It is clear that, when the silicon excess is higher, the agglomerated Si is present and with thermal annealing the generation of nc-Si is obtained, the diffusion coefficient of Si in the SRO at 1100 ºC was calculated as 1 x 10-16 cm2s-1. Diffraction peaks indicate the formation of nc-Si and it should be interpreted as a phase separation between nc-Si and SiO2 induced by thermal annealing.

Si-ncs sizes obtained by XRD are similar to those observed by HRTEM and EFTEM. For Ro = 10, HRTEM images clearly show Si-ncs with a mean size of 5.7 nm but Si-ncs were not observed for Ro = 20 and 30. However, bright zones associated with amorphous silicon clusters were observed by EFTEM on SRO films with Ro = 20 and 30. Si clusters with a mean size of 2.7 nm were observed for Ro = 20, whereas for Ro = 30 the shape was not very clear, because of low contrast and sharpness, but Si clusters with 1 nm size could be identified.

AFM images of the SRO films on silicon substrates show that the shape and size of the roughness change with Ro. For SRO films with Ro = 10, the roughness increased after thermal annealing. However, for SRO with Ro = 20 and 30, the roughness reduced, as shown in Figure 4. For Ro = 10 the roughness can be as high as a maximum of 24 nm. However, for Ro = 30 the surface looks smooth and the maximum height is about 5 nm. Then, for Ro = 10, the roughness observed through AFM can be associated to Si-ncs. Nevertheless, for SRO films with Ro = 20 and 30 the roughness observed can be ascribed to

UV-Vis Photodetector with Silicon Nanoparticles 45

with different combinations of Si-Si4-nOn, compounds (with n = 0 to 4), it can be terminated by double bond to an oxygen atom (SiO), oxygen vacancies or by a bond to a hydroxyl

Taking these results into account, it seems that the types of radiative centres producing the luminescence in SRO films deposited on silicon substrate are similar in the A band, but different in the B band. Therefore, depending on the substrate type, specific structural defects are formed in the SRO films. The Si content in the SRO system to form the phase (crystalline or amorphous) and the structure (crystal size and morphology) of the Si-nps, and especially of their surrounding Si-SiO2 interface, has significant influence on the optical properties. Also, the optical band gap tends towards the optical band gap of the Bulk silicon by increasing the silicon excess, and with a lower silicon excess the optical band gap will

The large variety of factors that can be used to influence the structural and optical properties makes these materials very interesting for optoelectronics applications. A shift of the optical band gap is possible by varying the silicon excess. Also, the PL emission can be tuned by

Therefore, MOS-like structures were designed to obtain high photocurrent. The results show that Al/SRO/Si grid structures have high photocurrent at different wavelengths. The large photocurrent is due to the photocarriers generation in the SRO films and also in the Si substrate. The mechanism of the carrier transport in the SRO layers can be via tunneling (Yu et al, 2008). High density of Si-nps and defects in the SRO films create conduction paths and then the photogenerated carriers can move through them allowing a large photocurrent.

This simple structure presents high optical sensitivity for UV and the visible to near infrared range (from 250 nm to 1000 nm), as depicted in Figures 10 and 11. The UV response of the SRO layer has been demonstrated in different reports, apparently the silicon nanoparticles (embedded or not) are in some way sensible to the UV. In (Nayfeh et al, 2004) the authors use silicon nanocrystals (not embedded) to develop an UV sensor in the 250 nm to 350 nm range. Also, our group has developed an UV-Vis-NIR sensor using SRO (with embedded Si nps) (Berman et al., 2008). Moreover, it is well known that red emission is obtained when SRO is irradiated with UV. So, UV response of the SRO could cooperate in two ways to the observed response of the devices reported in this work. First, the red photoluminescence produced by the SRO impinges into the silicon, which has a high sensitivity to these wavelengths. Second, the SRO itself has demonstrated some conduction properties under illumination, so this current component could be added to

Now, the high photoresponse can only be explained if the whole device is considered photosensible. That includes the components due to the SRO (already discussed), and the components in the silicon. As Figure 12 shows, there are two zones in the silicon substrate where the light could produce electron-hole pairs (EHP): (1) under the Al fingers and (2) in the exposed regions between the fingers. Due to the bias of the device, under the Al fingers the silicon surface is in inversion (layers of holes formed), forming an induced-PN junction. However, owing to the fact that this PN junction is covered by an opaque Al layer, very few photons will produce photocurrent; then the current component due to this layer should be

group (Si-O-H).

tend towards that of the SiO2.

the photocurrent (Luna et al, 2010).

thermal annealing, silicon excess and substrate type.

silicon compounds, or silicon clusters, rather than to Si-ncs. Therefore, elemental Si, SiOx, and SiO2 phases separate after annealing, and depending on the silicon excess, one of those phases can be dominant.

Si clusters in the SRO films are produced by diffusion of the Si excess at high temperature. That is, when SRO is annealed, the silicon particles diffuse creating silicon agglomerates around a nucleation site. If the Si excess is high enough, the Si clusters will be crystallized forming Si-ncs. However, when the Si excess reduces, Si clusters are amorphous due to the large lattice mismatch between SiO2 and Si. If the Si excess is low enough, there are no pure Si clusters created, instead, Si-O compounds are formed in SRO films. Then a more homogeneous material is obtained. Of course, these two mechanisms are not exclusive each other, and they can exist simultaneously. However, one of them will dominate depending on the silicon excess.

On the other hand, there are different factors that have an important impact on the optical properties of the SRO films. Therefore, it is necessary to understand these factors for designing materials to be used in photodetectors. In our case, factors such as substrate type, silicon excess, annealing temperature and composition are very important for the optical properties of SRO films. PL also depends on Ro and thermal-annealing conditions. In Figure 6, all the as-deposited SRO films show a weak PL band at 400–600 nm (B band). After thermal annealing, the B band disappears and another PL band (A band) dominates. The intensity of the A band increases with the time and annealing temperature. The PL property of SRO has been extensively studied in literature. Two major mechanisms for PL in this kind of materials are generally accepted: quantum confinement effects in the Si-ncs and defectrelated effects. In our samples, intense PL (A band) in SRO30 was observed; however, SRO30 has silicon compounds rather than Si-ncs. Therefore, the A band cannot be associated with quantum confinement effects; instead, it should be associated with Si-related defects. As reported previously (Lee et al., 1979, Lin et al., 2005), the B band is also associated with F and E defects and different types of oxygen vacancies. Although there are many reports about the origin of PL, most of them relates PL with the bulk microstructures of SRO. A very few studies report on the correlation between the PL and the surface morphology of the SRO films. Other authors (Torchynska et al, 2002) have reported that the diameter and area of the grains on the porous silicon film surface have an important influence on the PL spectra. In our case, SRO is not porous Si but it is considered to be Si nanoclusters embedded in an oxide matrix with very stable optical and electrical properties—quite different than porous silicon. However, from AFM images, the surface is completely irregular, with bumps of different heights and shapes. So, these bumps can be treated as grains on the surface. The grains vary in size depending on the Ro value and annealing time. They can then be related to the emission, as it was done for porous silicon. The height of the roughness, equivalently to the height of the grains, was already related to the silicon cluster size (Luna et al, 2007).

It is observed that the main influence on the structural and optical properties is due to the Ro value of the SRO films, their thickness, annealing time and substrate type. In XRD and AFM measurements important changes in the structure were observed, which indicates that the substrate has influence on the structural properties of the deposited SRO films. In the PL spectra the unusual high intensity of the A band is remarkable. The explanation for this observation is that different oxidation mechanisms lead to different termination of the tetrahedron SiO4 and Si4 of the silicon oxide and silicon, respectively. Si can be terminated

silicon compounds, or silicon clusters, rather than to Si-ncs. Therefore, elemental Si, SiOx, and SiO2 phases separate after annealing, and depending on the silicon excess, one of those

Si clusters in the SRO films are produced by diffusion of the Si excess at high temperature. That is, when SRO is annealed, the silicon particles diffuse creating silicon agglomerates around a nucleation site. If the Si excess is high enough, the Si clusters will be crystallized forming Si-ncs. However, when the Si excess reduces, Si clusters are amorphous due to the large lattice mismatch between SiO2 and Si. If the Si excess is low enough, there are no pure Si clusters created, instead, Si-O compounds are formed in SRO films. Then a more homogeneous material is obtained. Of course, these two mechanisms are not exclusive each other, and they can exist simultaneously. However, one of them will dominate depending

On the other hand, there are different factors that have an important impact on the optical properties of the SRO films. Therefore, it is necessary to understand these factors for designing materials to be used in photodetectors. In our case, factors such as substrate type, silicon excess, annealing temperature and composition are very important for the optical properties of SRO films. PL also depends on Ro and thermal-annealing conditions. In Figure 6, all the as-deposited SRO films show a weak PL band at 400–600 nm (B band). After thermal annealing, the B band disappears and another PL band (A band) dominates. The intensity of the A band increases with the time and annealing temperature. The PL property of SRO has been extensively studied in literature. Two major mechanisms for PL in this kind of materials are generally accepted: quantum confinement effects in the Si-ncs and defectrelated effects. In our samples, intense PL (A band) in SRO30 was observed; however, SRO30 has silicon compounds rather than Si-ncs. Therefore, the A band cannot be associated with quantum confinement effects; instead, it should be associated with Si-related defects. As reported previously (Lee et al., 1979, Lin et al., 2005), the B band is also associated with F and E defects and different types of oxygen vacancies. Although there are many reports about the origin of PL, most of them relates PL with the bulk microstructures of SRO. A very few studies report on the correlation between the PL and the surface morphology of the SRO films. Other authors (Torchynska et al, 2002) have reported that the diameter and area of the grains on the porous silicon film surface have an important influence on the PL spectra. In our case, SRO is not porous Si but it is considered to be Si nanoclusters embedded in an oxide matrix with very stable optical and electrical properties—quite different than porous silicon. However, from AFM images, the surface is completely irregular, with bumps of different heights and shapes. So, these bumps can be treated as grains on the surface. The grains vary in size depending on the Ro value and annealing time. They can then be related to the emission, as it was done for porous silicon. The height of the roughness, equivalently to the height of the grains, was already related to the silicon cluster size (Luna et al, 2007). It is observed that the main influence on the structural and optical properties is due to the Ro value of the SRO films, their thickness, annealing time and substrate type. In XRD and AFM measurements important changes in the structure were observed, which indicates that the substrate has influence on the structural properties of the deposited SRO films. In the PL spectra the unusual high intensity of the A band is remarkable. The explanation for this observation is that different oxidation mechanisms lead to different termination of the tetrahedron SiO4 and Si4 of the silicon oxide and silicon, respectively. Si can be terminated

phases can be dominant.

on the silicon excess.

with different combinations of Si-Si4-nOn, compounds (with n = 0 to 4), it can be terminated by double bond to an oxygen atom (SiO), oxygen vacancies or by a bond to a hydroxyl group (Si-O-H).

Taking these results into account, it seems that the types of radiative centres producing the luminescence in SRO films deposited on silicon substrate are similar in the A band, but different in the B band. Therefore, depending on the substrate type, specific structural defects are formed in the SRO films. The Si content in the SRO system to form the phase (crystalline or amorphous) and the structure (crystal size and morphology) of the Si-nps, and especially of their surrounding Si-SiO2 interface, has significant influence on the optical properties. Also, the optical band gap tends towards the optical band gap of the Bulk silicon by increasing the silicon excess, and with a lower silicon excess the optical band gap will tend towards that of the SiO2.

The large variety of factors that can be used to influence the structural and optical properties makes these materials very interesting for optoelectronics applications. A shift of the optical band gap is possible by varying the silicon excess. Also, the PL emission can be tuned by thermal annealing, silicon excess and substrate type.

Therefore, MOS-like structures were designed to obtain high photocurrent. The results show that Al/SRO/Si grid structures have high photocurrent at different wavelengths. The large photocurrent is due to the photocarriers generation in the SRO films and also in the Si substrate. The mechanism of the carrier transport in the SRO layers can be via tunneling (Yu et al, 2008). High density of Si-nps and defects in the SRO films create conduction paths and then the photogenerated carriers can move through them allowing a large photocurrent.

This simple structure presents high optical sensitivity for UV and the visible to near infrared range (from 250 nm to 1000 nm), as depicted in Figures 10 and 11. The UV response of the SRO layer has been demonstrated in different reports, apparently the silicon nanoparticles (embedded or not) are in some way sensible to the UV. In (Nayfeh et al, 2004) the authors use silicon nanocrystals (not embedded) to develop an UV sensor in the 250 nm to 350 nm range. Also, our group has developed an UV-Vis-NIR sensor using SRO (with embedded Si nps) (Berman et al., 2008). Moreover, it is well known that red emission is obtained when SRO is irradiated with UV. So, UV response of the SRO could cooperate in two ways to the observed response of the devices reported in this work. First, the red photoluminescence produced by the SRO impinges into the silicon, which has a high sensitivity to these wavelengths. Second, the SRO itself has demonstrated some conduction properties under illumination, so this current component could be added to the photocurrent (Luna et al, 2010).

Now, the high photoresponse can only be explained if the whole device is considered photosensible. That includes the components due to the SRO (already discussed), and the components in the silicon. As Figure 12 shows, there are two zones in the silicon substrate where the light could produce electron-hole pairs (EHP): (1) under the Al fingers and (2) in the exposed regions between the fingers. Due to the bias of the device, under the Al fingers the silicon surface is in inversion (layers of holes formed), forming an induced-PN junction. However, owing to the fact that this PN junction is covered by an opaque Al layer, very few photons will produce photocurrent; then the current component due to this layer should be

UV-Vis Photodetector with Silicon Nanoparticles 47

group, generation lifetime of high resistivity silicon was estimated to be between 1 and 2 ms

Now, for UV radiation (200 - 300 nm) the absorption coefficient is approximately, 1x106 cm-1, and the absorption depth is 0.01 m (Sze, 2006). For other wavelengths, for example around 600 nm, the absorption depth is about 5 m. In spite of fact that the absorption is quite superficial for UV, most of the generated holes reach the induced PN junctions due to their long diffusion length, and in general this is true for all the detection range: UV-Vis-NIR. So, the high sensitivity observed in this device is due to a large silicon volume where all the generated carriers contribute to the photocurrent and, as a complement, some photo-effects

The structure with Ro = 30 produces the highest current. However, it is well known that the lower Ro values increase the SRO conductivity, so if the current goes through the SRO, the more conductive structure will be that with Ro = 10. On the other hand, it is also known that silicon excess of 5% (for Ro = 30) produces the highest PL emission (Morales et al., 2005, Luna et al., 2009), and then the more sensible to radiation, where the PL (red emission) generates e-h pares in the induced PN junction. Then, these two mechanisms have to be further studied to determine which one is more adequate to improve the photoresponse of

The influence of different factors on the structural (Si-nc, a-Si, Si-nanocluster, roughness) and optical properties (PL, absorption coefficients, optical band gap) of the SRO films has been discussed from experimental data. PL increased with the annealing time and decreased with a greater amount of silicon excess. In our case, it has been discussed that PL is not due to Si-ncs. Si-clusters are the main nanoparticles that produce maximum PL due to the different type of defects. The formation and growth of Si-ncs and the formation of a-Si in samples with different silicon excess after different annealing times were investigated with XRD and HRTEM. Interestingly, it was found that the optical band gap strongly correlates with the silicon excess and it is shifted over a large energy range via the reactants ratio.

These interesting properties of the SRO films were used to fabricate MOS-like structures, and an Al/SRO/Si MOS-like grid structure was fabricated and tested as photodetector. Silicon-rich oxide films with different silicon excesses were used as dielectric layers and deposited on silicon substrates of very high resistivity. This structure shows high optical sensitivity for the whole UV-Vis-NIR range. The high photocurrent results from three different components. One of them is due to photo-carriers generated in the SRO films; another one comes from the red light that is photo generated in the SRO, as mentioned above. This red light impinges into the high resistivity silicon producing electron-hole pairs. The third component comes from the electron-hole pairs generated in the silicon by the light absorbed in the silicon volume. In spite of the fact that the UV radiation produces electron-hole pairs near the silicon surface, they contribute to the photocurrent because there is not a death zone. Due to the high resistivity of the silicon a long diffusion length is obtained, and the photo generated minority carriers are able to reach the induced PN

(Luna et al., 2005). Then, a value of diffusion of 0.114 cm, or 1140 m is calculated.

in the SRO also contribute to the photocurrent.

our sensor.

**5. Conclusion** 

junctions under the Al grids.

small. So, the PN junctions could act as a collector of the minority carrier produced somewhere else, and then contribute to the photocurrent through the SRO that allows the electrons to move from the substrate to the Al electrode.

Fig. 12. A cross-section scheme of the zones in the SRO films and silicon high resistivity substrate where the light could produce electron hole-pairs (EHP). (1) Under the Al fingers the silicon surface is in inversion due to the bias of the device, in this case only a few photons arrive to the edge of the region, and (2) in the SRO films in the Al uncover between the fingers. The dash lines marks the maximum length where a photon can reach.

On the other hand, the SRO traps charge both positive and negative, so that the surface in this zone could be in accumulation if the SRO is positively charged, in inversion if the SRO is negatively charged, or neutral if there is no charge in the SRO. The resistivity of the silicon substrates used for this experiment is 2000-5000 ohms-cm, so small quantities of trapped charge could bias the surface. Moreover, the charge could get to be trapped during the fabrication process, especially during the high temperature process.

The inversion or accumulation of the surface produces channels where the carriers could move toward the fingered zone where they contribute to the photocurrent. However, in our devices the dark current is very low (< 10-8 A) meaning that the channels do not exist.

Then, the only possibility to increase so much the photocurrent is that the uncovered neutral region under illumination produces electron-hole pairs able to arrive at the Al covered area. As known, only the EHP produced within a diffusion length, *<sup>L</sup> <sup>D</sup>* (D is the diffusivity and is the lifetime), will arrive at the PN junctions. In this case, only the holes which are the minority carriers are significant. The diffusivity can be estimated from the mobility using the Einstein's relationship. Also, the resistivity of the silicon substrate used here is very high and the mobility (510 cm2/Vs) does not vary much for resistivities greater than 100 Ω-cm (Cronemeyer, 1957). Therefore, a conservative value of the diffusivity is 13.184 cm2/s (Grove, 1967). Härkönen et al. (Härkönen et al., 2006) studied the lifetime of high resistivity silicon under low and high injection levels of carriers. They found that for high resistivity silicon the recombination lifetime is longer than 1 ms. Just for reference in our

small. So, the PN junctions could act as a collector of the minority carrier produced somewhere else, and then contribute to the photocurrent through the SRO that allows the

Fig. 12. A cross-section scheme of the zones in the SRO films and silicon high resistivity substrate where the light could produce electron hole-pairs (EHP). (1) Under the Al fingers the silicon surface is in inversion due to the bias of the device, in this case only a few photons arrive to the edge of the region, and (2) in the SRO films in the Al uncover between

On the other hand, the SRO traps charge both positive and negative, so that the surface in this zone could be in accumulation if the SRO is positively charged, in inversion if the SRO is negatively charged, or neutral if there is no charge in the SRO. The resistivity of the silicon substrates used for this experiment is 2000-5000 ohms-cm, so small quantities of trapped charge could bias the surface. Moreover, the charge could get to be trapped during the

The inversion or accumulation of the surface produces channels where the carriers could move toward the fingered zone where they contribute to the photocurrent. However, in our devices the dark current is very low (< 10-8 A) meaning that the channels do not exist.

Then, the only possibility to increase so much the photocurrent is that the uncovered neutral region under illumination produces electron-hole pairs able to arrive at the Al covered area.

and is the lifetime), will arrive at the PN junctions. In this case, only the holes which are the minority carriers are significant. The diffusivity can be estimated from the mobility using the Einstein's relationship. Also, the resistivity of the silicon substrate used here is very high and the mobility (510 cm2/Vs) does not vary much for resistivities greater than 100 Ω-cm (Cronemeyer, 1957). Therefore, a conservative value of the diffusivity is 13.184 cm2/s (Grove, 1967). Härkönen et al. (Härkönen et al., 2006) studied the lifetime of high resistivity silicon under low and high injection levels of carriers. They found that for high resistivity silicon the recombination lifetime is longer than 1 ms. Just for reference in our

(D is the diffusivity

the fingers. The dash lines marks the maximum length where a photon can reach.

fabrication process, especially during the high temperature process.

As known, only the EHP produced within a diffusion length, *<sup>L</sup> <sup>D</sup>*

electrons to move from the substrate to the Al electrode.

group, generation lifetime of high resistivity silicon was estimated to be between 1 and 2 ms (Luna et al., 2005). Then, a value of diffusion of 0.114 cm, or 1140 m is calculated.

Now, for UV radiation (200 - 300 nm) the absorption coefficient is approximately, 1x106 cm-1, and the absorption depth is 0.01 m (Sze, 2006). For other wavelengths, for example around 600 nm, the absorption depth is about 5 m. In spite of fact that the absorption is quite superficial for UV, most of the generated holes reach the induced PN junctions due to their long diffusion length, and in general this is true for all the detection range: UV-Vis-NIR. So, the high sensitivity observed in this device is due to a large silicon volume where all the generated carriers contribute to the photocurrent and, as a complement, some photo-effects in the SRO also contribute to the photocurrent.

The structure with Ro = 30 produces the highest current. However, it is well known that the lower Ro values increase the SRO conductivity, so if the current goes through the SRO, the more conductive structure will be that with Ro = 10. On the other hand, it is also known that silicon excess of 5% (for Ro = 30) produces the highest PL emission (Morales et al., 2005, Luna et al., 2009), and then the more sensible to radiation, where the PL (red emission) generates e-h pares in the induced PN junction. Then, these two mechanisms have to be further studied to determine which one is more adequate to improve the photoresponse of our sensor.

### **5. Conclusion**

The influence of different factors on the structural (Si-nc, a-Si, Si-nanocluster, roughness) and optical properties (PL, absorption coefficients, optical band gap) of the SRO films has been discussed from experimental data. PL increased with the annealing time and decreased with a greater amount of silicon excess. In our case, it has been discussed that PL is not due to Si-ncs. Si-clusters are the main nanoparticles that produce maximum PL due to the different type of defects. The formation and growth of Si-ncs and the formation of a-Si in samples with different silicon excess after different annealing times were investigated with XRD and HRTEM. Interestingly, it was found that the optical band gap strongly correlates with the silicon excess and it is shifted over a large energy range via the reactants ratio.

These interesting properties of the SRO films were used to fabricate MOS-like structures, and an Al/SRO/Si MOS-like grid structure was fabricated and tested as photodetector. Silicon-rich oxide films with different silicon excesses were used as dielectric layers and deposited on silicon substrates of very high resistivity. This structure shows high optical sensitivity for the whole UV-Vis-NIR range. The high photocurrent results from three different components. One of them is due to photo-carriers generated in the SRO films; another one comes from the red light that is photo generated in the SRO, as mentioned above. This red light impinges into the high resistivity silicon producing electron-hole pairs. The third component comes from the electron-hole pairs generated in the silicon by the light absorbed in the silicon volume. In spite of the fact that the UV radiation produces electron-hole pairs near the silicon surface, they contribute to the photocurrent because there is not a death zone. Due to the high resistivity of the silicon a long diffusion length is obtained, and the photo generated minority carriers are able to reach the induced PN junctions under the Al grids.

UV-Vis Photodetector with Silicon Nanoparticles 49

Harkonen, J., E. Tuovinen, Z. Li, P. Luukka, E. Verbitskaya, V. Eremin, Recombination

Hwang, J. & Lin, C. (2005). Gallium nitride photoconductive ultraviolet sensor with a

Jøgensen, J. F., The Scanning Probe Image Processor (SPIP), Denmark, 2002,

Karunagaran, B., Rajendra, R. T., Mangalaraj, D., Narayandass, K., Mohan, G., (2002),

Luna López, J. A., A. Morales Sanchez, M. Aceves-Mijares, Z. Yu, C. Dominguez, Analysis of

Luna-López, A., M. Aceves-Mijares, O. Malik, R. Glaenzer, Low-and high-resistivity silicon

Luna-López, J. A., Aceves M., O. Malik, Z. Yu, A. Morales, C. Dominguez and J. Rickards,

Luna-López, J. A., M. Aceves-Mijares and O. Malik, Optical and electrical properties of silicon rich oxide films for optical sensor, Sensors and actuators A, 2006, 132, 278. Luna-López, J. A., M. Aceves-Mijares, A. Morales-Sánchez, J. Carrillo-López,

Morales A., J. Barreto, C. Domínguez, M. Riera, M. Aceves and J. Carrillo, (2005),

Pankove, J. I., Optical Processes in Semiconductors, (Dover New York, 1975), Chap. 4-6. Sabnis, V. A., Demir, H. V., Fidaner, O, Zheng, J. F., J S Harris, D A B Miller, N Li, T C Wu,

Spinella C., C. Bongiorno, G. Nicotra, and E. Rimini, Appl. Phys. Lett. 87, 044102 2005. Sze, S. M., Physics of semiconductor devices, John Wiley and sons, USA, (2006).

semiconductor, Proc. SPIE 7381 2009, pp. 7381 1H.

Yu Z., M. Aceves-Mijares, J. A. Luna Lopez, J. Deng, Nanocrystalline Si-based metal-oxide-

Yu Z., Mariano Aceves-Mijares, A. Luna-López, Enrique Quiroga and R. López Estopier,

"Photoluminescence and Single Electron Effect Of Nanosized Silicon Materials". In:

silicon-rich oxide films, J. Vac. Sci. Technol. A, 2009, 27(1), 57.

Semiconductor Processing, 2006, 9, 261.

www.imagemet.com.

2002, pp. 1285-1292.

2005, 23 (3), 534-538.

1927.

1255.

(May 2010), pp. 276-279, ISBN 842-6508-233.

Revista Mexicana de Física, S 53(7), 293 (2007).

Technol. A 28(2), Mar/Apr 2010.

Torchynska, T. V., J. Appl. Phys. 92, 4019 2002.

lifetime characterization and mapping of silicon wafers and detectors using the microwave photoconductivity decay (mPCD) technique, Materials Science in

sputtered transparent indium–tin–oxide ohmic contact, Thin Solid Films, Vol.491

Influence of thermal annealing on the composition and structural parameters of DC magnetron sputtered titanium dioxide thin films, Cryst. Res. Technol., Vol. 37 (12),

surface roughness and its relationship with photoluminescence properties of

substrate characterization using the Al/silicon-rich oxide/Si structure with comparison to the metal oxide semiconductor technique, J. Vac. Sci. Technol. A,

Photoconduction in silicon rich oxide films obtained by LPCVD, J. Vac. Sci.

Comparative study between silicon-rich oxide films obtained by LPCVD and PECVD, Physica E: Low-dimensional Systems and Nanostructures, 2007, 38(1), 54. Nayfeh, O. M., S. Rao, A. Smith, J. Therrien, and M. H. Nayfeh, Thin Film Silicon

Nanoparticle UV Photodetector, IEEE Photonics Technology Letters, 2004, 16(8),

H T Chen, Y M Houng (2005), Intimate Monolithic Integration of Chip-Scale Photonics Circuits, IEEE J. selected Topics in Quantum Electronics 2005, 11(6), pp.

### **6. Acknowledgment**

This work has been partially supported by CONACyT-154725, PROMEP-231 and VIEP-BUAP-2011. The authors acknowledge technicians of INAOE laboratory for their help in the samples preparation and measurements.

#### **7. References**


This work has been partially supported by CONACyT-154725, PROMEP-231 and VIEP-BUAP-2011. The authors acknowledge technicians of INAOE laboratory for their help in the

Aceves, M., O. Malik, V. Grimalsky, Application of silicon Rich Oxide Films in New

Aceves, M., W. Calleja , C. Falcony, J. A. Reynoso–Hernández, The Al/Silicon Rich

Ashkan, B.; Jason, J.; Yongho, C.; Leila, N.; Gunham, M.; Wu, Z.; Andrew, G.; Kapur, P. ;

Berman D. M., M. Aceves-Mijares, L. R. Berriel-Valdos, J. Pedraza, A. Vera-Marquina,

Chen Y., T. Cheng, . Cheng, C. Wang, C. Chen, C. Wie, Y. Chen, Highly sensitive MOS

Comedi, D., O.H.Y. Zalloum, E. A. Irving, J. Wojcik, T. Roschuk, M. J. Flynn, and P.

Di Maria, D. J., Kirtley, J. R., Pakulis J., Dong, D. W., Kuan, D., Pesavento, F. L., Theis, T. N.,

DiMaria, D. J., Dong, , Pesavento, D. W., F. L.; Enhanced conduction and minimized charge

Foster J., J. K. Doylend, P. Mascher, A. P. Knights, and P. G. Coleman, Optical attenuation in

Grove, A. S., Physic and technology of semiconductor devices, John Wiley and sons, USA

aluminum oxide membrane, Optics Express 2010, 18(1), pp. 56-62.

annealed silicon-rich silicon oxides, J. Appl. Phys. 99, 023512 (2006). Comedi, D., Zalloum, O.H.Y., Irving, E. A., Wojcik, J., Roschuk, T., Flynn, M. J., and

Optoelectronic Devices, Physics and Chemistry of Solid State 2004, 5(2), pp. 234-

Oxide/Si P-N Induced Junction As A Photodetector, Revista Química Analítica,

Kriehna, C. & Ant, U. (2008). Metal-Semiconductor-metal photodetectrors based on single-walled carbon nanotube film-GaAs Schottky contacts, J. Appl. Phys., Vol.13,

Fabrication, characterization, and optimization of an ultraviolet silicon sensor,

photodetector with wide band responsivity assisted by nanoporous anodic

Mascher, X-Ray-Diffraction study of crystalline Si nanocluster formation in

Mascher, P., (2006), X-Ray-Diffraction study of crystalline Si nanocluster formation in annealed silicon-rich silicon oxides, J. Appl. Phys. 99, 023512 (2006), ISSN 0021-

Cutro, N., Brorson, S. D.; (1984). Electroluminescence studies in silicon dioxide films containing tiny silicon islands, J. Appl. Phys.,Vol. 56(2) (1984), pp. 401-416,

trapping in electrically alterable read-only memories using off-stoichiometric silicon dioxide films, J. Appl. Phys., 1984, 55(8), pp. 3000-3019, ISSN 0021-8979. Donald C., Cronemeyer, Hall and Drift Mobility in High-Resistivity Single-Crystal Silicon,

defect-engineered silicon rib waveguides, J. Appl. Phys., 2006, 99, pp. 073101, ISSN

**6. Acknowledgment** 

**7. References** 

240.

8979

ISSN 0021-8979.

0021-8979.

1967.

Phys. Rev., 1957, 105(2), 522.

1999, 18, 102.

(June 2008), pp. 114315, ISSN 0021-8979

Optical Engineering, 2008, 47(10), 104001.

samples preparation and measurements.


**3** 

*Italy* 

**Silicon Photodetectors Based on Internal** 

Maurizio Casalino, Luigi Sirleto, Mario Iodice and Giuseppe Coppola *Istituto per la Microelettronica e Microsistemi, Consiglio Nazionale delle Ricerche, Naples* 

Silicon Photonics has emerged as an interesting field due to its potential for low-cost optical components integrated with electronic functionality. In the past two decades, there has been growing interest in photonic devices based on Si-compatible materials (Kimerling et al., 2004; Jalali & Fathpour, 2006) in the field both of the optical telecommunications and of the optical interconnects. In this contest, tremendous progresses in the technological processes based on the use silicon-on insulator (SOI) substrates have allowed to obtain reliable and effectiveness full complementary metal-oxide semiconductor (CMOS) compatible optical components such as, low loss waveguides, high-Q resonators, high speed modulators, couplers, and optically pumped lasers (Rowe et al., 2007 ; Vivien et al., 2006 ; Xu et al., 2007 ; Michael et al., 2007; Liu et al., 2007 ; Liu et al., 2006). All these devices have been developed to operate in the wavelength range from C optical band (1528–1561 nm) to L optical band (1561–1620 nm). However one of the crucial steps toward the integration of photonics with electronics resides in the development of efficient chip-scale photodetectors (PD) integrated on Si. Bulk photodetectors are perhaps the oldest and best understood silicon optoelectronic devices. Commercial products in Si operate at wavelengths below 1100 nm, where band-toband absorption occurs. For the realization of photodiodes integrated in photonics circuits operating at wavelengths beyond 1100 nm silicon is not the right material because its transparency. In the last years, in order to take advantage of low-cost standard Si-CMOS processing technology, a number of photodetectors have been proposed based on different physical effects, such as: defect-state absorption (Bradley et al., 2005), two photon absorption (TPA) (Liang et al., 2002) and internal photoemission absorption (Zhu et al., 2008a). Physical effects, working principles, main structures reported in literature and the most significant results obtained in recent years were reviewed and discussed in our previous paper (Casalino et al., 2010a). In this paragraph, we go into more depth on photodetectors based on the internal photoemission effect (IPE). Silicon infrared photodiodes based on IPE are not novel, in fact PtSi/p-Si, Ir/p-Si and Pd2Si/p-Si junctions are usually used in the infrared imaging systems (Kosonocky et al., 1985). The main advantages of these devices resides in their extremely high switching speed and in their simple fabrication process, but, due to high background current density these devices can only work at cryogenic temperature.

**1. Introduction** 

**Photoemission Effect: The Challenge of** 

**Detecting Near-Infrared Light** 

Focus on Nanomaterials Research, Editor: B. M. Carota, ISBN 1-59454-897-8 Nova Science Publishers, Inc., 2008, 233.

Yu, Z., M. Aceves-Mijares, K. Monfil, R. Kiebach, R. Lopez-Estopier, J. Carrillo, Room temperature current oscillations in naturally grown silicon nanocrystallites embedded in oxide films, J. Appl. Phys., 2008, 103, 063706.
