Near-infrared Silicon Schottky Photodiodes based on Non-metallic

Materials

Maurizio Casalino

, Mariano Gioffrè

, Mario Iodice

, Sandro Rao

and Giuseppe Coppola

Institute for Microelectronics and Microsystems (CNR), Via P. Castellino n.111, Naples, Italy

Università degli Studi “Mediterranea” di Reggio Calabria, Dipartimento di Ingegneria dell'Informazione, delle

Infrastrutture e dell'Energia Sostenibile (DIIES), Via Graziella Feo di Vito, 89122, Reggio Calabria, Italy

Keywords: Internal Photoemission Effect, Photodetector, Fabry-Perot, Near-infrared, Silicon, Graphene.

Abstract: In this work we have investigated the performance of Schottky photodetectors based on materials non-

conventionally used to detect near-infrared wavelengths. In the proposed devices the absorption mechanism

is based on the internal photoemission effect. Both three-dimensional (sputtered erbium and evaporated

germanium) and two-dimensional materials (graphene) have been considered and their performance

compared. Our insights show that silicon Schottky photodetectors have the potentialities to play a key role

in the telecommunications opening new frontiers in the field of low-cost silicon photonics.

1 INTRODUCTION

In order to develop all-Si photodetectors (PDs) and

to take advantage of low-cost standard Si-CMOS

processing technology without additional material or

process steps, a number of options have been

proposed: two-photons absorption (TPA) (Casalino,

2010), incorporation of optical dopants/defects with

midbandgap energy levels into the Si lattice

(Casalino, 2010) and internal photoemission effect

(IPE) (Casalino, 2010). IPE is the optical excitation

of electrons in the metal to energy above the

Schottky barrier and then transport of these electrons

to the conduction band of the semiconductor. The

standard IPE theory is due to Fowler (Fowler, 1931).

However, the Fowler’s theory was obtained without

taking into account the thickness of the Schottky

metal layer. The enhancement of IPE in thin metal

film was theoretically investigated by Vicker who

introduce a multiplicative factor to the Fowler’s

formula (Vickers, 1971). In addition, a further

enhancement in IPE can be obtained due to the

increase of the reverse voltage that lowers the

Schottky barrier increasing the amount of emitted

carriers (Casalino, 2010). Concerning PDs, due to

the very low signal-to-noise ratio, for a long time

IPE-based Si PDs at infrared (IR) wavelengths were

believed usable only at cryogenic temperature.

However, in order to enhance the photoemitted

current with respect to the noise (dark) current,

many structures have been proposed based on:

surface plasmon polaritons (SPP) (Berini, 2012),

micro- and nano-metric optical waveguide

(Goykhman 2012), optical microcavity (Casalino,

2006). Although optical microcavity (Casalino,

2009) has been adopted with success to increase

device performance (Casalino, 2008), efficiency of

Schottky PDs based on conventional metals or

silicides remains two order of magnitude lower than

commercial PDs based on III-V compounds (i.e.,

InGaAs).

In this work we investigate the performance of

Schottky PDs based on materials not conventionally

used to detect near-infrared wavelengths. Both three-

dimensional (sputtered erbium and evaporated

germanium) and two-dimensional materials

(graphene) have been considered and their

performance compared. Our insights show that

silicon devices based on IPE are already suitable for

power monitoring applications and they could play a

key role in the telecommunications opening new

frontiers in the field of low-cost silicon photonics.

Casalino, M., Gioffrè, M., Iodice, M., Rao, S. and Coppola, G.

Near-infrared Silicon Schottky Photodiodes based on Non-metallic Materials.

DOI: 10.5220/0005740902590263

In Proceedings of the 4th International Conference on Photonics, Optics and Laser Technology (PHOTOPTICS 2016), pages 261-265

ISBN: 978-989-758-174-8

261

2 IPE ENHANCEMENT BY AN

OPTICAL MICROCAVITY

In 2008, we experimentally demonstrated the

influence of the optical microcavity on IPE for a

Schottky Si junction (Casalino, 2008). In this work

we proposed a device realized by a resonant cavity

Fabry-Perot structure formed by a dielectric bottom

mirror, a metallic top mirror and, in the middle, a

silicon cavity (Fig. 1). The dielectric bottom mirror

was a distributed Bragg reflector (DBR) formed by

alternating layers of amorphous hydrogenated

silicon (a−Si:H) and silicon nitride (Si

) having

λ/4 thicknesses, while, the top mirror was realized

by a copper (Cu) layer working both as absorbing

material and as optical mirror.

Responsivity measurements were carried out in the

range of 1545–1558 nm (step of 0.05 nm) showing a

maximum value of 4.3 µA/W. In addition, a peak

responsivity enhancement, due to the increased

finesse, of about three times, was demonstrated.

Then, in the 2010 we proved that by reducing the

size of the device a further increase in responsivity

of twice (its value was about 8 μA) could be

obtained (Casalino, 2010). Finally, in 2012 we

proved that a significant increase in responsivity

could be obtained by fulfilling the critical coupling

conditions (Casalino, 2012). Main results of

(Casalino, 2012) are reported in Fig. 2 where

responsivity versus the wavelength is characterized

by four distinct peaks deriving from interference

phenomenon. The measured free spectral range of

3.3 nm is in a perfect agreement with the cavity

thickness (100 µm) while the maximum measured

responsivity is 0.063 mA/W. It is worth noting that

although responsivity has been increased due to the

cavity effect, its value remains quite low due to the

intrinsic poor characteristic of the active material

used. In fact, metals are characterized by a high

reflectivity reducing the amount of absorbed photons

that can be converted into electrons. In addition,

carriers excited in states far below the Fermi energy

gets very low probability to overcome the Schottky

barrier. Although efficiency has been increased in

integrated device (Casalino, 2013) it remains quite

poor in vertically-illuminated devices.

Figure 1: Schematic cross section of the PD reported in

(Casalino, 2008).

Figure 2: Responsivity vs wavelength of the device

proposed in (Casalino, 2012).

In this context, the intrinsic limitations of metals

and silicides could be overcome by investigating IPE

through new not conventional metallic materials.

Indeed, in this work, both three-dimensional

(sputtered erbium and evaporated germanium) and

two-dimensional materials (graphene), have been

considered. The advantage in using non-metallic

materials as absorbing active layer is linked to the

fact that metals behaves like mirrors reflecting back

the most part of the incoming radiation. In other

words, only a little amount of the infrared radiation

can be absorbed hindering device efficiency. For

instance our ellipsometric characterizations show

that at 1550 nm reflectivity at Si/Cu and Si/Er

interface are 0.97 and 0.2, respectively, as reported

in Fig. 3. This means that Cu and Er will be able to

absorb 0.03 and 0.8, respectively. Er absorbs much

more than Cu and we will expect that efficiency of

PDs based on Er/Si Schottky junction will be at least

PHOTOPTICS 2016 - 4th International Conference on Photonics, Optics and Laser Technology

262

one order of magnitude higher with respect to PDs

based on Cu/Si.

Figure 3: Reflectivity at Er/Si and Cu/Si interfaces.

3 DEVICE FABRICATION AND

PRELIMINAR RESULTS

Devices reported in Fig. 1 have been fabricated

starting from a bi-polished 200-μm-thick very lightly

doped P silicon. Substrate was chosen slightly doped

in order to avoid free carrier absorption. After a

RCA Si cleaning, substrate has been thermally

oxdidized in order to obtain a 100-nm-thick silicon

dioxide (SiO

The collecting Ohmic contact and the Schottky

contact were both realized on the top of the sample.

The collecting contact was made by a ring of 200-

nm-thick aluminium film, thermally evaporated at

3⋅10

-6

mbar and 150 °C, patterned by a SiO

wet

etching and lift-off process of a Shipley S1813

photoresist which, deposited by a spin-coater at

4000 rpm, had a thickness of 1.4 µm. Then, an

annealing at 475 °C in nitrogen for 30 min, in order

to get a not-rectifying behaviour, was carried out

(Card, 1976).

On the wafer back side, a multilayer Bragg

mirror was fabricated by Plasma Enhanced

Chemical Vapor Deposition technique (PECVD).

The mirror was composed by a quarter-wave stack

of a−Si:H and Si

layers, having nominal

refractive index, at 1550 nm, of 3.52 and 1.82,

respectively. The reflector was realized with five

periods of a−Si:H/Si

pairs, whose nominal

thicknesses are 110 nm and 213 nm, respectively.

Silicon nitride was deposited at pressure of 1.2

mbar, temperature of 250 °C, at 30 W of RF power.

In the deposition chamber 10 sccm of NH

, 88 sccm

of SiH

(5% in He) and 632 sccm of N

are flowed.

The deposition rate was about 23 nm/min and the

suited Si

thickness was obtained with a process

time of about 9 min. Amorphous hydrogenated

silicon, instead, was deposited at pressure of 0.8

mbar, temperature of 250°C, power of 2 W and a

SiH

(5% in He) flow of 600 sccm. The a-Si:H

deposition rate was about 3 nm/min and the desired

thickness was obtained with a process time of about

35 min. A second photolithographic process has

been carried out in order to realize two large PADs

useful to connect the device to the macroscopic

world. To this aim 5-nm-thick chromium (Cr) and

100-nm-thick- Gold (Au) have been thermally

evaporated at 3⋅10

-6

mbar and 150 °C and patterned

by a lift-off process.

Finally, the Schottky contact was fabricated. The

top of the wafer was covered by Shipley S1813

photoresist, exposed and developed in order to

obtain a disk surrounded by the Al ring Ohmic

contact, as shown in Fig. 3.

Figure 4: Top view of the proposed device before

depositing the active material.

Then the active materials was deposited and

eventually patterned. On the device shown in Fig. 4

we have deposited three materials: erbium,

germanium and graphene.

A Radio-Frequency (RF) Sputtering technique

was used for depositing the Erbium (Er) thin film

directly on the device, from a 99,9% pure metal Er

target. The device was placed on the substrate

holder and the deposition chamber was pumped

down to a base pressure of 3·10

–6

mbar before

introducing the process gas (Ar). Er film was then

deposited at r.t., with 30 W RF power, at 2.5·10

–2

mbar pressure, with a constant 40 scan Ar flux and

11 min deposition time. To overcome the target

surface oxidation a 30 min presputtering process at

150 W RF power was necessary before the

Near-infrared Silicon Schottky Photodiodes based on Non-metallic Materials

263

deposition process.

Germanium (Ge) was thermally evaporated at a

pressure of 3

-6

mbar and at a temperature of 150

°C. Substrate was put in rotation in order to increase

the uniformity of the deposited layer resulting 100-

nm-thick film.

Graphene has been grown on a copper foil at

1000°C in a chemical vapour deposition (CVD)

chamber under the flow of CH

and H

. After

growth, PMMA has been deposited on top of

graphene followed by etching of the copper foil.

Then, graphene has been transferred onto the silicon

substrate and the PMMA layer has been removed. A

characterization of the Raman spectrum of graphene

transferred onto substrate has been performed in

order to confirm the monolayer nature of the CVD-

grown graphene.

Figure 5: IV characteristic of the Er/pSi PD.

Figure 6: IV characteristic of the Ge/pSi PD.

We have characterized the I-V behaviour of the

three junctions in order to extract the Schottky

barrier height Φ

(Aubry, 1994). In all three cases,

the rectifying I-V characteristic confirms the

Schottky nature of the junction. The IV

characteristics of the Er/p-Si and Ge/p-Si Schottky

junction have been reported in Fig. 5 and 6,

respectively.

The potential barrier of Er/p-Si and Ge/p-Si are

0.49 eV, 0.50 eV, respectively. Moreover the

graphene/p-Si Schottky barrier has been measured as

0.55 eV. It can be derived that the cut-off

wavelengths for the aforementioned junction are

2.53 μm, 2.48 μm and 2.25 μm, respectively, and

thus they are able to detect near-infrared

wavelength.

In addition, preliminary responsivity

measurements have been carried out with the

experimental set-up shown in Fig. 7. The IR light

beam emitted by a wavelength tunable laser, has

been collimated, chopped and focused onto the

device by a long working distance 50X microscope

objective providing a beam diameter of about 5 µm.

On the other hand, a 20X collecting objective

microscope placed at the back of the device has been

used to addresses light on an InGaAs Near-Ifrared

CCD showing when the light impinges on the active

material disk and thus simplifying the alignment

procedure. A lock-in amplifier measures the

photocurrent produced by our device. A trans-

impedance amplifier is employed to provide a

reverse bias voltage to the photodetector, and at the

same time for reducing the dark current. The dark

current cancellation circuit realized by using a trans-

impedance amplifier has a limited bandwidth,

however it is adequate for our scope, that is DC or

quasi-static measurements. Incident optical power

was measured with the same experimental set-up,

replacing our PD with a calibrate commercial

InGaAs PD whose responsivity is very close to the

unity.

Figure 7: Experimental set-up for external responsivity

measurements.

PHOTOPTICS 2016 - 4th International Conference on Photonics, Optics and Laser Technology

264

Some preliminary measurements show that the

devices based on Ge are characterized by a very low

responsivity of 0.02 mA/W. On the other hand, Er

and graphene are characterized by a responsivity of

about 0.2 mA/W and 0.08 mA/W, respectively.

These values are two and one order of magnitude

higher than the same devices realized with metals,

respectively (Casalino, 2008). Proposed devices

show the potentialities to play a key role in the field

of silicon photonics.

4 CONCLUSIONS

In this paper we have investigated the responsivity

of Schottky photodetectors based on materials non-

conventionally used to detect near-infrared

wavelengths. Both three-dimensional (sputtered

erbium and evaporated germanium) and two-

dimensional materials (graphene) have been

considered and their performance compared. We

have characterized the I-V behaviour of the three

junctions in order to extract the Schottky barrier

height Φ

. The potential barrier of Er/p-Si and Ge/p-

Si are 0.49 eV, 0.50 eV, respectively. Moreover the

graphene/p-Si Schottky barrier has been measured as

0.55 eV. It can be derived that the cut-off

wavelengths for the aforementioned junction are

2.53 μm, 2.48 μm and 2.25 μm, respectively, and

thus they are able to detect near-infrared

wavelength. Some preliminary measurements show

that the devices based on Er and graphene are

characterized by a responsivity of about 0.2 mA/W

and 0.08 mA/W, respectively. These values are two

and one order of magnitude higher than the same

devices realized with metals, respectively. Our

insights show that silicon Schottky photodetectors

have the potentialities to play a key role in the

telecommunications opening new frontiers in the

field of low-cost silicon photonics.

REFERENCES

Casalino, M., Coppola, G., Iodice, M., Rendina, I., Sirleto,

L., 2010. Near Infrared All-Silicon Photodetectors:

State of the Art and Perspectives, Sensors.

Fowler, R. H., 1931. The analysis of photoelectric

sensitivity curves for clean metals at various

temperatures. Physical Review.

Vickers, V. E., 1971. Model of Schottky barrier hot-

electron-mode photodetection. Applied Optics.

Berini, P., Olivieri, A., Chen, C., 2012. Thin Au surface

plasmon waveguide Schottky detectors on p-Si.

Nanotechnology.

Goykhman, I., Desiatov, B., Khurgin, J., Shappir, J., Levy,

U., 2012. Waveguide based compact silicon Schottky

photodetector with enhanced responsivity in the

telecom spectral band. Optics Express.

Casalino, M., Sirleto, L., Moretti, L., Della Corte, F.,

Rendina, I., 2006. Design of a silicon resonant cavity

enhanced photodetector based on the internal

photoemission effect at 1.55 µm. Journal of Optics A:

Pure and applied optics.

Casalino, M., Sirleto, L., Moretti, L., Iodice, M., Rendina,

I., 2009. Back-illuminated silicon resonant cavity-

enhanced photodetector at 1550 nm. Physica E: Low-

Dimensional Systems and Nanostructures.

Casalino, M., Sirleto, L., Moretti, L., Gioffrè, M.,

Coppola, G., Rendina, I., 2008. Silicon resonant cavity

enhanced photodetector based on the internal

photoemission effect at 1.55 micron: Fabrication and

characterization. Appl. Phys. Letters.

Casalino, M., Coppola, G., Gioffrè, M., Iodice, M.,

Moretti, L., Rendina, I., Sirleto, L., 2010. Silicon

technology compatible photodetectors at 1.55 µm.

Journal of lightwave technology.

Casalino, M., Coppola, G., Iodice, M., Rendina, I., Sirleto,

L., 2012. Critically coupled silicon Fabry-Perot

photodetectors based on the internal photoemission

effect at 1550 nm. Optics Express.

Casalino, M., Iodice, M., Sirleto, L., Rendina, I., Coppola,

G., 2013. Low dark current silicon-on-insulator

waveguide metal-semiconductor-metal photodetector

based on internal photoemission effect at 1550 nm.

Journal of Applied Physics.

Card, H. C., 1976. Aluminum-silicon Schottky barriers

and ohmic contacts in integrated circuits. IEEE

Tansactions on electron devices.

Aubry, V., Meyer, F., 1994. Schottky diodes with high

series resistance: limitations of forward I-V methods.

J. of Appl. Phys.

Near-infrared Silicon Schottky Photodiodes based on Non-metallic Materials

265