properties, this can be a useful tool for tailoring material

properties to a particular use.

An important feature of a-Si, regardless of the deposition

technique, is the presence of dangling bonds, which arise due

to the non-periodic nature of amorphous silicon. In order to

minimize the negative effects of these dangling bonds, hydro-

gen is incorporated in the film to compensate for these bonds,

making the material suitable for electronic applications by

reduction of the defect state density and the associated levels

in the energy gap. The dangling bond density in non-hydroge-

on key properties of the material, e.g., the effective energy

gap can range from 1.61 to 1.88 eV as the hydrogen content

varies from 7±21 at.-% and the resistivity can increase by

some two orders of magnitude on moving from zero hydrogen

to about 20 at.-% (in effect the material changes from a semi-

metal to a semiconductor). In addition to this, sputtered a-Si

has excellent adhesion to the substrate and the films can be

deposited for large-area devices limited only by target size

4. Research Program

Our workinvolves the fabrication of thin-film amorphous

silicon and crystalline devices, both of which show promising

results. The thin-film silicon workis founded on a long

research program studying the characteristics and properties

of sputtered amorphous silicon, in particular its hydrogena-

the thin-film PSDs will be described here. Note that the sensi-

tivities reported here are measured directly and are not ampli-

fied. Most of our devices are 20 mm ” 10 mm.

Structure 1: ITO is dc sputtered onto glass followed by an

RF deposition of a-Si:H for about 1 h, which produces films

of approximately 0.5±1 lm thick. A Schottky barrier is then

formed by evaporating Pt onto the a-Si:H layer. Finally two

contacts are made to the Pt layer in the x-direction and two

contacts to the ITO layer in the y-direction.

Structure 2: ITO is dc sputtered onto glass followed by the

deposition of non-hydrogenated a-Si for 5±10 min, followed

by a-Si:H for 1 h. A Schottky barrier is then formed as above

with contacts in the x-direction and y-direction as in Struc-

ture 1.

Structure 3: Al contacts are evaporated onto glass, followed

by 60 min of a-Si:H. A Schottky barrier is formed by evapo-

rating Pt onto a-Si:H and finally, a second pair of contacts is

made to the Pt film (x-direction), at right angles to the Al con-

tacts (y-direction).

5. Results

All of the devices in this workshowed optimal results in

photovoltaic mode and only these results are reported here.

Representative results from each structure are summarized in

Table 1. This includes the sensitivity and nonlinearity mea-

sured for 500 lm increments under a red laser (5 mW) and

focused white light (15 mW). Sensitivities for measurements

along x- and y-directions were taken to assess the potential

for future two-dimensional structures.

For an ideal device, when the voltage output versus distance

moved is plotted, a linear relationship should be found. A

figure of merit of a PSD is its nonlinearity, with a good device

as follows:

Nonlinearity = d =2s/F = 2 Â RMS deviation

Measured full scale

(1)

and it is a measure of the distortion of the sensor output.

Another indication of device quality is the correlation co-

efficient, r, defined mathematically as

y± y=r

(x± x)

(2)

where r is the standard deviation. The quantity r gives a

good indication of device linearity with perfect linearity being

indicated when r is ±1. Another important figure of merit is

the spatial resolution of a device. This indicates the minimum

distance that can be clearly measured when the light spot is

moved from one position to another.

Structure 1: These devices have shown some good linear

results and sensitivities of around 16 lV per increment with

excellent linearities both in the x- and y-directions. An opti-

mized two-dimensional structure is now being designed.

Structure 2: These devices have shown high outputs, an im-

provement on the Structure 1 devices, producing outputs of

350 lV per increment and excellent linearity. We believe that

the non-hydrogenated a-Si acts like a semi-metal and makes

an excellent contact between the a-Si:H and the ITO. A graph

showing the response of this device under white light is shown

in Figure 3.

Structure 3: Measurements were taken between the a-Si

contacts and sensitivities of 2.9 mV per increment obtained,

although with reduced linearity. Increased linearity was found

across the Pt layer but with reduced sensitivities of 12 lV per

increment.

The maximum spatial resolution is calculated by measuring

the nonlinearities of devices for decreasing spatial increments

Adv. Mater. 2001, 13, No. 12±13, Julyl 4

Ó WILEY-VCH Verlag GmbH, D-69469 Weinheim, 2001 0935-9648/01/12±1307-03 $ 17.50+.50/0

3

J. Henry, J. Livingstone/Thin-Film Amorphous Silicon Position-Sensitive Detectors

Table 1. Results for Structures 1, 2, and 3 tested under a red laser and focused

white light with 500 lm increments. The x and y notation indicates measure-

ments made along two directions at right angles on the PSD.