Enfoque Pedagógico

To briefly summarise, the structures developed within this work, are designed, nominally, to operate at a wavelength corresponding to a gas absorption line [chapter 1]. In the case o f the devices that w ill be considered in this thesis, the design wavelength is chosen to be 1680nm Wiich corresponds to a strong absorption line o f CH* The material for the absorber region is GaSb and the materials for the DBRs are GaAs and AlAs. At 1.68pm nAiAs=2.889, ncAs=3.3615,

( 'H. W TER S O p tic a l Results: R i 'E P h o to d etecto r

noasb=3 8Q2, and a; 68nm^5xl0^cm'% see chapter 6, and these parameters will be used in the design o f the structure considered

Tlie structure tliat will be modelled is shown m fig 8 1 The back mirror is highly reflecting and

consists o f 18 period o f GaAs/AlAs, [Mansoor e t a!., 1995], grown on a GaAs substrate. Tlie

design and characterisation o f this structure have already been covered in chapter 5.

Front Reflector |air-semicondiictor interface] GaSb Cavity Back Reflector [18 period DBR] GaAs GaAs Substrate

Figure 8.1 Schematic oj RCE detector stnicture. The .stnicture is a p '-n junction with the DBR being heavily doped n-type and the GaSh being p-type of the order o f iO'^cm'^

The cavity thickness, d, is designed to be some multiple o f mX,/2n and, as seen in chapter 5, a thickness o f 0 647pm was chosen as tliis corresponds to m=3. It was concluded from the modelling carried out m chapter 4 that this thickness could be adequately depleted with ~4V o f reverse bias if the carrier concentration o f the GaSb was in the low lO'^cm'^ Due to the complexities o f tlie growth it was decided tliat the front mirror, in tliese initial structures, should be the air-semiconductor interface, thereby making the structure less complicated. This would give a front mirror reflectivity o f -34% for the value o f the refractive index considered above The back mirror reflectivity is required to be as highly reflecting as possible and M2 was highly reflecting in the centre o f the wafer [F. Mansoor e t a!., 1995].

Modelling the reflectance as a function o f wavelength for the resonant cavity structure for operation at 1 68pm gives tlie spectra o f fig 8 .2. This shows reflectance spectra for tlie structure shown in fig. 8 1 with no absorption and for a=5000cm'V Also shown is a spectra for the DBR as was seen m chapter 6. With no absorption, fig. 8.2[A], there is a small dip in the spectra at the operating wavelength and the reflectance is -9 5 % This dip is due to the destructive interference o f the light reflected from the front mirror and the light reflected fonu the cavity and the back mirror With an increase in the absorption, fig 8 2[B], the reflectance drops to -4% as more o f

CH APTER S O p tic a l Results: R(. 'E P h o to J etecto r

the incident light is absorbed The reflectance spectra o f fig. 8.2[C] is plotted to show the reflectance o f a Bragg reflector stack which is not incorporated into a cavity and is lossless.

0.8 y 0.6 0.2 0.0 14(X) 15(H) 17(H) 19(H) 2(HH) W(n>elength jn n ij

Figure 8.2 Spectra showing réflectance as a Jiinction oJ wavelength for various structures A: resonant cavity stnicture with a= O a n '; B: resonant cavity stnicture with

a=5000ctn' ; C: Bragg reflector stack

As well as the absorption coefficient causing tlie reflectance o f a cavity structure to be minimised, the thickness o f the cavity region also influences the reflectance. The graph o f fig. 8 .3 plots the thickness o f tlie cavity by representing it as the integer m which has been derived in chapter 5. As m increases the reflectance decreases due to the thickness o f the cavity also increasing according to equation 5.3.1.3 in chapter 5. At m=4, the reflectance is nearly zero as more o f the incident light is absorbed and the conditions for minimum reflectance in a cavity are satisfied Tlie absorption coefficient used in the modelling for fig. 8.3 is taken to be 5xl0^cm ’ with a unity back reflector and a front reflector o f 34% to simulate the air-semiconductor interface.

( 'HAPTF.R (S’ O p tic a l R esults: R C E P h o to J etecto r m = l 02- 0.0 I40f.l 1500 1700 1900 2000 Wavelength In m l

Figure 8.3 Reflectance as a function of wavelength fbr m = l-4 with a unity' hack reflector and a 34% front miiror reflectance: a is taken to be 5000cm''

From the results o f fig 8 2, we can model tlie effect o f increasing the absorption coefficient o f the cavity on the reflectance spectra The reflectance decreases, fig 8 4, as tlie absorption coefficient increases up to a=9000cni ' and then increases agaui as the absorption coefficient increases flirther such that at a=2xl0^cni'', the reflectance has mcreased to -18 % from bemg zero at a=9xl0^cm V This is due to the fact that the front mirror is overcompensated at a=2xlO ‘*cm*' and the equations for optimising the cavity given in chapter 5 are not satisfied Tliis increase m reflectance will mean a corresponding decrease m the quantum efficiency o f tlie detector stnicture a

I

Figure 8.4 Reflectance o f a cavity stm cture with va tying value of cavity absorption coefficient fo r unity back reflector and front reflector o f 34%. The cavity thickness is taken

to be d^O.647pm fcorresponding to tn= 3f

CH.4FTER 8 O p tic a l R esults: R C E P h o to d e tec to r

Table 8-1 list the designs o f the various structures grown. In the case o f the resonant cavity structures, reflector stack M2 was used as the back mirror. In all cases the cavity thickness was designed to be some multiple o f A,/2n where 1= 1 .68pm. d=0.647pm was a thickness which we thought would be a good one for an initial structure, D, as we could, in theory, fully deplete this with ~4V reverse bias at an assumed carrier concentration o f ~2xI0'^cm'^. Non-resonant structures were grown in the same growth run with the same absorber region thickness but on GaAs:Si substrates. The second RCE detector structure, E, was also designed to have the same cavity thickness and was grown on stack M2 in order to try and improve on D. Table 8-1 lists the structures and the designed and actual absorber region thicknesses. As can be seen there is a discrepancy between the values o f this thickness and analysis o f these structures from the measured results will be discussed in detail in the following sections.

Table 8-1 Structures o f the resonant cavity and non-resonant cavity devices where Ei, E2

refers to wafer E [device 1 and device 2]

Sample Cavity Thickness designed Measured Cavity thickness [polaron at Oxford} Calculated Cavity Thickness Ifrom Reflectance Measurements] Layer Dupaut Carrier Concentra tion 1cm ^1 D Layerl 0.647pm 1 pm +/- 10% 0.92 pm GaSb p-type -2x10'"

Layer2 n GaAs: Si n-t\pe 2.35x10'*^

Layer 3 n" AlAs: Si n-t\pe 3.15x10'*

La> er 4 n GaAs: Si n-t\pe 2.35x10'*

1 DBR