Los procesos clave y las debilidades de FORVIC vistas desde el concepto de paz positiva

An initial step towards the control of the nitrogen composition is to study the influence of the nitrogen flux on the nitrogen composition of layers grown by plasma assisted-MBE (at fixed plasma power). Three sets of InAsN epilayers were grown on InAs substrates with different nitrogen flux at fixed RF powers, growth temperatures and As fluxes. The nitrogen composition of these samples is plotted as a function of the nitrogen flux in Figure 5.8. A first set, grown at 160 W with varying nitrogen flux, shows a linear increase in the nitrogen composition (up to 0.96 %) with increasing N flux. Such a behaviour was previously reported for dilute nitrides grown by plasma assisted-MBE, in GaInNAs/GaAs MQW [42] and GaNAs [121, 159]. Two other batches of samples were grown at RF powers of 210 W and 260 W. A radically different behaviour is observed at higher RF powers and a decrease in the nitrogen composition is observed with increasing N flux.

Figure 5.8 Nitrogen composition as a function of the N flux at fixed RF power. The solid line is

a linear fit and dashed lines are a guide to the eye

The dependence of the nitrogen composition on the nitrogen flux at fixed RF power can be understood in the light of the nitrogen plasma characterisation by optical spectroscopy presented earlier. At low RF power (160 W), the overall smooth increase of the atomic nitrogen line intensity IN in Figure 4.12 (a) (see Section 4.1.6) probably leads to a linear increase of atomic nitrogen supplied for N flux from 3 to 8×10-7

mbar. On the contrary, plasma characterisations at higher RF powers of 210 W and 260 W showed a drop in atomic nitrogen when the nitrogen flux increased from 5 to 8×10-7 mbar. This trend at high RF powers has been reported for bulk GaAsN and GaInAsN MQW grown by RF-MBE [160]. The maximum nitrogen composition achieved at a plasma power of 160 W is inferior to one percent and a higher composition is required for the purpose of this thesis. The comparison of the samples grown with increasing power at ~ 5×10-7 mbar N flux (dashed circle) – at different growth temperatures though – suggests that higher powers should help incorporate more nitrogen. 1 2 3 4 5 6 7 8 9 10 11 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 260 W (_Tg=380oC, As~1.4x10-6mbar) (_Tg=420oC, As~2.5x10-6mbar) 210 W 160 W (T_g=360oC, As~2x10-6mbar) N c o m p o s it io n ( % )

For some nitrogen plasma sources, the RF power controls more effectively the generation of atomic nitrogen than the nitrogen flow rate [161]. To find out whether the nitrogen composition could efficiently be tuned by the variation of RF power at fixed nitrogen flux (all other parameters identical), a set of 200 nm-thick samples were

grown with varying RF power at ~ 5×10-7 mbar N flux (growth rate ~ 1 µm/h).

Figure 5.9 Nitrogen composition as a function of the N plasma power. The dashed line is a guide

to the eye

The dependence of the nitrogen composition on the plasma power is shown in Figure 5.9. While the characterisation of the nitrogen plasma by optical spectroscopy (see Figure 4.13 (a) in Section 4.1.6) showed atomic nitrogen lines with increasing intensity IN when the RF power was increased, the nitrogen incorporation in these samples increases by a factor of three up to 210 W (more nitrogen atoms are produced) but drops at RF power of 260 W. This behaviour contradicts the generally reported trend in GaNAs and GaInAsN [121, 126, 159, 160] for which the nitrogen composition increases and saturates with increasing RF power. In our case, the drop can be attributed to a likely decrease in the brightness of the plasma. Indeed, it was shown for GaAsN that the nitrogen composition is a linear function of parameter

A0076 A0083 A0077 160 180 200 220 240 260 280 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 N c o m p o s it io n ( % ) N plasma power (W) N ~ 5x10-7mbar (T_g=360oC, As~2x10-6mbar)

CSOURCE, a function of the plasma brightness and expressed as h × VOPT / RGaAs, where h is the number of holes in the aperture plate of the plasma, VOPT the plasma light intensity and RGaAs the GaAs growth rate [162]. At fixed growth rate (as is the case for these samples) and identical plate in front of the plasma cell (all other parameters identical), the nitrogen composition is strictly proportional to the brightness of the plasma. Hence for sample A0077 (1.10 % N), a plasma setting still in the bright mode but with a lower brightness could result in an InAsN epilayer with lower nitrogen composition, due to lower efficiency of the cell in cracking N2 molecules.

The effective available RF power ranges from 160 W to 380 W. High RF powers produce a higher amount of ionic N species [163] and can damage the crystal as reported for GaInNAs SQW lasers [164] ; they will be avoided. As demonstrated in this section, the plasma set to an inappropriate N flux might operate with reduced efficiency to crack dimers. Finally, it was shown in the characterisation of the plasma (see Figure 4.12 (a), Section 4.1.6) that there is an appropriate flux for each RF power that maximises the generation of atomic nitrogen. InAsN growth will thus be preferred with the plasma operated at optimal setpoints to assure maximum efficiency with RF powers ranging from 160 W to 260 W. Since the variation of the N flux at fixed RF

power and the variation of the RF power at fixed N flux does not vary regularly the

nitrogen composition, the latter can probably be finely tuned by other growth parameters.