Non-Radiative Competition in the Excitation of Erbium Implanted Silicon Light Emitting Devices

  • PDF / 1,225,196 Bytes
  • 6 Pages / 414.72 x 648 pts Page_size
  • 116 Downloads / 179 Views

DOWNLOAD

REPORT


ABSTRACT This paper reports a study of the non-radiative processes competing with the excitation of the erbium ion in layers implanted with high concentrations of erbium and oxygen. These processes reduce the luminescence efficiency of the Si:Er system and dramatically increase the threshold current density calculated to be necessary for an ultimate goal, the Si/Ge:Er LASER. Using cross sectional TEM, photoluminescence as a function of temperature and DLTS, it is demonstrated that a two stage anneal procedure which avoids the formation of extended defects and removes specific deep states is necessary to obtain efficient Er 3 + excitation at high erbium concentrations. Comparisons are made with damage resulting from germanium implantation into silicon. The role of multiple stage anneals is discussed in relation to the removal of Shockley-Hall-Read recombination centres

INTRODUCTION Light emission from erbium in silicon [1] is currently receiving substantial world attention because of the possibility of making an efficient LED or LASER as part of a silicon integrated circuit. The dominant emission is from a transition within the erbium atom at a wavelength of 1.541Lm. It has a narrow line width and excellent wavelength/temperature stability afforded by the screened 4113/2 to 4115/2 transition within the Er 3+ ion. There are, however, many intractable problems to conquer before there is any real chance of achieving viable devices. Central to this is the efficiency of the light emission from the Si:Er system. Although several reports of room temperature emission at 1.541Lm have been published [2,3], the estimates of efficiency at these temperatures are very low. A typical value for the internal quantum efficiency is of the order of 10- 3 %. At low temperatures (77K) the efficiency is much higher but still compares very unfavourably with available III-V communication lasers which operate with a room temperature internal quantum efficiency approaching 100%. There are several separate energy loss mechanisms responsible for this problem which can be divided into two distinct classes. Firstly, processes which compete directly with the 1.54jim emission causing de-excitation of the erbium ion and secondly processes which compete with the excitation of the erbium , whether from photo or electrical excitation. The first process is associated with the interaction between the erbium ion and the relatively small bandgap host material, ie, silicon; this results in a rapid quenching of the luminescence with increasing temperature. In wide bandgap materials, such as glasses, this process does not 223 Mat. Res. Soc. Symp. Proc. Vol. 392 01995 Materials Research Society

occur but in hosts of smaller bandgap, the quenching becomes significant and moves to lower temperatures as the host bandgap becomes smaller [4]. This quenching is one of the critical problems and the attention of several groups (including our own) is focused on it. However, there are other very important loss mechanisms which degrade the excitation of the Si:Er, som