One of the fascinating aspects of this model is that this is a new mechanism for the formation of shallow donor levels, and only initially requires a defect with a deep level state. In effect the compressive neighbouring Oi atoms act to form a localised quantum well that contains the electronic gap state. This quantum well is compressed due to the Coulombic repulsion and dative bonding of the confining atoms, and this increases the energy of the locally confined electronic state, which is trapped in this well.
There are many implications of this work beyond the direct attempt to determine the structure of STDs. This mechanism could provide a mechanism for creating shallow donor levels in a wide range of materials, including commercially important shallow donors in diamond. Ultimately it may even be possible to attempt defect level engineering, treating a material in such a way as to produce defects with energetically squeezed levels lying at precisely the energies required for device design.
In order that the wonderbra mechanism can operate, a core defect has to have only two fundamental properties:
It is possible to imagine various ways in which this could occur, although the 110 type N-O2 structure is possibly one of the simplest. For example, a similar mechanism could operate through a 111 dangling bond state, which could be compressed by a non-bonded substitutional atom sitting on the original neighbour site. A third caveat is that where the electrostatic attraction towards the defect core is only weak, such as in the (CH)- shallow thermal donors, some form of external compression may also have to be applied, such as by the addition of further Oi atoms.
The mechanism provides an alternative method for producing shallow donor states in materials that are traditionally hard to dope, for example, n- type diamond. Instead of finding ways of introducing shallow dopants, the challenge now becomes one of finding ways to compress deep donor states already available in the material in order to shift them upwards in the gap. It is possible to dope diamond with substitutional nitrogen, which in the neutral charge state has a donor level 1.7 eV below the conduction band . This sits in an anti-bonding orbital between N and one of its C neighbours, and hence N does not remain on the Td lattice site but instead drifts away to form a C3v defect. It could be imagined that the anti-bonded C could be replaced with some other defect such as Ge, Si, O or F in order to compress the N donor state through this mechanism. However in practise incorporating such impurities is extremely difficult. In addition any charge redistribution effects will be less marked due to the higher electronegativity of C than Si, and so any application of the wonderbra mechanism to diamond doping will probably be more complicated than this. Another alternative source of this mechanism could come through sp2 hybridised carbon atoms where the filled non-bonded p- orbital could be compressed.
Some preliminary test calculations by Latham and Jones have shown that a NsO2i structure similar to that described above is potentially a single shallow donor in diamond . This could prove useful, as Ns is a common impurity in diamond. The challenge is now thrown open to experimental groups to determine ways in which these defects can be constructed in real materials.