In Figure 9.1 one of the loss mechanisms is from radiative recombination (loss 4). In most devices this is assumed to be a minimum loss that cannot be reduced – for a cell at the radiative limit, i. e. no nonradiative recombination. This is necessary as a reciprocal device that can absorb solar wavelengths must also be able to emit those same wavelengths [Kirchoff, 1860]. However, as mentioned in Section 2.2.2 in this volume, it is possible that a nonreciprocal device could reuse some of this emitted radiation and boost efficiencies beyond the radiative limit. This is possible in theory because Kirchoff’s law applies only to time-symmetric processes. Time symmetry can be violated in a detailed process if the deeper CPT symmetry is observed. Thus, either charge or parity must also be asymmetric for a process to be time asymmetric [Green, 2003; Weinberg, 2013]. A device that can be nonreciprocal in this time asymmetric way is known as a multiport circulator; with incoming light incident on port 1 emitted at port 2 but incident light on port 2 emitted at port 3, [Ries, 1983]. In principle, such a circulator can be used to redirect light emitted from a solar cell onto a second cell, light from this cell can then go through a second circulator to a third cell, etc. It has been shown that such an approach can boost efficiencies to 93%, the Landsberg limit, for an infinite number of circulators illuminating tandem cells with an infinite number of bandgaps [Brown and Green, 2002; Green, 2003]. The obvious practical difficulties of this are offset to some extent by the fact that most of the efficiency gain is obtained with the first circulator and, as we have seen, a tandem cell has most of its increase in efficiency for the first few layers. The nonreciprocal rotation of the polarisation of light by a magnetic field can be and is used to fabricate such esoteric-sounding circulator devices, for microwave and laser optics [Fujii, 1991]. However, the complexity of the components and the very small efficiency gains make such an approach applied to photovoltaics only appropriate for theoretical consideration.