Figure 1. The State of the Art of Nulling Interferometry.

Progress in nulling interferometry is summarized in Figure 1. The plot shows null depth as a function of bandwidth, for laboratory experiments that have been reported since 1999. On the far left-hand side of the plot are shown the results obtained using lasers at visible, near-infrared, and mid-infrared wavelengths. Experiments with bandwidths as large as 40% are shown. Of principal concern to TPF-I are the experiments that have been conducted at mid-infrared wavelengths, which are indicated by the red circles in the plot. The light blue region is the performance required of TPF-I testbeds, being rejection ratios (inverse of the null depth) better than 100,000:1 with a bandwidth of 25% or larger.

The reported results are almost all limited to a single-polarization input. All but one of the results are from two-beam nulling interferometers. The result by Martin et al. (2005) is from the four-beam Planet Detection Testbed. The results by Gappinger (2007) are unpublished results from the Achromatic Nulling Testbed. The result by Peters (2007) is an unpublished result from the Adaptive Nuller. The literature references, where available, are listed below.

From the results to date, we can draw the following conclusions.

1. Laser experiments have shown that achromatic effects (predominantly pathlength variations) can be controlled in the lab at a level that allows rejection ratios better than 1,000,000:1 to be achieved repeatedly. This level of performance exceeds the requirements for TPF-I. Narrow-bandwidth laser nulls have been attained with mid-infrared rejection ratios of 2,000,000:1 and at visible wavelengths of 10,000,000:1.

2. The best broad-band mid-infrared results were obtained in May 2007 with the Adaptive Nuller tesbed, noted in Fig. 1 as "Peters 2007." A rejection ratio of 82,000:1 was obtained with a 32% bandwidth. This rejection ratio is very close to 100,000:1 and therefore almost at the level of the flight requirements. The approach used by the Adaptive Nuller is noteworthy because it is not intrinsically bandwidth limited; a spectrum is imaged onto a line of pixels on a deformable mirror to correct phase and intensity across the band, and this method can easily be extended to work well across the full 6-20 micron science band.

If one takes into account the relaxation of requirements due to advances in our understanding of instability noise, the best broad-band mid-infrared null have nearly demonstrated the flight requirements for TPF-I. Work is ongoing to achieve deeper nulls and to demonstrate similar capability with a four-beam testbed.

Updated 23 May 2007