Potential of carbon nanotubes films for infrared bolometers

ABSTRACT We investigate in this paper the potential of carbon nanotubes for infrared bolometers. A method to obtain CNT film layer and technological processes to obtain matrix of devices are presented. The electrical characterization of samples establishes the quality of our technology i.e. low contact resistance, and weak dispersion between devices. The potential of carbon nanotub es films as bolometric material is investigated by measuring the thermal dependence of their resistance and by comparison with amorphous silicon (one of the leading material for bolometric applications). Optical measurements of CNT films in the infrared and THz ranges show a relatively high absorption for a few hundreds nanometers thick material. Eventually the infrared (8-12  m) photo-response of a first demonstrator is presented and discussed. Keywords : uncooled IR detector, microbolometer, single wall carbon nanotube. INTRODUCTION Over the last decade, the market of infrared bolometer’s focal plane array (FPA) has shown a rapid growth. The vanadium oxides originally developed by Honeywell, and the amorphous silicon pushed forward by ULIS and CEA-LETI are the main technologies [1]. Since the first generation of FPA, various improvements have been made. In the case of ULIS, they gradually decrease the pitch of their pixels from 50 P m in 1998 [2], to 17 P m in 2009 [3]. This enables the realization of large format FPA: 1024*768 pixels and decrease of manufacturing costs. In order to compensate the inherent decrease of the Noise Equivalent Temperature Detectivity (NETD) due to the reduction of the pitch, improvements were made on the thermal insulation of the suspended membranes by means of advanced UV-photolithography processes. Attention was also paid to the reduction of the noise by an improvement of the fabrication process of the amorphous silicon and by advanced design of pixels. Nevertheless this roadmap will be in the next future less and less relevant, especially because further reduction of the pitch will reach the diffraction limit for the LWIR (8-14 P m). Absorption and fill factor of the FPA are already closed to their ideal values. Improvement of performance or reduction of the cost can still be made by progress in smart and efficient read-out integrated circuit ROIC, which is a key point of the uncooled FPA success. Recently progress has been reported in the packaging, by pixel level encapsulation [4]. Improvement of the performances can also be made by a change of the bolometric materials [5]. It is worth noticing that the two same materials have been driving the market for tw enty years. Recently, following th e results shown by Itkis et al. [6] on the bolometric behavior of carbon nanotube (C NT) films, ONERA, the French aerospace labs and LPN-CNRS founded by the French MOD, have chosen to investigate the potential of carbon nanotubes films as bolometers. First we will present the fabrication and report processes of our material: Single Wall Carbon Nanotubes (SWCNT) films. Then we will focus on the fabrication processes developed in our clean room facilities, and the electrical characterization of the fabricated devices. The next part will deal with the potential of SWCNT films as a thermistor. Then we will detail the optical characterization of our material, and eventually show the infrared (8-12  m) photo-response of our first generation of bolometric sensors.