1.1 Objectives and originality of the joint research activity

1.1.1 Objectives
Instruments of high-intensity neutron sources, dedicated to single crystal diffractometry (SXD), reflectometry, and small angle neutron scattering (SANS), are limited nowadays by the position resolution of their detector and/or its maximum sustained counting rate. This situation results from continuous progress made in neutron optics, allowing increased neutron flux on samples, and from an increasing demand to use samples of less than 1mm3 volume. For example, on the SANS machines of the ILL, it is sometimes necessary to prevent detector saturation by attenuating the beam upstream the sample. This situation has triggered the SANS-2MHz project, which goal is to improve the actual counting rate of the D22 instrument by a factor of 50. As a result, a new detector that has been studied in the framework of the TECHNI-FP5 network is planned to start its operation on D22 at the beginning of 2004. It is based on position sensitive proportional counters, a technology well suited for very high counting applications, but limited in position resolution to about 5 mm. In order to provide both a high count rate and a high position resolution for SXD and reflectometry instruments, intense development have been engaged in several institutes, which have gained a strong interest among the future users of the two spallation sources under construction (SNS in the US, JSNS in Japan); actually, the peak count rate of these sources will be more than one order of magnitude higher than ISIS, the actual most intense pulsed source, and progress made in the field of neutron detectors during the last years do not allow to predict an optimum use of SNS and JSNS at the beginning of their operation.

The project will be devoted to the development of a new detector that will significantly improve the performances of SXD and reflectometry neutron instruments, in terms of sensitive area, counting rate, and position resolution. In order to broadly benefit to the neutron user community, techniques that can be transfer to the industry will be given priority. In Table1 are listed the specifications of the MILAND detector.

1.1.2 State-of-the-art
Important parameters concerning current detectors used in some “world class” reflectometers and diffractometers are given in Table2. RR, defined as the position resolution in percentage of the detector dimension, provides an arbitrary but useful parameter to compare detectors of different sizes. To this respect, most of the detectors indicated above are equivalent. Surprisingly, RR is also equal to 0.8% for the actually most performing SANS detector, namely the 1 m x 1 m MultiWire Proportional Chamber of the D22 instrument at the ILL, and for the 5 cm x 5 cm very high resolution detector developed at the BNL. Moreover, the high performance charge division position sensitive counter developed recently by Reuter Stokes, has also an RR value equal to 0.8% (8 mm position resolution over 1 m length). RR=0.8% can then be considered as to the actual limit of resolution for gas detectors, independently of their size. Several factors can be pointed out to explain this limit:

⇒ The mechanical stress induced by the inner gas pressure on the entrance window increases linearly with its dimension. For large area detectors, in order to maintain this stress at an acceptable level without affecting the neutron detection efficiency, it is necessary to reduce the differential pressure between the two sides of the window. This can be done by increasing the conversion gap and reducing the inner pressure gas, while keeping unchanged the CF4 partial pressure (a pressure of 4 bars of CF4 is necessary to achieve 1 mm position resolution) and the 3He volume. But the limit imposed by parallax error is rapidly reached and complementary techniques have been used: spherical window or field electrodes have be designed to correct the field shape inside the conversion gap and limit the parallax. Another solution consists of using a double volume vessel, with a first volume limited by a spherical entrance window and containing a transparent gas. All these techniques were successfully applied, but they are rather complicated and were never investigated to achieve a resolution as good as 1 mm
⇒ Increasing the density of channels also requires a higher precision of the internal mechanical components, and induces new difficulties to implement electronics. It has also a strong impact on the detector cost.

The specifications of the MILAND detector were defined according to previous international detector meetings. Compared to the best-performing 2D neutron gas detectors available today, it will improve the RR by a factor of 3; global as well as local counting rate capability will be increased by a factor of 5; and all the other parameters (gamma discrimination, counting stability and uniformity, efficiency) will be maintained at the same performance level.

1.2 Implementation plan of the joint research activity
Two detector options based on ³He as the gas converter will be studied in the first 2 years of the MILAND project. Both of them rely on the same process of neutron interaction in 3He: a primary charge cloud is generated in the gas after a nuclear interaction took part, and a layout of polarised electrodes mounted in the gas volume amplify this charge signal during a gas avalanche process. One option is the Multi Wire Proportional Chamber (MWPC) where each neutron is counted individually. This charge readout gas detector is known since 30 years in neutron instrumentation. Thanks to the use of photolithographic techniques, the project will move it back beyond its actual limit of resolution. During the avalanche process, charges are amplified by collision between electrons and molecules of the gas. Photons are also emitted by excitation of these molecules in a range of wavelength which extends from UV to NIR, and depends on the gas components. The second detector studied in MILAND is a combination of a Gas Scintillator Proportional Chamber (GSPC) coupled with a CCD detector where the avalanche light is integrated on the CCD during the acquisition time. This light readout gas detector, introduced 20 years ago in high energy physics, has never been used on a neutron instrument.