To p rep a re a NEA GaAs photocathode, one m u st first p rep a re a clean GaAs surface. For different GaAs surfaces, th ere are different ways of cleaning them . The [110] surface is th e cleavage plane. A p olarised electron source u sin g a crystal cleaved in u ltra high vacuum to obtain the [110] surface h as been reported by E rb u d ak and Reihl (1978). To clean a [100] surface, as in th e p resen t situation, it is essen tial to chem ically etch th e cry stal ju s t before placing it in vacuum . The cry stal can e ith e r be etched by applying an extensive m ethod proposed by Pierce et al (1980) or by sim ply cleaning th em in am m onia. In th e p re se n t ex p erim en ts, th e la tte r m ethod is applied. It h as been shown th a t th e quality of th e em itter depends more on th e origin of th e crystal th a n on th e cleaning procedure used (Kolac et al 1988). A fter th e etching process, th e cry stal is quickly in stalled into th e source cham ber and th e system evacuated to IO-10 Torr.
re tra c te d position about 2 cm from th e fro n t a p e rtu re of th e ex tractio n electron optics. The cry stal is th e n h e a te d g rad u ally up to 690°C. The h e atin g tem p era tu re used by us is a b it h ig h er th a n th e h eatin g sta n d a rd recom m ended in lite ra tu re s (Kolac et al 1988, Pierce et al 1980). This could be due to th e different conditions of th e rm a l conductance in each cry stal h e a tin g system , as well as d ifferen t te m p e ra tu re m ea su rin g poin ts in different source system s.
A fter th e GaAs cry stal is cooled to less th a t 50°C, it is read y for activation. A typical re su lt of th e activation process is illu stra te d in figure 3.6, w here the p hotocurrent is plotted as a function of tim e. F igure 3.6(a) show s th e r e s u lt from an e x p e rim e n t w ith a low stan d -b y cesiatio n c u rre n t on. The source cham ber p ressu re a t th e s ta r t point is usually less th a n 2 x IO-10 Torr. At first, a c u rre n t of 6 A is passed th ro u g h th e cesium dispenser, and th e photoelectron c u rre n t is m onitored while th e crystal is illu m in a ted by th e GaAlAs la s e r diode a t 0.8 mW. Typically, w ith in 10 m in u tes, a rapidly risin g p h o to cu rren t is observed. The cesiation c u rre n t is th en reduced to ~3.5 A. The em ission c u rre n t gradually increases to a m axim um of 47 pA an d th e n s ta r ts to drop. A t th is point, oxygen is allowed to e n te r th e system th ro u g h a lea k valve, w hile th e cesiation c u rre n t is still k ep t a t a stand-by level of about 2.5A. The long counting tim e s re q u ire d in th e (e, 2e) m e a s u re m e n ts m ean s t h a t it is v ery im p o rta n t to achieve a long lifetim e, sta b le electron beam c u rre n t a t a reaso n ab le c u rre n t level. U n d er th e in flu en ce of low-level continuous cesiation, th e cry stal lifetim e in creases considerably.. This is co n sisten t w ith observations m ade elsew here (T ang et al 1986, Guo et al 1990). The p h o to cu rren t can be k ep t w ith in th e 20 pA - 30 pA ran g e for m ore th a n one w eek, u n til th e G aA s s u rfa c e is too c o n ta m in a te d by th e accu m u latin g cesiation or th e cesium d isp e n se r is em pty. F ig u re 3.6(b) show s th e a ctiv atio n perform ance w ith o u t th e low stan d -b y cesiatio n c u rren t on. It tu rn s out th a t th e ph o to cu rren t drops rapidly in th e absence of any ad d itio n al cesium deposition. The lifetim e of th e p h o to cu rren t is only about 4 hrs.
time (min.)
o 30
time (min.)
F ig u r e 3.6 Typical performance of the Cs-02 activation on GaAs surface (a) with and (b) without the low cesiation stand-by current.