A single borosilicate-glass capillary was drawn into a 30.5 cm long ellipti
cal shape. The inside diameter was 0.40 mm. at the large base end and 0.13
mm at the tip. With 12 keV X-rays from the CHESS DI bending magnet, the sin
gle-bounce capillary produced a focus of better than 18 mum in diameter (FH
WM) at a 3 cm distance from the capillary tip. A flux gain of 110 in the fo
cus position was observed along with a total flux in the spot of 4 x 10(10)
X-rays/s (conditions: 5.3 GeV. 182 mA, 1.5% bandwidth multilayer, 12 keV X
-rays). A measurement of the far field focus ring diameter yielded a diverg
ence of 3.8 mrad, in good agreement with the 4 mrad design of the optic for
protein crystallography. Using a small 25 mum square beam, we measured the
local reflectivity to be greater than 95% and the inner slope errors of th
e capillary to average about +/- 150 mu rad, both from raw and elliptically
shaped tubing. Our conclusion is that more perfect starting tubing (i.e. o
ne with lower slope errors) is needed to make even more perfect ellipticall
y figured optics in the future. (C) 2001 Elsevier Science B.V. All rights r
eserved.