Supplementary Material
Studying macromolecular solutions without wall effects by
stroboscopic small-angle X-ray scattering
Rita Graceffa, Manfred Burghammer, Richard J. Davies, Cyril Ponchut, and
Christian Riekel
European Synchrotron Radiation Facility (ESRF), B.P.220, F- 38043 Grenoble Cedex, France
Methods
Integration of microdrop generator into data acquisition system. The scheme for the
stroboscopic generation of microdrops using a drop-on demand inkjet system from Microdrop
Technologies GmbH is shown in Figure 1.
Figure 1 Schematic design of the experimental setup featuring a Medipix2 data acquisition
system. Values for tf and tc correspond to typical values used during the experiment
(see main text)
Optical observation of the microdrops is possible after triggering stroboscopically a light
emitting diode (LED).
Experiments were performed at the ESRF ID13 beamline. A 3x3 m2 monochromatic
synchrotron radiation beam at a wavelength of 0.0965 nm is used for SAXS solution
scattering.1 The stroboscopic mode is imposed by triggering the readout of a fast framing
silicon pixel with integrated readout electronics based on the Medipix2 chip (256x256 pixels
of 55 m pitch).2 This detector is advantageous for very weak SAXS signals due to its zero
readout noise and can be better synchronized with the data acquisition system than the
previously used CCD with image intensifier.3 By operating the system in a stroboscopic
acquisition mode, counts are only accumulated on the readout electronics during a defined
exposure window (tf). At the end of the stroboscopic sequence, the counts are readout to a
custom readout board and transferred to a PC. The overall counting time to is user defined.
The minimum exposure time used in the present work was tf=5 sec, which is compatible
with the 1 µs maximum dead time of the Medipix pixels. All microdrops were generated at a
frequency of 4 KHz corresponding to a strobe period of tc=0.250 ms.
Influence of framing time on microdrop shape. A stroboscopically frozen microdrop of 2M
HCl solution is visualized in Figure 2 for a framing time tf=5 sec of after integration of the
refraction/reflection streak using a proprietary batch processing software.4
Figure 2 Influence of detector framing time on observed microdrop shape in mesh scan.
Optical images are obtained with LED illumination for identical stroboscopic
conditions (inset).
The orientation of the streak is indicated as a white line. An increase in framing time increases
the path length along the flight direction over which the interface scattering is recorded.
(Figure 2) This effect is analogous to the blurring observed when imaging a fast moving
object with long exposure times. A framing time of tf=5 sec shows practically no distortion
whilst a framing time of tf=30 sec results in the appearance of an elongated microdrop as
interface scattering extends across several data collection positions.
Speed of microdrops. By observing the position of a sequence of microdrops under
stroboscopic conditions the speed of the microdrops can be determined. (Figure 3)
Figure 3 Microdrop speed derived from the optically determined position of the microdrops
from the capillary exit. A 2nd order polynomial (barred curve) was fitted to the data
(solid squares). The star corresponds to the microdrop position for the SAXS
experiments
The fit of a 2nd order polynomial suggests a speed of about 1.7 m/s at the 1.5 mm distance
from the nozzle exit where the SAXS experiments were performed.
Protein microdrop data collection and data reduction. We determined the center of the
cytochrome C microdrop by horizontal and vertical scans with 4 m increment. The protein
solution scattering was recorded close to the center of the microdrop with 20 s framing time
for an accumulated detector time of 400 ms during 50 cycles in order to maximize the
counting statistics and avoid the scattering from the edge. The background scattering
measured in the same way outside of the microdrop was subtracted. We determined the
scattering from 0.015 M HCl solution microdrops by a mesh-scan with 2.5 m steps, 5 s
framing time and 100 ms accumulated detector time per step. The buffer scattering was
integrated over 203 mesh points avoiding any contributions from interface scattering. The
background scattering, integrated over 406 mesh points outside the microdrop, was
subtracted. All data were scaled to the incoming beam intensity determined by a microionization chamber upstream from the sample position. The scattering volumes of protein
solution can be calculated at the know position of the beam on the microdrop by assuming 80
m diameter microdrops and a 3 m diameter beam. This allows subtracting the scaled HCl
buffer data from the protein microdrop data taking into consideration the different framing
times. This procedure could be further improved by measuring protein solution and buffer
scattering under the same conditions.
References
1
2
3
4
R. Gebhardt, M. Hanfland, M. Mezouar, and C. Riekel, Biomacromolecules 8 (7), 2092 2097 (2007).
X. Llopart, M. Campbell, R. Dinapoli, D. San Segundo, and E. Pernigotti, IEEE
Transactions On Nuclear Science 49, 2279-2283 (2002); C. Ponchut, J. Clément, J. M.
Rigal, E. Papillon, J. Vallerga, D. LaMarra, and B. Mikulec, Nucl. Instrum. Methods A
576, 109-112 (2007).
R. Graceffa, M. Burghammer, R. Davies, and C. Riekel, Rev. Sci. Instrum. 79, 086106086101 (2008).
R. Davies, J. Appl. Cryst. 39, 267-272 (2006).