MICROFOCUSING SOURCE AND MULTILAYER OPTICS BASED X- RAY DIFFRACTION SYSTEMS

Transcription

1 THE RIGAKU JOURNAL VOL. 19 / NO.1 / 2002 MICROFOCUSING SOURCE AND MULTILAYER OPTICS BASED X- RAY DIFFRACTION SYSTEMS BORIS VERMAN, LICAI JIANG AND BONGLEA KIM Osmic, Inc., 1900 Taylor Rd., Auburn Hills, Michigan, 48362, USA 1. Introduction X-ray diffraction is an important method for investigating the material structure. Most of the diffraction methods require an incident beam with well-defined spatial configuration and spectrum purity. The beam size at the sample position should be comparable to the sample size. A small beam spot at detector position is always desired. Some applications, such as powder diffraction, reflectometry, thin film measurements and high resolution diffractometry require well-defined beam in one plane (diffraction plane) and tolerate a much larger beam divergence in the other plane (perpendicular or axial plane). Other applications, such as a protein crystallography, small/macro molecule crystallography or micro-diffraction, require a well-defined beam in the both planes, and commonly use area detectors to collect diffracted X-rays in two dimensions. Multilayer Kirkpatrick-Baez side-by-side optics, so called CMF optics [1], have been the best optical systems in approaching these requirements and have been applied to many applications, especially to protein crystallography. A typical system includes a rotating anode generator, a CMF optic with two elliptical mirrors, a pinhole collimator before the sample of interest to cut unused beam, and an area detector. The focal spot of the X-ray source is positioned at one focus of the optic. X- rays are reflected by both mirrors before reaching the second focus, and diverge after reaching the second focus. Sample position is between the optic and its second focus, typically closer to the focus. An area detector is either at or behind the focus. Although the beam spot is larger when the detector is behind the focus, the longer distance between the sample and the detector within certain range usually yields a smaller angular width of the spot on the detector, i.e. better resolution. We use ray-tracing simulation to optimize an optical system for a particular source and a specific set of requirements from a specific application [2]. A major subject of optimization is to find maximum flux for a given detection capability. For a protein diffraction system, this detection capability is often described as the maximum unit cell size that can be resolved with a certain detector performance, such as the pixel size and the spreading function. This approach has been proven to be a reliable way to predict the performance of a diffraction system. However, the ray tracing method in general reveals the system performance in terms of a series of performance parameters. Ray tracing method is not a very convenient way in understanding the relations among different parts of a system, identifying a key issue of a system, revealing some major tendencies and limitations in a system design. It has also been proven that the simple geometrical method is often a very effective way in the performance comparison among a variety of system configurations, such as with different X-ray sources and different optical configurations. 2. Geometrical Analysis of the System Efficiency A simplified sketch of an X-ray diffraction system with an X-ray optic is shown in Fig. 1. Let us assume that the lens in this system is an ideal X-ray lens similar to a lens for visible light. This means that the lens has 100% efficiency and a larger enough capture angle (numerical aperture) and does not have any aberration. The sizes of the object and image, capture ang1e and convergent angle are bond by the following formula: α D f = = 2 (1) γ F f 1 where α is the capture angle, γ is the convergent angle, D is the diameter of the pre-sample pinhole, 4 The Rigaku Journal

2 F is the area of the focal spot from where X-rays fall into pre-sample pinhole, f 1 is the distance between the source and the lens and f 2 is the distance between lens and the image. Two common limitations exist for the considered X-ray diffraction system. The pre-sample pinhole size is predefined (because the sample is small) and the beam convergence is limited. We will use below γ=3 mrad and D=0.3 mm as an example of the typical (for protein crystallography) requirements for the estimations. If we assume a uniform intensity distribution over the focal spot f of the X-ray tube and a symmetric intensity distribution in two perpendicular planes, we can calculate the flux through the presample pinhole by the following: Φ 0 P F α Φ 0 P D γ Φ = = (2) 2 2 U e f U e f where Φ 0 is quantity of characteristic photons emitted from the X-ray tube target per electron in a unit solid angle, P is X-ray tube power, U is the voltage applied to the X-ray tube and e is the charge of electron. The term P/(U e f 2 ) describes quantity of electrons hit the unit area of the target per second. Several different methods have been developed for the calculation of Φ 0 with the quantitative microanalysis. The value photons per electron is in good agreement with our measurements for Cu Ka radiation at 6 degrees of takeoff angle and 50 kv accelerating voltage. The term Φ 0 x P/(U e f 2 ) is the source brilliance B or the quantity of the photons irradiated from the unit target area within the unit solid angle per second. The right hand side of Eq. (2) was obtained by using Eq. (1). One distinction of the right hand side of Eq. (2) is that the optical parameters are not presented. This is because it is assumed that F < f and the lens is ideal X-ray lens under two limitations of D and γ. Only a few variables remain in Eq. (2) for the flux optimization. Usually, the voltage on X-ray tube is near to optimal for copper or softer characteristic radiations, and below optimal for heavier targets. But it is limited by the X-ray tube and power supply design and there would be no a huge improvement with the higher voltage even for silver characteristic radiations, the hardest in the laboratory diffractometers. The power loading P/f 2 is worthy of more discussion. Two types of X-ray tubes are well known and are widely used: stationary target X-ray tube (usually called sealed tube) and rotating anode X-ray tube. An effectively increased target surface caused by a high-speed anode rotation under electron beam illumination allows a significant increase of power loading for the rotating anode X- ray tube. The power loading is limited by the target overheating in both types of the tubes. The maximum power loading was calculated in reference [3] with the assumption that the maximum target temperature is equal to the target melting temperature. The direct experimental testing of the target melting under an electron beam supports this prediction [3,4]. The maximum power loading for a specific target material depends on the focal spot size and shape. A smaller focal spot allows a higher power Vol. 19 No

3 loading because of the better heat dissipation conditions. The power loading does not grow with the focal spot elongation, but total power increases. An elongated focal spot can be seen as a symmetrical spot at a shallow takeoff angle with an increased brilliance. This feature is used in the most of X-ray system with the typical takeoff angle of 6 degrees. We will consider this takeoff angle and the focal spot aspect ratio 10 to 1 (the length to the width) in the further discussions. The brilliance, calculated on the base of permissible power load, is presented in Fig. 2 (solid lines) as the functions of the effective focal spot size for both kinds of X-ray tubes. The permissible power loading (and source brilliance correspondingly) increases about 30 times for the rotating anode X-ray tubes and about 500 times for the sealed X-ray tubes in the considered focal spot width range extending from 1 micrometer to 1 millimeter. The brilliance from rotating anode X- ray tubes can be several times higher compared to the sealed X-ray tubes (especially with a higher speed of the target rotation) even with a small focal spot. The maximum power loading and the respectively brilliance of a real X-ray tubes are some what lower than the theoretical predictions. A few examples, based on manufacturers specifications are shown in Fig. 2 as single dots for both types of X-ray tubes. These examples are limited to the X- ray tubes with the focal spot with the aspect ratio of 10: 1. More examples are given in Table 1 with variety of the aspect ratios. Most of the X-ray tubes are loaded at 40 to 70 percents of the calculated power. This suggests that another limitation for the maximum power loading may exist. We investigated the microfocusing X-ray tubes presented at line 7 and line 8 in Table 1 and observed an intensity decrease with time with the rate of about 0.02% per hour at the conditions shown in the table. A ten percent increase of power loading will accelerate this degradation rate dramatically. Also, it is well known that the target surface of the rotating anode X-ray tube is required to be re-polished at regular time intervals and the electron beam track is visible before repolishing. Both facts evidence that something is changing on the target with the working time. Some types of chemical reactions are possible between the target material and residual gases in the tube at high temperature and high level of ionization, which are typical target working conditions. The investigation of the physical and chemical properties of the electron beam track on the target surfaces may shed light on about the mechanism of the target degradation and, possi- # Power e-beam incident angle Table 1. X-ray tubes power loading and brilliance. Focal spot Takeoff Target Power loading Length Width angle speed real Calculated relative Brilliance kw degrees mm mm degress M/sec kw/mm 2 kw/mm 2 kw/mm 2 photons sec mm 2 mrad 2 Rotating anode x-ray tubes E E E E E E E+10 Stationary target x-ray tubes E E E E E E E+09 6 The Rigaku Journal

4 Vol. 19 No

5 bly, the ways of increasing target lifetime. The best that we can expect is 30% to 50% gain in the X-ray tube brilliance. Both Fig. 2 and Table 1 suggest that a higher brilliance may be obtained with a smaller focal spot size. But there are some limitations in this attempt. It is known that electrons scatter inside the target. For an electron beam with a cross section of sub-micrometers, the effective area of the characteristic radiation generation may exceed the cross section of the electron beam itself. An estimation based on the model in reference [5] shows that the dimension of the excited X-ray generation volume by an electron at 50 kev is about 6.8 µm for CuK α radiation. The calculated volume includes some peripheral area, which provides an insignificant contribution to the X-ray intensity. We can expect much smaller FWHM of the generation area. An estimation of the effective depth of the X-ray generation on the base of Philibert's correction [6] gives the value of 1.5 µm. A lower electron energy results a smaller area of excitation. At the same time, it causes a lower efficiency of characteristic line production. This means that for a conventional X-ray beam production, when an electron beam hits a solid copper target, the maximum obtainable brilliance is at the level of 5x photons/(mm 2 x mrad 2 x s) for a stationary target. This can be several times higher for a rotating target. These numbers are about one order of magnitude higher than what is available now. It is assumed above that the focal spot size f is larger than the area of the target F from which photons pass the lens, then go through the pinhole D. Because the brilliance is higher for a smaller focal spot, we can let f = F for maximum flux. With this condition the expression for the flux is simplified greatly: F=B x γ 2 x D 2 (3) It is useful to discuss two features of multilayer optics. First, the multilayer is a wavelengthselective reflector, by utilizing Bragg's law. This important feature allows one to reduce continuous radiation and K β line from the original X-ray spectrum. But this feature leads to a limited range of the incident angles for a characteristic line reflection. A quantitative description of the multilayer angular acceptance is the width of its rocking curve. The rocking curve widths δθ of commonly used multilayers normally range from 0.01 degrees to 0.1 degrees for copper characteristic radiation. Equation (3) is valid if the rocking curve width is larger then the angular size of the pinhole φ: δθ > φ (as shown in Fig. 2). This condition must be checked before Eq. (3) is used. Second, it is convenient to discuss relative performances of focusing (elliptical) and collimating (parabolic) optics by using Fig. 1 and Eq. (3). In the case of the parabolic optic and parallel beam, the capture angle α=d/f 1 and the working area of the target F= φ x f 1 = D x f 1 /f 2. For the flux we have: Φ = B x D 4 /(f 2 ) 2 = B x D 2 x φ 2 (4) It can be seen from Eqs. (3) and (4) that the flux provided by a parabolic optic can not exceed the flux from an elliptical optic because γ = φ. The maximum value of φ is δθ. It means that, if the acceptable beam convergence is not very low, focusing optic may provide higher or significantly higher flux. 3. Calculation of Multilayer Optics Parameters The simple geometrical approach of the system description is helpful to conduct a preliminary design of a multilayer optic. Let us use the abovementioned example of the requirements to the system performances. In addition, let us assume that the distance from the optic to its second focus is equal to 500 mm and the d-spacing at the center of the optics is 30 Å. We can calculate the major optic parameters on the base of Eq. (1) for the various sizes of the X-ray tube focal spots, as shown in the top portion of Table 2. Several other parameters are also shown in Table 2: the mirror lengths, the d-spacing ranges, the ranges of the radius along the mirror surface. These parameters are calculated for the mirrors with elliptical surfaces and the center positions of the mirrors as defined above. Calculated mirror lengths provide the specified convergence angle. The capture angles may differ from the geometrical calculations because mirror surface with a variable curvature is extended along the beam path. The data in the table suggest several obvious tendencies. First, as the focal spot becomes smaller, the optic must be positioned closer to the focal spot. However, a larger d-spacing change and a larger radius change are needed. These 8 The Rigaku Journal

6 Table 2. Calculated optics parameters. Different configurations Geometrical calculations Parameter Expression Units F mm M D/F f 1 f 2 /M mm α γ*m mrad degrees L mirror mm d min Angstroms d max Angstroms R min mm R max mm Efficiency 6.45E E E E E-04 Flux photons/sec 8.22E E E E E+09 Ray-tracing simulations f 1 mm α degrees γ mrad L mirror mm d min Angstroms d max Angstroms R min mm R max mm Efficiency 1.19E-05 3/88E E-05 Flux photons/sec 4/58E E E+08 geometrical calculations ignore the real performances of multilayers and various technological limitations. Several specific features of multilayer optic need to be considered for the optimization of the optics performance and the system performance. The reflectivity of all the commonly used coatings drops quickly when d-spacing decreases below 30 Å [5] and is below 50% for a d-spacing of 15 Å. The width of the rocking curve (and angular acceptance accordingly) decreases nearly linearly as d-spacing decreases. Other technological complications appear for the optic for a small focal spot. The surface radius of an optic changes more rapidly and the ratio between the smallest and the largest radius on the mirror surface increases from 1.4 for the version 1 to 30 for the version 4. This makes the traditional method of surface formation not applicable to the versions 3 and 4. The gradient of d-spacing distribution grows quickly as well. The d-spacing distributions and their gradients are shown in Fig. 3 for all four versions of the Optics. The gradient for Optics 3 is slightly and for Optics 4 is significantly beyond of the current technological capabilities. One more specific feature of the versions 3 and 4 is that the distance from the focal spot to the closer end of the mirror is very short: 3.7 mm and 1 mm respectively. This not only requires a specific X-ray tube design, but also produces reflected beam with an ununiform beam intensity distribution as well. For the Optics version 3, for example, the distances from the focal spot to the closer and further ends of the mirror vary 8 times. This means that the beam intensity that strikes the different area of the mirror changes even more dramatically. This is the primary cause of ununiform intensity distribution of the reflected X-rays. By nature, a multilayer coating has a lower reflectivity and narrow rocking curve at lower d- spacing. Since the mirror surface near the source requires a larger incident angle (therefore a smaller d-spacing), reflected beam uniformity can be improved, but only partially. Finally, version 4 does not improve the system performances compared to the version 3, because of a low reflectivity and a narrow acceptance angle at the end of the optic closer to the source, and the ununiformity of intensity in the beam cross section. Vol. 19 No

7 The simple geometrical analysis above shows explicitly the problems arising from the design of an optic with a high brilliance X-ray source. This approach cannot suggest optimal solutions, but it helps to find a space of variables, from which an optimal solution could be found. The data in Table 2 show that an optimal system can be built on the base of the microfocusing source with effective focal spot in the range between 10 and 30 µm. The system optimization was done on the base of ray tracing simulations. 4. Optics Optimization by Ray-tracing Simulations We considered beam convergence γ, the pinhole diameter D and the position (the distance between optic and pinhole f 2 ) as given parameters, and the focal spot to optic distance f 1, optic d-spacing and length as variables. One constraint applied was that the reflected beam ununiformity must not exceed 50%. Results of such optimization for the focal spots of 0.03 mm and 0.01 are given in the bottom part of Table 2 as the version 2 and 3. Optimal optic parameters are similar to the geometrical calculation for the version 2. In contrast, optimal optic parameters differ greatly for the version 3. The major reasons for the difference were discussed above (low reflectivity and small acceptance angle at extremely low d- spacing, ununiformity of the mirror surface irradiation). Both versions of the system produce roughly the same flux, which supports the previous assumption that maximal flux may be obtained for the system with the focal spot in the range between 10 and 30 µm. Next step of the optimization had been performed with the focal spot size as one more variable. The results of this optimization are shown as version 5 in Table Examples of Applications 5.1. System for Protein Crystallography The beam requirements considered above as an example are typical for protein crystallography. The acceptable beam convergence angle γ ranges from 2 mrad for the system intended for large unit cells measurements to 5 mrad for the faster measurements of the structures with small and moderate unit cells. The above analysis had shown that for 3 mrad beam convergence, the effective focal spot size of 20 µm is close to optimum, and it was chosen as an initial value for the system design. The microfocusing sources from Bede Scientific Instruments provide required size of the focal spot. Major parameters of this type of source were presented at line 12 in Table 1. The system design differs from what was calculated as optimal optics design for this focal spot size. Several constraints and practical limitations were taken into account. First, the distance between the spot and the closer end of the optics cannot be less than 8 mm because of the source design. Additional requirements, such as room for gas sealed Optics housing with an X-ray shielding connection and a mechanism for the Optics alignment, extend this distance to about 20 mm. Second, optic design accommodates existing technological limitations: minimal optic radius is 1 m, minimal d-spacing is 15 Å and maximum gradient of d-spacing is 1 Å per millimeter. Optic ray-tracing optimization with these limitations gave the following optics parameters: the focal spot to optic distance f 1 =65 mm, the optic length I m =80 mm, d-spacing at the optic center d=35 Å, the distance from optic to pinhole f 2 =635 mm. System provides 1.5 x 10 8 photons per second at the sample position with the beam convergence of 2.9 mrad System for Small Molecule Crystallography We used convergence γ=3.7 mrad and presample pinhole diameter D=0.1 mm as the requirements to the beam for this application. This D is not a typical size of the crystal. Instead, it represents a common size of the small crystals. But we chose this value because precise focusing is possible with a microfocusing system and small crystals become more common as subjects of studies. Another state-of-the-art application is to use Mo characteristic radiation with λ=0.71 Å. This shorter wavelength is not favorable to the multilayer optic because the width of the rocking curve drops proportionally to the wavelength. The calculation and ray-tracing simulation similar to those described above for the protein crystallography system gave the optimal system parameters: focal spot effective size F=0.017 mm, the focal spot to the optic distance f 1 =57 mm, optics length I m =80 mm and central optic d-spacing d=34 Å for the optic to the presample pinhole distance f 2 =243 mm. The system design based on the available mi- 10 The Rigaku Journal

8 Table 3. System performances comparison for small molecule crystallography. System parameters Existing system Sealed tube with CMF Optics Microfocusing source with CMF Optics Sealed x-ray tube Sealed x-ray tube Microfocusing source Source 0.4 x 0.8 mm focal spot 0.4 x 0.4 mm focal spot x mm 6 degree takeoff angle 3 degree takeoff angle focal spot 2 kw power 2 kw power 30 W power Optics Graphite monochromator CMF CMF Flux through 0.1 mm sample (10 6 photons/s) Beam divergence for a 0.1 mm pinhole (mrad) 7.3 x Flux through 0.2 mm sample (10 6 photons/s) Beam divergence for a 0.2 mm pinhole (mrad) 5.1 x Flux through 0.3 mm sample (10 6 photons/s) Beam divergence for a 0.3 mm pinhole (mrad) 3.5 x crofocusing source with molybdenum target and accommodated the above-mentioned requirements to the source-optic coupling. Finally the system parameters are as the following: F=0.017 mm, f 1 =65 mm, I m =80 mm, central optic d-spacing d=35 Å for the optic to the pre-sample pinhole distance f 2 =235 mm. The experimental comparisons of the beam performances of several systems for small molecule diffractometry are given in Table 3. These results show that a system based on microfocusing source provides more flux for small samples compared to the systems based on the traditional sealed X-ray tubes. The first and the last system in Table 3 were compared with the collected data from the same crystal of cytidine with the dimensions of 0.05 mm x 0.05 mm x mm. The data quality obtained with microfocusing system is superior compared to the traditional system with 6 times longer of exposure time. One specific feature of the systems utilizing a convergence beam for the single crystal diffraction is their capability to provide adjustable trade between the flux and the angular size of the pattern (and maximal accessible unit cell correspondingly) [8]. This feature is important for both protein and small molecule crystallography system and is elucidated in Figs. 4 and 5. The best position for a small sample to obtain a maximum usable flux is at the narrowest beam cross section around optic focal plane. A larger sample may be placed closer to the optic (Fig. 5) to illuminate entire volume. The detector can be moved toward to the optic as well (keeping the same sample to detector distance), and closer to the optic focal plane. Smaller pattern sizes at the detector position provide a better angular resolution and a higher peak to background ratio. This is an important advantage of the systems with focusing optics. We incorporated an appropriate adjustability in the systems design. The position of microfocusing Vol. 19 No

9 source-optics unit may be easily adjusted relative to the sample and detector Small Angle Scattering System for Biomedical Applications A tiny and low divergence beam is extremely important for small angle scattering experiments. Two approaches are used commonly to shape the beam: systems with two and three pinholes, depends on the sample scattering power and angular scattering range. A two-pinhole camera provides higher flux, but background is higher due to parasitic scattering on the optics and the first and the second slits, with a significant contribution of the latter. Microfocusing source provides additional advantages in reducing this component of the background. This is owing to the better-defined shape of the beam. When the angular size of the source coincides with the angular acceptance of optic, intensity in the beam cross section drops quickly on the beam wings", Two-dimensional focusing multilayer optics coupled with microfocusing source provides converging beam with well defined cross section. A pinhole with a diameter that exceeds the beam cross section is hit with a low intensity beam and provides a lower scattering. The requirements to the beam for small angle scattering experiment cannot be defined as clearly as for protein crystallography. Because of this, the pinholes were included in the ray-tracing simulations. The optimization procedure was to search a reasonable trade between system efficiency and its resolution or its minimal accessible angle (for three pinhole system). The same microfocusing source (line 12 in Table 1) was used for this system. Optimal optic parameters are found as the following: the focal spot to optic distance f 1 =65 mm, the optic length I m =80 mm, d-spacing at the optics center d=45 Å, the multilayer coating type is Ni/C as it provides a higher reflectivity and narrower rocking curve for a better beam shaping. To assess the level of microfocusing system performances, the characteristics of several existing systems was evaluated by ray-tracing simulations. Some results, for a comparison of two pinhole systems, are given in Table 4. We consider a smaller beam width at the detector as an evidence of the system capability for providing improved performance for the low angle scattering measurements. One of the subjects of the microfocusing system testing is to prove that a better shaped beam scatters less on the second pinhole, and therefore provides lower background. We expect a more complicated alignment procedure of the system with a tiny shaped beam. To overcome this problem the system design incorporates a set of features for much easier alignment Micro-diffraction System We consider the beam convergence of γ=0.05 degrees=0.87 mrad and beam cross-section at the sample D=0.02 mm, as a typical set of requirements for the microdiffraction application. Considering above discussed physical limitations, an ideal X-ray tube with a stationary target coupled with an ideal X-ray optic may provide a flux of 1.5 x 10 7 photons per second with these beam condition according to Eq. (3). Taking into account a typical power underloading of the real X- ray tubes (coefficient 0.6), we came with the flux 9x10 6 photons per second. Even this optimistic prediction seems disappointing for many crystallographers, because this is lower than the expected flux that causes a long time of the sample studying. Table 4. Small angle scattering system performances. Parameter System Source type Rotating anode Rotating anode Microfocusing Effective focal spot size, mm X-ray tube power, kwatts Optics Parabolic Kirkpatrick-Baez CMF CMF Flux at the sample, photons/sec 1.7 x x x 10 8 Beam cross section at the sample, FWHM, mm Beam cross section at the detector, Level 0.1 of maximum, mm The Rigaku Journal

10 We believe, nevertheless, that a reasonable set of the requirements to the beam for this application may be found. It is encouraging that both geometrical calculation and ray tracing simulations predict a higher system efficiency (closer to the mentioned limit) than for other applications and without extreme requirements to the multilayer. Two versions of the system may be considered. The first provides beam with FWHM of 0.02 mm at the sample. The total flux is, in this case, 4.2 x 10 6 photons/sec. and the flux within the FWHM area is 3x10 6 photons/sec. The systems parameters of this version are: effective focal spot size F=2 micrometers, 6 degrees of takeoff angle, X-ray tube power of 3.5 Watts, the focal spot to the optic distance f 1 = 50 mm, mirror length I m = 21.8 mm, central d-spacing d= 42.5 Å, optic to the sample distance f 2 =450 mm. Another version assumes a pinhole before the sample. The flux trough the pinhole is slightly higher (F=4x10 6 ) and its intensity is more uniform. The system parameters are different from the previous version and are as the following: f 1 = 50 mm, I m =21.0 mm, d=41.6 Å. The optic parameters in both versions are in the range of a conventional CMF optic: the d- spacing range is 31 to 52 Å, the gradient of d- spacing does not exceed 0.8 Å per mm, the minimum radius is larger than 1.5 meters. Only one requirement is very stringent for this optics: the precision of the optic surface must be extremely high, at the level of 10 µrad in order to avoid the beam widening at the focal plane. This requirement is about one order of magnitude higher than the requirement imposed on conventional CMF optic, and the existing CMF technology can not achieve this. Few approaches to overcome this problem are under our considerations and testing. We believe that CMF optic is the most promising candidate as optic for micro-diffractometry because other kinds of optics have even severe limitations: multireflection single capillary or policapillary have a high beam divergence (above the glass critical total reflection angle); total reflection optics has a lower angular acceptance than multilayers; zone plate and refractive optics have limitations of short focusing distance, required for a high solid acceptance angle. 6. Conclusion Diffraction systems on the base of a microfocusing source and CMF Optics had been proven their efficiency and advanced features for protein and small molecule crystallography application. Small angle scattering system is under testing. Design study of the system for microdiffraction needs a refinement from the initial requirements. Microfocusing source design and multilayer optic technology must be upgraded for a higher system efficiency. References [1] B. Verman, L. Jiang, B. Kim, R. Smith, N. Grupido, Adv. X-ray Anal., 42, 321, (1998). [2] "Performance evaluation for multilayer optics using ray tracing method", B. Verman, K. D. Joenson, L. Jiang, XVIlith IUCr Congress & General Assembly, Glasgow, Scotland, Abstracts, p. 524, August 4-13, (1999). [3] A. Muller, Prog. Roy. Soc., A, CXXXII, 646, (1931). [4] D. E. Grider, A. Wright, P. K. Ausburn, J. Phys. D. Appl. Phys., 19, 2281, (1986). [5] D. E.Newbury, Microscopy, 22, 1,11, (1992). [6] J. Philibert, in "X-ray Optics and X-ray Microanalysis", N.Y., Acad. Press, 1963, [7] L. Jiang, B. Verman, B. Kim, Y. Platonov, Z. almosheky, R. Smith, N. Grupido, The Rigaku Journal, 18, 2, 13, (2001). [8] B. Verman, L. Jiang, "X-ray diffractometer with adjustable image distance", USA patent 6,069,934. Vol. 19 No