ULTKASOFT X-KAY MICROSCOPY 



In order to compare the two methods, let 

 us consider the error associated with an ob- 

 ject of one micron dimensions and of compo- 

 sition (organic or of the hghter elements) 

 requiring typically about 25 'A for measur- 

 able contrast. P'or contact microradiography 

 the effective emulsion thickness is a fraction 

 of a micron so that d is of the order of one 

 micron and the error becomes from (3) 

 about 0.05 micron. This is appreciably less 

 than the diffraction error resulting from the 

 light microscope used to measure the micro- 

 radiogram which, for high resolution meas- 

 urement, is about 0.2 n and which is conse- 

 quently the limit of resolution for contact 

 microradiography. To minimize diffraction 

 error in projection microscopy the sample 

 must be placed as close to the x-ray point 

 source as is practical. For this distance 

 equal to 0.1 mm the error is ec^ual to 0.5 ju 

 and for the sample to source distance eciual 

 to 25 microns or 0.001", the error becomes 

 0.25 M- If should be noted that even if the 

 wavelength could be reduced to one fourth 

 (6A) as permitted by heavy element sam- 

 ples, this error is reduced by only one hah. 

 It is therefore concluded that projection mi- 

 croscopy has a resolution limit comparable to 

 that of light microscopy and contact micro- 

 radiography. 



Relative Efficiency of X-Ray Microscope 

 Methods 



At present, the most serious practical limi- 

 tation on x-ra}' microscopy is the very low 

 efficiency of production of the required low- 

 voltage x-radiations. With three or four kilo- 

 volts and maximum beam current, present 

 day projection microscopes yield barely 

 enough intensit}^ for the required fluorescent 

 screen focusing of the point source as gen- 

 erated by magnetic lens demagnification of a 

 point source of electrons. Exposure times are 

 often longer than can be tolerated because 

 of the instability of the focal spot position. 

 To date no projection microscopes have been 

 developed for the ultrasoft wavelengths. 



There is a possibility of successfully meeting 

 the intensity problem by operating the pres- 

 ent projection microscopes at normally high 

 voltages with transmission targets of such 

 thickness and composition that a sufficient 

 amount of the desired ultrasoft radiation is 

 generated along with the shorter wave- 

 lengths. The harder radiations would permit 

 precise adjustment for high resolution work 

 and might then be rejected in the microra- 

 diographic measurement by a recording ma- 

 terial sensitive only to the ultrasoft compo- 

 nent, e.g., the usual, very thin concentrated 

 Lippmann emulsion mounted upon a thin 

 plastic film, the combination being trans- 

 parent to the harder radiation. 



The relative efficiency of the two meth- 

 ods, defined as the reciprocal of the exposure 

 time required per unit sample area, may be 

 estabhshed as follows: 



The camera speed, 8, defined as the recip- 

 rocal of the time rec^uired to record a given 

 sample, is equal to the product of io , the 

 radiant flux per unit solid angle and per unit 

 projected source area; of 12, the solid angle 

 of the x-radiation which is utilized; of the 

 projected source area, which is proportional 

 to the source diameter, w, squared; of the 

 sensitivity of the recording material which is 

 assmned proportional to the square of its 

 resolution error, A^ ; and divided by the area 

 of the irradiated recording material, M-A, 

 where M is the primary magnification and 

 A is the sample area. Thus, 



S = 





{5) 



For an optimum condition of operation, the 

 penmubra error. A, is equated to the record- 

 ing error, A^ . Also, from Fig. 3 it is noted 

 that 



A = wx = Mb = (1 -t- x)d 



(6) 



where x = d/c and 8 is the error. A, as meas- 

 ured at the sample. The maxim lun camera 

 speed may then be written as 



S = Kio5'{il/A){l + 1/x)^. (7) 



679 



