Electron Diffraction at Multiple Slits articles contain extensive references to the more 200 A Silver recent literature. -D.B.and S.H Glass Plate I.INTRODUCTION 10-50A Polymer Strip In the past,several interference and dif- fraction experiments have been carried over from light optics into electron optics:diffraction at a 5000 A Copper half plane,diffraction at holes of various shapes,? diffraction at a straight wire and the production of electron double-beam interference by means of an electron optical biprism,and its subsequent ap- plication to the measurement of internal po- tentials in an electron interferometer,3-5 the realization of a Mach-Zehnder interferometer for FIG.1.Production of the slits electron waves,and the attainment of double beam interference by means of diffraction at two nearby holes.7 to electrons as there are for light.An electron This paper is a report of the transfer of yet beam propagates without scattering only through another interference experiment from light optics a vacuum.Therefore,it is not possible to make to electron optics:the diffraction of electron waves the slits by preparing them on a transparent at a single slit and at periodic arrangements of carrier as is commonly the case in light optics. up to five slits.This experiment yields for the One must find a method which allows the produc- first time three-,four-,and five-slit diffraction tion of slits free of material,with slit dimensions patterns in electron optics. small enough so that they can be illuminated both The design of these multiple-slit diffraction sufficiently intensively and coherently. experiments has encountered some technical difficulties which have until now prevented them II.PRODUCTION OF THE SLITS from being realized.First of all there are the difficulties associated with the small wavelength After some preliminary experiments a method of the electron beam.One must work with was finally found which permitted the produc- moderately fast electrons in order to obtain tion of the desired slits.It is based on the fact reasonable results.In this experiment we have that a layer of polymerized hydrocarbon is pro- used electrons accelerated through 50 kV with a duced at the target of an electron beam if organic corresponding deBroglie wavelength of 0.05 A. molecules are present in the system.This layer This wavelength is significantly smaller than the is characterized by both an exceptional chemical dimensions of an atom,so it is fundamentally im- inactivity and a very low electrical conductivity. possible to make slits in any solid material if These properties are exploited in the following both the width and the separation of these slits manner (Fig.1): are to be of the order of magnitude of one wave- A thin layer of silver is evaporated in a high length.Only slits with considerably greater vacuum on a glass plate,serving as a preliminary dimensions can be produced.In principle this is substrate,so that the plate becomes electrically not a serious constraint for such experiments, conducting.Strips of polymerized hydrocarbon but one must then worry about both the coherent are then printed on the silver layer by aiming an illumination of the slits and the subsequent en- electron beam at the plate.The number,width, largement of the very fine interference patterns. length,and separation of the strips correspond to From the fact that the wavelength is very small the number and dimensions of the desired slits. compared to atomic dimensions,a further dif- The plate is then coated with a copper layer in an ficulty arises:There are no materials transparent electrolytic bath.In this process those portions AJP Volume 48 /5