nature physics portal - looking back - Waves of mat...

http://www.nature.com/physics/looking-back/davis...

Wa ves o f m a tter II s ca ttering o f electro ns

In 1927, Davis s on and Germer bombarded a s ingle crys tal of nickel with a beam of electrons , and obs erved s everal beams of s cattered electrons that were almos t as well defined as the incident beam. The phenomenological s imilarities with X-ray diffraction were s triking, and s howed that a wavelength could indeed be as s ociated with the electrons . Nature 119 , 558 -560 (1927) | click here for a PDF vers ion (400 K) |
Na ture P ublis hing Group 20 0 6

T he Scat t e ring of Ele ct rons by a Single Cryst al of Nicke l.

In a s eries of experiments now in progres s , we are directing a narrow beam of electrons normally agains t a target cut from a s ingle crys tal of nickel, and are meas uring the intens ity of s cattering (number of electrons per unit s olid angle with s peeds near that of the bombarding electrons ) in various directions in front of the target. The experimental arrangement is s uch that the intens ity of s cattering can be meas ured in any latitude from the equator (plane of the target) to within 20 of the pole (incident beam) and in any az imuth. Fig u re 1 Intens ity of electron s cattering vs . co-latitude angle for various bombarding voltages az imuth-{111}-330. high-res olution vers ion

The face of the target is cut parallel to a s et of {111}-planes of the crys tal lattice, and etching by vaporis ation has been employed to develop its s urface into {111}-facets . The bombardment covers an area of about 2 mm.2 and is normal to thes e facets . As viewed along the incident beam the arrangement of atoms in the crys tal exhibits a threefold s ymmetry. Three {100}-normals equally s paced in az imuth emerge from the crys tal in latitude 35, and, midway in az imuth between thes e, three {111}-normals emerge in latitude 20. It will be convenient to refer to the az imuth of any one of the {100}-normals as a {100}-az imuth, and to that of any one of the {111}-normals as a {111}-az imuth. A third s et of az imuths mus t als o be s pecified; this bis ects the dihedral angle between adjacent {100}- and {111}-az imuths and includes a {110}-normal lying in the plane of the target. There are s ix s uch az imuths , and any one of thes e will be referred to as a {110}-az imuth. It follows from cons iderations of s ymmetry that if the intens ity of s cattering exhibits a dependence upon az imuth as we pas s from a {100}-az imuth to the next adjacent {111}-az imuth (60), the s ame dependence mus t be exhibited in the revers e order as we continue on through 60 to the next following {100}-az imuth. Dependence on az imuth mus t be an even function of period 2p/3. In general, if bombarding potential and az imuth are fixed and exploration is made in latitude, nothing very s triking is obs erved. The intens ity of s cattering increas es continuous ly and regularly from z ero in the plane

1 de 3

01/03/09 13:11

nature physics portal - looking back - Waves of mat...

http://www.nature.com/physics/looking-back/davis...

of the target to a highes t value in co-latitude 20, the limit of obs ervations . If bombarding potential and co-latitude are fixed and exploration is made in az imuth, a variation in the intens ity of s cattering of the type to be expected is always obs erved, but in general this variation is s light, amounting in s ome cas es to not more than a few per cent. of the average intens ity. This is the nature of the s cattering f or bombarding potentials in the range from 15 volts to near 40 volts . At 40 volts a s light hump appears near 60 in the co-latitude curve for az imuth-{111}. This hump develops rapidly with increas ing voltage into a s trong s pur, at the s ame time moving s lowly upward toward the incident beam. It attains a maximum intens ity in co-latitude 50 for a bombarding potential of 54 volts , then decreas es in intens ity, and dis appears in co-latitude 45 at about 66 volts . The growth and decay of this s pur are traced in Fig. 1. A s ection in az imuth through this s pur at its maximum (Fig. 2Az imuth-330) s hows that it is s harp in az imuth as well as in latitude, and that it forms one of a s et of three s uch s purs , as was to be expected. The width of thes e s purs both in latitude and in az imuth is almos t completely accounted for by the low res olving power of the meas uring device. The spurs are due to beams of scattered electrons which are nearly if not quite as well defined as the primary beam. The minor peaks occurring in the {100}-az imuth are s ections of a s imilar s et of s purs that attains its maximum development in co-latitude 44 for a bombarding potential of 65 volts . Thirteen s ets of beams s imilar to the one jus t des cribed have been dis covered in an exploration in the principal az imuths covering a voltage range from 15 volts to 200 volts . The data for thes e are s et down on the left in Table 1.(columns 1-4). Small corrections have been applied to the obs erved co-latitude angles to allow for the variation with angle of the 'background s cattering,' and for a s mall angular dis placement of the normal to the facets from the incident beam. T a b le 1 Electron Beams . Az imuth. Bomb. Pot (volts ). 54 100 174 174 {100} 65 126 190 159 {110} 138 170 {111} {100} {110} 110 110 110 X-ray Beams . v x 10-8 cm./s ec. 4.36 5.94 7.8 4 7.8 4 4.79 6.67 8 .19 7.49 6.98 7.75 6.23 6.23 6.23 nl x 108 cm. 1.65 1.11 0.77 1.76 1.49 1.04 0.74 1.8 8 1.06 0.8 9 1.8 2 1.8 2 1.05 0.99 0.91 0.8 3 2(0.95) 0.98 0.95 0.98 2(0.97) 1.02 0.95 1.56 1.56 0.90

CoCoColx lat. Inten- Refleclat. s ity. tions . 108 cm. lat. q. q. q'. 50 31 21 55 44 29 20 61 59 46 58 58 58 0.5 0.5 0.9 0.15 0.5 1.0 1.0 0.4 0.07 0.07 0.15 0.15 0.25 {220} {331} {442} {440} {311} {422} {533} {511} {420} {531} .. .. .. 2.03 1.49 1.13 1.01 1.8 4 1.35 1.04 1.05 1.22 1.04 .. .. .. 70.5 52.7 44.0 31.6 31.6 22.4 70.5 52.7 59.0 43.2 38 .9 27.8 28 .8 20.4 77.9 59.0 78 .5 59.5 57.1 41.7 .. .. .. .. .. ..

{111}

If the incident electron beam were replaced by a beam of monochromatic X-rays of adjus table wave-length, very s imilar phenomena would, of cours e, be obs erved. At particular values of wave-length, s ets of three or of s ix diffraction beams would emerge from the incident s ide of the target. On the right in Table 1. (columns 5, 6 and 7) are s et down data for the ten s ets of X-ray beams of longes t wave-length which would occur within the angular range of our obs ervations . Each of thes e firs t ten occurs in one of our three principal az imuths . Several points of correlation will be noted between the two s ets of data. Two points of difference will als o be noted; the co-latitude angles of the electron beams are not thos e of the X-ray beams , and the three electron beams lis ted at the end of the Table appear to have no X-ray analogues .

2 de 3

01/03/09 13:11

nature physics portal - looking back - Waves of mat...

http://www.nature.com/physics/looking-back/davis...

The firs t of thes e differences is s ys tematic and may be s ummaris ed quantitatively in a s imple manner. If the crys tal were contracted in the direction of the incident beam by a factor 0.7, the X-ray beams would be s hifted to the s maller co-latitude angles q' (column 8 ), and would then agree in pos ition fairly well with the obs erved electron beams the average difference being 1.7. As s ociated in this way there is a s et of electron beams for each of the firs t ten s ets of X-ray beams occurring in the range of obs ervations , the electron beams for 110 volts alone being unaccounted for. Thes e res ults are highly s ugges tive, of cours e, of the ideas underlying the theory of wave mechanics , and we naturally inquire if the wave-length of the X-ray beam which we thus as s ociate with a beam of electrons is in fact the h/mv of L. de Broglie. The comparis on may be made, as it happens , without as s uming a particular corres pondence between X-ray and electron beams , and without us e of the contraction factor. Quite independently of this factor, the wave-lengths of all pos s ible X-ray beams s atis fy the optical grating formula n l = d s in q, where d is the dis tance between lines or rows of atoms in the s urface of the crys tal thes e lines being normal to the az imuth plane of the beam cons idered. For az imuths -{111} and -{100}, d = 2.15 x 10-8 cm. and for az imuth {110}, d = 1.24 x 10-8 cm. We apply this formula to the electron beams without regard to the conditions which determine their dis tribution in co-latitude angle. The correlation obtained by this procedure between wave-length and electron s peed v is s et down in the las t three columns of Table 1. Fig u re 2 Intens ity of electron s cattering vs . az imuth angle54 volts , co-latitude 50. high-res olution vers ion

In cons idering the computed values of n(lmv/h), lis ted in the las t column, we s hould perhaps dis regard thos e for the 110-volt beams at the bottom of the Table, as we have had reas on already to regard thes e beams as in s ome way anomalous . The values for the other beams do, indeed, s how a s trong bias toward s mall integers , quite in agreement with the type of phenomenon s ugges ted by the theory of wave mechanics . Thes e integers , one and two, occur jus t as predicted upon the bas is of the correlation between electron beams and X-ray beams obtained by us e of the contraction factor. The s ys tematic character of the departures from integers may be s ignificant. We believe, however, that this res ults from imperfect alignment of the incident beam, or from other s tructural deficiencies in the apparatus . The greates t departures are for beams lying near the limit of our co-latitude range. The data for thes e are the leas t trus tworthy. C. DAVISSON . L. H . GERM ER. Bell Telephone Laboratories , Inc., N ew York, N .Y., M ar. 3. | return to looking back index page |

3 de 3

01/03/09 13:11