the philip,s. electron microscope em 300 phillps technical review volume 29 the philip,s.electron...

370 PHILlPS TECHNICAL REVIEW VOLUME 29

The Philip,s. electron microscope EM 300

e. J. Rakels, Le. Tiemeijet and K. W. Witteveen

Two parallellines of attack can be distinguished in the development of the Philips electronmicroscopes. Improvement of the resolving power goes hand in hand with improvement ofthe arrangement and design of the instrument, giving maximum efficiency of operation andease of use. The latest Philips electron microscope, the EM 300, offers some noteworthy

. advances in both of these aspecis.

Introduetion

The range of electron microscopes put on the marketby Philips in the last 20 years, the EM 100, the EM 75and the EM 200 [1], has recently been extended by theaddition of a new instrument, the EM 300. The maindifference between this microscope and the others is itseven better resolving power (less than 5 A). The EM 300is also simpler to operate than its predecessors, and inits equipment ánd design even more attention than be-fore has been paid, Jo the convenience of the user. Alarge number of matching accessories have been devel-oped; so that the microscope can be used for a verywjC!~variety ·of investigations without requiring majorchanges ofthe instrument. The picture can be displayedoil'a1elevisio~ mbnito;.[2] as well as on the usual fluor-escent 'screen. The magnification can be varied in 31calibrated steps from 220 tirnes to 500 000 times; therange of possible magnifications thus partly overlapsthat of the optical, microscope., As can be seen from fig. 1, the microscope tube ofthe~.. . .

new instrument is vertical, .as-it is jn the EM 200 andthe'EM 75. This arrangement has proved to be the bestone for stability, ready accessibility of the operatingcontrols, easy attachment of accessories and simplemaintenance. The cabinet on the left-hand side containsthe sensitive electronic circuits, such as the stabilizingcircuit>f;r the' high voltage and' the lens currents, andvarious auxiliary circuits. The vacuum system is accorn-modated in ~!1Ccentral 'section, behind the space forthe-operator'sfèet. This 'central section, which carriesthe' microscope tube, is very .sturdily built. to ensuremechanical stability, and its.weight is taken directly bythe floor, independently of the two side cabinets:

Ir. C. J. Rakels, J. C. Tiemeijer and Ir. K. W. Witteveen are withthe Philips Industrial Equipment Division (PIT), Eindhoven ..

In the projection chamber, fitted with three viewingwindows, the fluorescent screen is located at the heightof the table-top. This screen can be swung aside to givethe electrons access to a plate camera or to a televisioncamera tube ("Plumbicon" [*1 tube). When the cameratube is used, an image intensifier can be interposed.Facilities are also available for taking photographswith two types of roll film camera in addition to theplate camera.The electronic circuits are all transistor circuits.

Some of them are water-cooled. All circuits are con-tained in standard exchangeable modules. The operat-ing controls not regularly needed are on sliding panelsimmediately below the table-top (fig. 2). There are nocontrols on the table-top itself.One of the distinguishing features of the vacuum

system is that it works automatically, and another isthat the roughing pump can be switched off for hoursonce a sufficiently high vacuum has been reached. Thiseliminates a major source of vibration and noise. Thevacuum is better and is reached more quickly than inthe predecessors of the EM 300. A number of vacuumlocks are provided, so that the filament, photographicplates or films and the specimen can be changed with-out having to admit air into the entire microscope tube.Photographic exposures are also very easy to make.

The film and plate transport is motor-driven, and theEM 300 is fitted with an exposure meter which is ar-ranged so that photographs may be taken in much thesame way as with a semi-automatic camera: for correct

[*] Registered Trade Mark for television camera tubes.[1] A description of the EM 100 electron microscope is given in:

A. C. van Dorsten, H. Nieuwdorp and A. Verhoeff, Philipstech. Rev. 12, 33, 1950/51, and of the EM 75 in A. C. vanDorsten and J. B. Le Poole, Philips tech. Rev.17, 47, 1955/56.

[2] See P. H. Broerse, A. C. van Dorsten and H. F. Premsela,Philips tech. Rev. 29, 294, 1968 (No. 10).

1968, No. 12 ELECTRON MICROSCOPE EM 300

Fig. I. The Philips EM 300 electron microscope.

371

372 PHILlPS TECHNICAL REVIEW

exposure the operator adjusts the shutter speed or cur-rent density on the screen until the pointer of a meteron the control panel is at the centre of the scale. Tomake an ordinary exposure and a diffraction exposureof the same object one after the other, it is simply nec-essary to turn the knob of the "function switch".

An im portant feature of the new microscope is thatall parts of the electron-optical system can be separatelyaligned. This minimizes displacement of the image inthe image plane, due say to variations in the strength ofthe lenses.

The very high resolving power of the new microscopeis illustrated by fig. 3. It is so good partly because oftbe mechanical stability noted earlier and partly be-cause of the excellent stabilization of lens currentsand high voltage, the low sensitivity to external fieldsand, in particular, the extremely short focal length ofthe objective lens (1.6 mm). The importance of makingthe focal length short is explained in the next section,where the image formation and the electron-optical

VOLUME 29

Fig. 2. The operating controls.that are most frequently neededare within easy reach above thetable-top; the others are on slid-ing panels immediately below it.

factors that deterrnine theresolving power are dealtwith. The other sections willbe concerned with technicalfeatures of the microscopeand accessories.

Electron optics and resolv-ing power

Fig. 4a shows schemati-cally the principal electron-opticai corn ponents insidethe microscope tube. Apartfrom various diaphragms,these components (from

Fig. 3. Two photomicrographs of the sarne specimen in which de-tails situated 5 A apart can be distinguished. The fact that thesarne features are seen on both exposures proves that they are notgrains from the photographic emulsion.

1968, No. 12

G[UJ 0

-ea, Di

C21i-IlD2I

O~~~

Dill""m-D4 Pr

ImDprll.m

'I, 1, 1, 1, 1, 1, 1

S-'---'-

Q 12.

ELECTRON MICROSCOPE EM 300

r

0'I'Sp

0---1 ,Sp

- -03- -04

Di DiI

[-- + I

IPr Pr

s.

373

Fig. 4. a) Schematic arrangement ofthe microscope tube. G emission chamber with electron gun. Cl firstcondenser lens. C2 second condenser lens. 0 objective lens. Sp specimen. Di diffraction lens. I intermediatelens. Pr projector lens. Sfluorescent screen. Dl and D2 condenser diaphragms. De objective diaphragm.D4 selector diaphragm.b) Path ofrays in normal use (high magnifications). The intermediate lens and the diffraction lens are bothin operation. ,I

c) Path of rays for medium magnifications. The objective lens forms a virtual image of the object. Theintermediate lens is switched off. .d) The diffraction lens used as objective (very small magnifications, for initial investigation ofa specimen).In the case illustrated the objective lens is weakly excited.e) Electron diffraction. At the object plane of Pr the diffraction lens now no longer forms an image of theobject, but ofthe focal plane F2 ofthe objective. Ifnecessary the intermediate lens canbe switched off.

top to bottom) are the electron gun, two condenserlenses, the objective lens, the diffraction lens, the inter-mediate lens and the projector lens. Altogether, there-fore, there are six lenses, all of them electromagnetic.The specimen is located approximately at the upperfocal plane of the objective lens. There is a choice ofaccelerating voltages of 20,40,60, 80 or 100kV.The gun and the two condenser lenses constitute the

"illumination system". The first condenser lens Cl is apowerful lens whose excitation can be varied. The mini-mum focallengthis 1.5 mm. This lens forms a reducedimage of the electron source, with a diameter of about0.3 [Lmat maximum excitation (for most applicationsthe maximum excitation will not be used, but one givinga spot diameter between 5 and 1 [Lm).The second con-denser lens C2 is a relatively weak one, whose excitationcan also be varied. This lens is adjusted in such a waythat the image given by Cl appears with unit magnifica-tion at or near the specimen ~ usually slightly abovethe specimen since this setting gives a small illuminatingaperture, which is desirable in certain cases (see below).The size of the spot projected on the specimen, and

its intensity (current density) can be varied over a widerange by varying the strength of the two condenserlenses (fig. 5). This has considerable advantages whichare not present in an illumination system using onlyone condenser lens [31. In the first place, the electron

, ;

loJ,'....-----..------. 10o'i-------,;---;-----,-~___,um A/cm2 ' ia~3;.5 ~ .•rad:

~/9'. "I-~~----,HHr:--~~~_I5 a

la21--__:",,~_+_-t;fj~~-/-'--I ;;~ ; ,Fj m~

if----?L--ftt------><--~_/f:Tdl----I-----lTCJ"

5

2Td!.Q'.'=5-------:1:'=:a-------=

• 2_(ni)

a

T.a '_(ni)2

bFig. 5. a) Variation of the radius r of the electron spot on thespecimen as a function of the excitation of C2 with constant ex-citation of Cl (ni is the number of ampere-turns in arbitrary units)and constant diameter of the condenser diaphragms.b) Similar curves for the aperture GC and the current density jinthe specimen plane; these curves coincide ifthe ordinate scale isproperly chosen.

In (a) and (b), the solid-line curves apply for Cl strongly ener-gized, the dashed curves for weak energization. The chain-dottedlines are found when Cl is switched off. ,When the strength of Cl is varied, while C2 keeps the spot

in focus on the specimen, the size of the illuminated part of thespecimen changes, but not the aperture or the current density.These can be reduced only by defocusing C2. ,

[a] For a discussion of the advantages of a condenser with twolenses, see S. Leisegang, Proc. 3rd Int. Conf. on Electron Mi-croscopy, London 1954.


spot can be made almost as smaIl as the part of the spec-imen to be investigated, without iIIuminating it morestrongly than is necessary. This permits the thermalloading of the specimen to be kept to a minimum. Sec-ondly, when the light spot is smaIl, only a few elec-trons are scattered in parts of the specimen that areoutside the field of view, and this improves the qualityofthe image. Thirdly, when the currentpassingthroughthe specimen is smaIl, there is less trouble from impuri-ties liberated from parts of the microscope or fromcharging effects.

The specimen can be iIIuminated in the normal way,or with oblique iIIumination (dark-field iIIumination).The maximum value that can be chosen for the anglebetween the axis of the beam and the optical axis of themicroscope is as much as 5°.The optical characteristics of the next component,

the objective lens, which are of great importance to thequality of the image formation and the resolving powerwill be dealt with in more detail later. It is sufficient tonote here that this lens can be varied in strength, thatthe coil is situated above the specimen, and that thespecimen is located inside the magnetic field of this lens(immersion objective),The diffraction lens, which is relatively weak and

has a wide bore, and the intermediate lens can also bothbe varied in strength. If these two lenses are used -.either together or separately - as an intermediate lens,the whole range of magnifications can be covered with-out having to alter the strength ofthe projector lens [4].

For high arid very high magnifications (greater than20000 times) the combination shown in fig. 4b is used,for moderate magnifications (4000 to 20000 times) thecombination of fig. 4c, and for smaIl magnificationsthat of fig. 4d. In the last case the diffraction lens infact functions as a (non-immersion) objective lens; thediffraction lens could hardly be used in this way if thespecimen were above the coilof the objective lens, be-cause the distance from the specimen to the lens wouldthen be too great.

When both the diffraction and the intermediate lensesare used, as in fig. 4b, the magnification is varied byvarying the strength of the intermediate lens. (Thestrength of the objective lens is also varied slightly inorder to keep the image.in focus.j.The diffraction lens isalways set to maximum strength. With the arrangementof fig. 4c, the interrnediate lens is switched off and themagnification is varied by means of the objective andthe diffraction lenses. As can be seen, the objective hereforms a virtual image of the object. This image, whichmoves upwards with increasing magnification, is pro-jected by the diffraction lens on to the object plane ofthe projection lens. With the mode of operation shownin fig. 4d, the objective lens is usually weakly energized

so as to cause the electron paths to cut the optical axisat the location of the selector diaphragm D4. This canthen act as an objective diaphragm, so that good con-trast is stiIl obtained when the instrument is used in thisway. If it is desired not to subject the specimen to amagnetic field, or to subject it only to a field whosestrength is accurately known [5], the objective lens isnot excited. As a result of the arrangements andmodes of operation illustrated in fig. 4a-d, the distor-tion of the magnified image obtained with the EM 300is Iow even at the weakest magnifications. FinaIly,fig. 4e shows the path of the rays when the microscopeis used as an instrument for making electron diffractionphotographs.

Resolving power

The resolving power of the EM 300 is primarily de-termined by the objective lens. The less fundamentalfactors affecting the resolving power, such as the me-chanical and electrical stability, are oflesser importancein the new microscope. We shaIl therefore briefly reviewthe way in which the resolving power is affected by theprincipal lens errors: spherical aberration, chromaticaberration and astigmatism.

Chromatic aberration cannot be neglected because an electronbeam in an electron microscope can never be completely mono-energetic: there is energy dispersion immediately outside the gunsince the electrons are liberated by thermionic emission, veloci-ties are affected by space charge in regions where the paths crossone another (Boersch effect) and finally not all the electrons aresubject to the same amount of energy loss in the specimen.These effects make it necessary to take account of chromaticaberration even in an electron microscope that is electrically per-fectly stabilized [61.

Every magnetic lens with circular symmetry suffersfrom positive spherical aberration, that is to say the fo-cal length is smaIler for the peripheral rays than for theparaxial rays. Unfortunately, no practical method hasyet been found for correcting this lens error. Thespherical aberration can, however, be minimized bychoosing appropriate values for the spacing and boreof the pole-pieces, and in particular for the excitationof the lens. This minimum of spherical aberration de-creases as the focallength of the lens is made smaller.This is the primary reason for making the focallengthofthe objective lens as small as possible in order to ob-tain a good resolving power.The chromatic aberration of an electron lens is also

less when the focal length is smaller. The chromaticaberration constant Cc of the EM 300 (see below) istherefore only 1.3mm. This means that the spread ofthe energy of the electrons will prevent the theoreticalresolving power from being achieved only when theelectrons pass through a relatively thick specimen. Pro-vided thespecimen is sufficiently thin (e.g. a carbon film

1968, No. 12 ELECTRON MICROSCOPE EM 300 375

of 20 nm), only a small fraction of the electrons loosesenergy in the specimen. .If spherical aberration only were present, and if the

situation in an electron microscope were entirely anal-ogous to that in an optical microscope - in particularwith respect to the illumination of the object - thetheoretical resolving power found from the classicaltreatment due to Rayleigh would be 2.3 A (0.23 nm) atan accelerating voltage of 100kV.In ordinary optics the classical treatment of resolving

power has been supplemented in the last 20 years by thetheory of contrast transmission, which uses contrast-transfer functions and curves. For some years now thistype of treatment has also been applied to electronlenses [7]. It will appear that the value of resolvingpower quoted above can be approached very closely,but also that the interpretation of the image of an ob-ject whose details are of the order of magnitude ofängströrns is a complicated matter.We shall now consider how the transfer of contrast

in the object is affected by spherical aberration, de-focusing of the microscope and, through chromaticaberration, by fluctuations in the high voltage and thelens currents.Let us consider a beam of rays emitted from the

first focal point. If there were no defects in the lens, thespherical wave-front of the beam would emerge fromthe lens as a planar wave-front. The lens defects do how-ever cause some deviation, known as the wave-front ab-erration ,1, which depends on the radius vector rb inthe plane under consideration and on the position ofthat plane. (For object points lying in the first focalplane but away from the optical axis, different values of,1 are found at a given rb, but this complication may beneglected in an electron microscope since the objectsare relatively small.) Taking only the spherical aberra-tion into account, then the wave-front aberration inthe second focal plane is given by LI(rb) = -Arb4.If the object point is on the axis but not at thefirst focal point, an extra term Brb2 enters into theequation, where B depends on the distance L1zFof bothpoints, the "defocusing". The total wave-front aberra-tion is then:

Fig. 6 gives a curve of this function for the objectivelens of the EM 300, for the case B = 0 and also for fourcases where there is a certain amount of defocusing.The wave-front aberration ,1 is expressed in terms of thewavelength of electrons at 100keY (0.037A). A secondhorizontal scale beneath the graph gives rb in terms ofthe line spacing (grating constant) d of a sine gratingthat diffracts an electron beam of 100keV through anangle Cl. such that the diffracted beam is focused into

00 3~5d_

Fig. 6. The wave-front aberration zl in the second focal plane ofthe objective, caused by spherical aberration and defocusing, as afunction ofthe radius vector rb, for an electron beam of 100 keYenergy, emitted from the primary focus Fi or from a point a dis-tance of LlZF away from it on the optical axis. The scale beneaththe figure gives the grating constant d corresponding to rb.

(1)

the second focal plane at a distance rb from the opticalaxis.

From the curves of fig. 6 we can immediately find theeffect of the spherical aberration and the defocusingon the diffracted image obtained from a given gratingin the second focal plane, and thus at the same timeobtain some idea of the imperfections that the magnifiedimage on the screen will have. This comes about sincean object can be regarded as a superposition of a varietyof phase and amplitude gratings with differentconstants.(An object is a phase object when the emergent 'wavehas the same amplitude and thus the same intensityat all points, but differs locally in phase.) For example,at zero defocusing the beam diffracted by an am-plitude grating with d = 5.3 A has a wave-front aberra-tion of -tÀ. This means that the image ofthis gratinglooks like that of a phase grating, i.e. the grating struc-ture is completely imperceptible on the screen. On theother hand, if the object is a phase grating, its imagewillhave maximum contrast. Because of the sphericalaber-ration the structure of the phase grating, which in theideal case would be imperceptible, can now be observed., If we determine the contrast with which amplitudeand phase gratings with a different grating constantare reproduced, and we plot the result in a graph, weobtain what are known as contrast-transfer curves. The

[4J The advantage of using a third magnifying lens, in additionto the objective and the projector lens, was first pointed out byJ. B. Le Poole in Philips tech. Rev. 9, 33, 1947/48.

[SJ See for example H. B. Haanstra and H. Zijlstra, Philips tech.Rev. 29, 218, 1968 (No. 7). .

[6J See the second article quoted in reference [IJ.

[7J See for example K.-J. Hanszen, Naturwiss. 54, 125, 1967(No;6).


-1I I I I I I 3A20 15 10 8 7 3.5

d-+1

Fig. 7. Contrast transfer curves for the objective lens of theEM 300, derived from the curves of fig. 6; upper curves for ampli-tude gratings and lower curves for phase gratings.

curves for the EM 300 are shown in fig. 7. It can beseen from this figure even more clearly than from theprevious one that there is no setting at which the finestgrating structures (d < 10 A) can all be seen at thesame time. For instance, with a defocusing of 120 nmamplitude gratings with d = 3 A and 3.5 < d < 6.5 Aare reproduced with good contrast, but those where dis in the region of 3.25 and 9 A are not. The best settingin this respect is the one with a defocusing of 40 nm.The corresponding defocusing for phase structures is80 nm. We should add that there are many specimens(particularly biologicalones) that yield only very weakamplitude contrasts. In the investigation of such speci-mens the presence of phase contrasts can be a help in-stead of an interfering effect.From the foregoing we can find the stability re-

quirements which the accelerating voltage Vo and theexcitation current Iofthe objective lens should meet inorder to minimize the effect of fluctuations in I and Voon the resolving power, The displacement oz of thefocal point caused by variations 0 Vo and Ol is given by:

OZF';:;Cc {(OVoÎVo)-2(0I/I)}. (2)

The quantity Ccis the chromatic aberration constantmentioned previously. We can now calculate that thecontrast decreases to 90% if, in the period of observa-tion (one minute at the most), the wave-front aberra-tion LIvaries between +tÀ and -!À, and all valuesare equally likely. Taking this as the criterion, we findfrom the wave-front aberration curves what variationin defocusing is permissible, and then from equation

(2) what relative variation of Vo and I this correspondsto at a given value of d (Table I). For the EM 300 therelative drift of Vo is less than 5X 10-6 per minute, andthat of I less than 2.5 X 10-6 per minute. It is evidentfrom this that the drift of I and Vo in the EM 300does not prevent the theoretical resolving power frombeing reached.Table J. Maximum deviation in the accelerating voltage Vo andthe lens current I at which a phase object with a periodic structure(phase grating) of line spacing d, still gives an image with goodcontrast.

d (Á) r5 Vol Vo - 2 MIl

105432.3

110 X 10-626 X 10-617 X 10-69.7 X 10-65.7 X 10-6

If the same is to apply for the effect of the astigma-tism ofthe objective lens, then there should be virtuallyno change in the contrast-transfer curves when theplane in which the diffracted beam lieschanges position.If this plane is rotated one full turn around the opticalaxis, the defocusing range must not be greater thanabout 10 nm. This means that the requirements forthe circular symmetry of the pole-pieces are very exact-ing. These requirements cannot be met using normalmachining methods and tools. The pole-pieces of theEM 300are therefore made on a specially designed ultra-high precision machine equipped with hydrostaticbearings [81. Before the pole-pieces are mounted in themicroscope they are checked with a precision instru-ment, one of whose special features is that its own me-chanical deviations do not affect the measurement [91.

Even with these special devices and provisions theastigmatism still cannot be reduced to a sufficientlysmall value. The residual astigmatism is compensatedby means of a stigmator, which in principle is simply acombination of two cylindrical magnetic lenses whosestrength can be varied. In the EM 300 the stigmatoris a set of eight small coils mounted directly under theobjective lens.Apart from the above-mentioned characteristics of

the objective lens and the more trivial factors referredto in the introduction, like the mechanical stability andsensitivity to stray magnetic fields, other factors affect-ing the resolving power are the quality of the specimenstage and the brightness of the electron source.

[8] See J. G. C. de Gast and H. J. J. Kraakman, shortly to be. . published in Philips tech. Rev.[9] This instrument was built by the Central National Organiz-

ation for Applied Scientific Research in the Netherlands(TNO), Technical University of Delft, and was describedby J. Kramer, J. B. Le Poole and A. B. Bok in TNO-Nieuws20,297, 1965 (in Dutch, with English summary). .

1968, No. 12 ELECTRON MlCROSCOPE EM 300

The specimen stage must be capa-ble of holding the specimen inposition to within a few ángströrnsduring the period of an exposure.The specimen stage in the EM 300is the same as the one used in theEM 200, which meets these require-ments.

The brightness of the electronsource comes into the picture in thefollowing way. We have just seenthat in order to make the very finestdetails visible, it is often necessaryto make use of phase contrasts. Thiscalls for highly coherent illumina-tion, and this condition is most easilymet with a small illuminating aper-ture. (The condenser system of theEM 300 does in fact permit operationwith a small illuminating aperture,as we noted earlier.) J n order toobtain a sufficient current densityat the fluorescent screen when usi nga small illuminating aperture andthe maximum magnification of theinstrument, it is necessary to havean extremely bright electron source. Oi---l'+'

The microscope tube

In this section we shall considerthe principal design features of thecomponents contained in the micro-scope tube. A simplified cross-sec-tional drawing of the tube is shownin fig. 8. In addition to the compo-nents in fig.4, the drawing shows

-- --- - --- ---------

Fig. 8. Cross-section of the microscopetube. G gun. An anode. Cl and C2 conden-ser lenses. 0 objective lens. St specimenstage. Di diffraction lens. 1 intermediatelens. Pr projector lens. j girn bal ring. 2knob for diaphragm Dl (fig. 4). 3 knob fordiaphragm D2. 4 coils for oblique illumina-tion and focusing. 5 knob for objective dia-phragm D3. 6 knob for selector diaphragmD4. 7 alignment controlsfor lenses and pole pieces.8 vacuum valves. 9 shut-ter.!10camera (35 mm). 11binocular microscope. 12plate camera. 13 vacuumline. 14 rapid couplingdevice. The camera 10can be replaced by anelectron-diffraction de-vice (i.e, a specimen hol-der for large specimens).

13

377


the specimen stage, the stigmator, the coils for obliqueillumination, the photographic shutter and a numberof vacuum valves and locks, as well as a number ofcontrols for operating the various diaphragms andfor the alignment of lenses, pole-pieces, etc.

Electron-optical section

The electron gun is of the conventional type, com-prising a filament, a Wehnelt cylinder and an anode. Itcan be aligned by displacing the cathode (here taken tobe the combination of filament and Wehnelt cylinder)with respect to the anode and rotating it about twoaxes perpendicular to the centre-line of the microscopetube and passing through the emitting tip of the fila-ment. The mounting of the cathode in a gimbal bearingensures that it moves smoothly and can be accuratelyaligned. The height of the anode can be adjusted whilethe gun is in operation. This is important in order toachieve maximum brightness from the electron gun atvoltages lower than 100 kV. Facilities such as the rapidcoupling arrangement and the vacuum lock situatedimmediately below the electron gun enable a filamentto be replaced in less than a minute.

Each of the condenser lenses is fitted with a dia-phragm holder containing three interchangeable dia-phragms. Fitted against the lower pole-piece of C2 is adisc which contains the stigmator. This corn pensates fordepartures from circularity of the illuminating beamcaused by astigmatism in the condenser system. (If thebeam is not of circular cross-section, the specimenmay receive more thermal energy from the beam thannecessary, and there may be unwanted charging effectsat the edges of diaphragms, etc.)

In addition to the above-mentioned facilities for dis-placing and tilting the gun, the illumination system hasa variety of other correction and alignment facilitieswhich cannot all be mentioned here.

The design of the objective lens is illustrated in jig. 9.This lens, together with the specimen stage (shownshaded), forms a single assembly, which however caneasily be removed in order to change from one type ofspecimen stage (see below) to another.

The upper pole-piece of the objective lens cannot bedisplaced and constitutes one of the fixed points usedfor alignment (the other is the centre of the fluorescentscreen). The lower pole-piece can be aligned. When thecurrent in the coil is at its maximum value (5 A),the magnetic field between the pole-pieces is aboutJ.5 Wb/rn". Situated immediately under the lowerpole-piece is the stigmator of the objective lens.

Fitted in the bore of the magnet coil, above the lensfield, are the deflection coils for oblique illumination.These operate on the same principle as Le Poole'sfocusing device [lOl, and they can also be used for this

purpose. Here, however, there are two pairs of coilsinstead of one. These pairs are arranged at right anglesto one another, so that the azimuth of the plane inwhich the deflection takes place can be chosen as re-quired.

The deflection coils of the focusing device are woundin such a way that the magnetic field they produce ishighly uniform. There are also a few correction coils tocompensate for any difference in orientation that mayexist between the upper and the lower coils. With thesefacilities the angle between the beam and the opticalaxis can be increased to the high value of 50 mentionedearlier, with no noticeable change in the location atwhich the beam meets the specimen. When a set ofthese coils is used for focusing, very sharp images canbe obtained even at quite high magnifications.

Fig. 9. Schematic cross-section of the objective lens and specimenstage (shaded). The pole-pieces are shown black. When the upperpart (including the coil) is detached from the ring W, the specimenstage can easily be removed. The lower pole-piece can be aligned.

Fig. lOa shows how the specimen stage is mountedin the microscope tube. Fig. J Ob shows the centralsection, which contains across-slide system. This systemis driven by a system of pushrods and two levers, eachdriven in turn by micrometer screws, operated fromthe control panel. Exceptionally smooth and accuratemovement of the specimen is obtained by proper bal-ancing of frictional forces, spring forces and massesand by the appropriate choice of material. As in theprevious Philips microscopes, the specimen is intro-duced through a vacuum lock by means of an injector.

The diffraction lens, the intermediate lens and theprojector lens are relatively simple in design. The diffrac-tion and intermediate lenses can be fully aligned, i.e.the lens can be displaced as a whole and in additionone of the two pole-pieces can be displaced with re-spect to the other. The projector lens can only be dis-


placed as a whole, which is sufficient because this lensis not varied in strength and no further magnificationtakes place after it.

Projection chamber and aids for viewing and recording

The aperture of the beam that converges at one pointof the image formed in the projection chamber of anelectron microscope is very much smaller than the

picks up a plate from a magazine, carries it into positionunder the microscope tube and, after the exposure ismade, lets it drop into a box. The disc also carries thevacuum valve which shuts off the projection chamberfrom underneath. This is done by turning the disc untilthe vacuum valve is immediately under the tube open-ing and then raising it until it closes the opening.

Between the projector lens and the 35 mm camera

Fig. 10. a) The standard specimen stage in the microscope. The upper section of the tube shown here isthe ring W (fig. 9) of the objective lens. b) A detailed view of the central section. 1 upper cross-table slide.2 lower slide. 3 push-rod for the movement of the upper slide. 4 specimen holder. 5 specimen. 6 coolingelement (-180°C); this accessory does not belong to the stage but is intro duced separately from the side.

a

aperture of the projector lens itself, and therefore thedepth of focus is very great. This gives the designer con-siderable freedom in the positioning of the variousviewing aids. A 35 mm camera is accommodated direct-ly under the projector lens, and a plate camera justbelow the screen. Below this there is room for a cam-era for 70 mm film or for a television camera tube,which may if necessary be used in conjunction with animage intensifier. When the image intensifier is used,exposures can be made at a very low current density, per-mitting the investigation of specimens that would bedamaged at the normal current density. Tn addition tothe ordinary fluorescent screen there is a small screenwith a very fine grain; the image formed on this can beviewed with a binocular microscope at a magnificationof 10 times.

The arrangement of the plate camera is illustratedschematically infig.ll. The main component is a rotat-ing disc situated eccentrically with respect to the axisof the microscope tube. At every revolution this disc

[101 See the first article quoted in reference [IJ.

b

there is a rotating shutter, which can be used for allphotographic exposures. A rotating shutter has con-structional advantages in view of the requirement thatopening and closing must cause the least possible vibra-tion.

Fig. 11. Schematic representation of plate camera. The tootheddisc D meshing in a smaller gear is motor-driven. S plate maga-zine. R box for receiving exposed plates. V vacuum valve. Whenthe opening H arrives above S, a plate is pushed in. This travelsaround with the disc into position under the microscope tube,where it is exposed and then dropped into R.

380 PHILIPS TECHNICAL REVIEW VOLUME 29

Vacuum system and electrical equipmentThe vacuum system of the EM 300 (fig.12) is basic-

ally conventional - comprising a roughing pump,vacuum reservoir, mercury ejector pump and oil diffu-sion pump - but differs in some significant details fromthe system used in the earlier Philips microscopes.In the first place facilities are provided which make

the entire pumping process completely automatic. Be-sides making the instrument very simple to operate,automation has the advantage that the pumping pro-cess is optimized and human errors in operation are

v.... --

D

closes, the roughing pump stops and V4 opens. Sinceafter the second phase the roughing pump is isolatedfrom the microscope tube by the closing of V2,contam-ination of the vacuum system by oil from the rough-ing pump is minimal; there is no tack diffusion of oilvapour above about 0.1 torr. Another favourable fea-ture here is that the roughing pump normally operatesfor so short a time that it barely gets warm.

There are of course various safety devices in the auto-matic system which protect the microscope from dam-age if the water-cooling should fail, if there should be a

Fig. 12. The vacuum system. The microscope tube, shown on the left, is connected by threebroad tubes with the oil diffusion pump ODP. MDP mercury ejector pump. B fore-vacuumvessel. P roughing pU!l1P. M; Pirani gauge. M2 Penning gauge, Ma mercury pressure gauge.Vvacuum valves . i: 'vacuum lock, which can be separatelyevacuated. K vessel. with dryingagent. Valves Vl, V2 and V4 are electromagnetic, Va is motor-driven.

excluded, thus reducing the risk of breakdown. In thefirst phase only the roughing pump is in operation andall the valves, apart from VI, are closed. When thepressure at the location of the Pirani gauge MI hasdropped to about 0.1 torr and the two diffusion pumpsare hot enough, VI closes and V2opens. If the pressureat MI again has dropped to 0.1 torr, V2 closes and VIand Va open. The buffer vessel, the mercury ejectorpump and the oil diffusion pump are then in series be-tween the roughing pump and the tube: Once the pres-sure above the oil diffusion pump at the location of thePenning gauge M2 has dropped to ID-a torr, valve VI

serious vacuum leak, if a heating element should breakdown, and so on. The high-vacuum valve Va, for ex-ample, is arranged so that the valve closes immediatelyif there is a mains failure.

Other differences between the equipment of this andthe earlier Philips microscopes are the incorporation ofa pump unit with a very much higher pumping speedand the provision of a number of vacuum valves in themicroscope tube and in the viewing chamber. As wenoted earlier, there is no need to let air into the entiremicroscope tube of the EM 300 when changing a fila-ment or loading the plate camera.


Fig. 13. The electrical circuits are contained in modules that fit into hinged racks.

The electrical equipment of the EM 300 is ratherdifferent from that of the earlier Philips microscopes.The most conspicuous difference is the use of trans-istors instead of valves; in the whole of the electricalequipment there are only two valves, in the high-volt-age stabilizing circuit. The advantage of transistor cir-cuits is that they are more reliable and develop lessheat. ]t is also much easier to use printed wiring withtransistor circuits than with valve circuits. The change-over to transistors made it necessary to develop newcircuits for most sections of the electrical equipment.

The equipment has also been extended. In addition tothe vital components, like the circuits that supply thehigh voltage and the lens currents, there is a monitor-ing circuit that can be connected to no fewer than18 points of these circuits, enabling the user to measure(or record) the principal èurrents and voltages of themicroscope or to check their stability.

To minimize magnetic interference with the beam, thetransformers used in the circuits are all C-core types,which have an extremely weak stray field.

All the electrical circuits are contained in uniformmodular units. This makes the most efficient use of theavailable space, and all the components are readilyaccessible (fig. 13). The power supply for the completeequipment comes from a separate transformer, ar-

ranged in such a way that the microscope can be fedfrom any 50 or 60 Hz single-phase mains supplywhose voltage is between 110 and 440 V.

In the following parts of this section we shall deal atsomewhat greater length with some important circuits.

The universal control amplifier

In order to be able to stabilize the lens currents andhigh voltage to such a degree that their fluctuations donot prevent the theoretical resolving power from beingachieved, fluctuations of about I in 106 have to be de-

tected. Since low voltages are used in transistor cir-cuits, the control voltage derived from a lens current isalso low, only a few volts. Variations of a few micro-volts in the control voltage must therefore still be readi-ly detectable. This calls for a highly sensitive amplifierwith a carefully designed wiring layout: there must beno risk of interference from magnetic or electric fieldsor from coupling with other circuits. To offset thesedifficulties, there is the advantage that in transistorcircuits the resistances across which the control volt-ages appear are small, so that the amplifier does notneed to have a high input impedance.

The preamplifier which we have developed for thecontrol circuits required in the EM 300, and whichmeets the above-mentioned requirements, is a special

.--~--~-~---------------~--- - -_ - - -


Fig. 14. Universal control amplifier. The circuit consists of a transistor chopper which convertsthe d.c. input voltage into an a.c. voltage, a transformer, an a.c. voltage amplifier and a syn-chronous detector. The chopper and the synchronous detector operate at a frequency of1200Hz and are controlled by a common oscillator via a transformer. Data: Floating input. In-put impedance >20000 n. Amplification 1000 times. Frequency response 0-100 Hz (-3 dB).Output voltage swing ±IO V. Between 0 and 50°C the input drift at a maximum sourceimpedance of 1000 n is less than 1 [LVrC and less than 1 [LVj8hours; the noise (peak-to-peak;0-10 Hz) is less than 1 [LV.At a maximum source impedance of20 000 n these values are threetimes as high.

type of chopper amplifier. Its principle of operation isillustrated infig.14, and the main features are summar-ized in the caption. The d.c. input voltage is convertedby the transistor. circuit on the left (the chopper) into analternating voltage with a frequency of 1200 Hz andthen passed by a transformer to an a.c. amplifier. Theoutput signal from this amplifier is converted into ad.c. voltage again by the circuit on the right.

The distinguishing feature of the universal controlamplifier is that the conversion from d.c. into a.c.voltage-is not effected mechanically, but by means of atransistor circuit. This arrangement was chosen formaximum reliability and long life. As can be seen, theinput circuit is separated by a transformer from the a.c.

amplifier; this gives considerable freedom from inter-ference, and permits the unit containing the samplingresistor and the input circuit to be earthed separately.The output signal from the a.c. amplifier is not recti-fied by an ordinary rectifier but by a circuit which ismore or less the counterpart of the chopper circuit,connected to the same oscillator (synchronous detec-tion).

The chopper circuit in fig. 14, like any other chopper,causes interference signals, mainly during switching.Various measures have been taken to minimize these,but they can still reach a peak value of about 200 fLV,which is far in excess of the variations of the controlsignal. Nevertheless, the new amplifier will give faith-ful amplification of control signals of 1 fLV. This is duein the first place to the synchronous detection, and also

to the fact that the amplification has not been madeunduly high, so that input signals may be as high as200 fLV without blocking occurring in the later stagesof the amplifier.

Finally, some remarks on the chopper circuit. As can.be seen, it consists of a symmetrical arrangement offour transistors. These are germanium transistors,selected in matched pairs for identical collector-emittervoltage at zero collector current. Germanium tran-sistors are preferable here to silicon transistors, becausein the transition from "completely" off to "completely"on (and vice versa) they require a much smaller signalon the base (70 mV r.m.s.). The use of germaniumtransistors therefore means that there is much less in-

terference from crosstalk between the switching signaland the control signal- the principal type of interfer-ence - than there would be if silicon transistors wereused. Germanium transistors, however, have a greaterleakage current; as a result, a fairly large drift canoccur if the voltage source delivering the signal to beamplified has a high internal impedance.

Although it might appear that only two transistorscould be used for the chopper circuit, the arrangementwith the two pairs of transistors connected in opposi-tion has been chosen since a transistor is not such anideal switch as a mechanical switch. In the present cir-cuit the zero errors of the members of each pair gener-ally compensate each other.

The chopper amplifier is included with a d.c. differ-ence amplifier of conventional design (gain 2000) in a


rack-mounted module of thekind already mentioned (fig. 15).The electrical equipment of theEM 300 contains altogether ninesuch modular units. They areused for the lens-current andhigh-voltage stabilizers, for themonitoring circuit and also forthe circuit that supplies the ref-erence voltage with which thecontrol signal is compared in eachstabilizing circuit.

Fig. IS. Standard control amplifier, containing a circuit like that of fig. 14 and a con-ventional d.c. difference amplifier. The two amplifiers are mounted on printed wiringboards. Note the mu-metal screening around the transfermers.

The centra! reference voltage

For stabilizing the high voltageand the lens current a referencevoltage is needed that meets atleast the same requirementsofsta-bility. A simple and very effectivesolution is to derive the referencevoltagefrom a battery. but in thatcase the battery must not deliverany current. In the EM 300this is not practicable for thefollowing reason. As the power for the lenses comesfrom transistor circuits, the currents used are relativelyhigh. The sampling resistors must therefore be small(of the order of 1 Q). This means that the current can-not be varied by varying the sampling resistance, be-cause in a resistance of about 1Q the variations in thecontact resistance would be too large compared withthe total value. The lens current can therefore be variedonly by varying the reference voltage, that is to say byconnecting a potentiometer circuit across tbe sourcesupplying this voltage. This rules out the use of abattery. For the EM 300 a reference-voltage source hastherefore been developed that can deliver a current andmeets the appropriate stability requirements. Its out-put voltage is about 10 V and varies less tban 1 in 106per minute and less than 3 in 105 in 24 hours. The cir-cuit can deliver 50 mA and is short-circuit proof. Thetemperature coefficient averages 10-6 per oe and isnever more than 3 X 10-6 per oe. All stabilizing circuitsderive their reference voltage from this single voltagesource.

Fig. 16 shows a block diagram of the reference volt-age source. The main units are the amplifiers discussedabove (fig. IS) and a bridge circuit consisting of threeresistors (RI, R2, R3) and a Zener diode selected for lownoise. The bridge ei rcui t is housed in a small thermostat.Tbe principle of operation is very simple. As soon asthe part of the output voltage determined by R2 and R3differs from the operating voltage of the Zener diode,the cbopper amplifier receives an input signal and the

c

-tGVn, ,','

Fig. 16. Generator for the central reference voltage. Ch chopperamplifier. Diff difference amplifier (see fig. IS). The dashed lineencloses a bridge circuit consisting of three resistors and a Zen erdiode, housed in a thermosrat. C feedback capacitor for the sup-pression of fast fluctuations.

output voltage is returned to the correct value. Varia-tions too fast to pass the chopper amplifier are fed viacapacitor C straight to the input of the difference ampli-fier.

Generation and stabilization of lens current and hightension

Fig. 17 gives a block diagram of the power supplysystem for the lenses and shows how the lens currentsare stabilized. All tbe lenses are fed from a single recti-fier which can supply a current of 20 A. The outputvoltage is 42 V unstabilized and 33 V stabilized. Fromthe output of the voltage stabilizer the current is distrib-uted to the six lens circuits, each of which is a seriesconfiguration of a current regulator Tp, a sampling re-sistor Rm and the lens coil with a polarity reversal unitRU. A power transistor, or a number ofpower transis-tors in parallel, act as the current regulator whicb, to-

384 .PHILlPS TECHNICAL. REVIEW VOLUME 29

gether with the samplingresistor, forms part of thecurrent stabilizing circuit;these transistors are mount-ed on a water-cooled plate.All stabilizing circuits areshort-circuit proof - ifthere is a short-circuit oroverload, the current is re-duced to a safe value. In ad-dition, the stabilizer forthe objective-lens currentcontains a cut-out whichswitches off the lens currentif one of the parallel tran-sistors becomes overloadedor fails. (With a parallelarrangement of five transis-tors this is not immediatelynoticed, but it is with oneor two.) The sampling re-sistors are resistance-alloy coils, housedcooled oil bath to ensure stability.

Fig. 17. Circuit for generating and stabilizing the lens currents. Reet rectifier. Avc voltage-stabilizing amplifier. Tp power transistor(s). RU current reversal unit. L lens coil. Rm sampl-ing resistor. Ch chopper amplifier. D(ffdifference amplifier.1 connection with central referencevoltage. 2 connection with 1.5 Hz generator. 3 connection with monitoring circuit. The resistorR3 is connected to the high-voltage selector switch. The lens current can be varied by means ofthe potentiometer R2.

in a water-

The current is reversed by means of mercury-wetted reed relays.The entire operation is completely automatic and its sequence isas follows. First the lens coil is short-circuited and almost sirnul-taneously cut off from the current source. After about half asecond, which is sufficient for the dissipation of the field energystored in the coil, thi coil connections are reversed, the short-circuit is removed and the current again switched on. These re-lays are extremely 'reliable, and they are also bounce-free andhave a<vèry;lo'wánd stable contact resistance.

~;~ .. '. 'r". " '

Slow variations in the lens èurrent are suppressed byapplying the difference between the voltage across thesampling resistor and th~ reference voltage as an inputsignal to the control amplifier described above (fig. IS).The output signal from this amplifier drives the powertransistor Tp• This control does not respond to fast:fluctuations, as the chopper amplifier cannot handlethem. The output of rp is therefore directly connectedto the input,of the difference amplifier through a capac-itor C (cf. fig. 16).

As explained above, changing over to another lenscurrent entails altering the reference voltage. This isdone by means of a potentiometer circuit (on far leftin fig. 17)consisting of a fixed resistor RI, a resistor withtap R2 and a variable resistor R3. 'The form of R2 differsfor each lens, varying from a very simple to a highlycomplicated arrangement. The objective-lens currentcan be varied by means of a divider in very small stepsover a large range (the smallest steps are 4 in 106).

More than one potentiometer is provided for threelenses. This makes it possible for the user to switchfrom normal magnification to very low magnification,

or to electron diffraction, by merely turning the func-tion switch mentioned previously.There is a similar facility, as in previous Philips

microscopes, for changing over to another highvoltage. The resistor R3, which controls the magnitudeof the current in the potentiometer circuits, is coupledto the high-voltage switch and controls the currentin such a way that when the high voltage is changed toanother value the illumination, magnification and fo-cusing remain practically unchanged.A low-frequency alternating current (1.5 Hz) can be

applied to the input of the chopper amplifier. In thiscase the lens current is modulated approximately 7%,which is necessary for alignment. (Changing thestrength of a lens causes a rotation of the image, sothat the point where the optical axis cuts the imageplane can be found.)The recording shown infig. 18 gives a good idea of

the stability achieved. There is little variation in currenteven over periods much longer than the one minutementioned earlier.The high voltage is stabilized by means of two con-

trol circuits, which have some sections in common.The first is completely analogous with the circuit offig. 17; the correction signal produced in this circuit isadded to the output voltage of the high-voltage genera-tor. This circuit has a limited control range, however(about 1200 V on either side), which would rapidly beexceeded if no further provision were made. This diffi-culty is avoided by arranging that the output voltagealso controls the stabilizing circuit that holds the pri-maryvoltage for the high-voltage transformer constant.In this way only slow variations can be smoothed out,

1968, No. 12 ELECTRON MICROSCOPE EM 300

Fig. 19. Horizontal cross-section of the microscope tube with goniometer specimen stage. W microscopetube (cf. fig. 9). CS specimen stage which has a system of tubes attached to its right-hand side. Inside thesetubes there is a much narrower tube Tl (lightly shaded), whose left-hand end is hemispherical andbears against the spherical surface BI. The position of the other end of Tl is deterrnined by the lever LI,which is moved by the tube T2 when the nut N is turned. In the plane perpendicular to the plane of thedrawing there is a second lever L2 (see detail at bottorn right). The end E of the specimen holder Hs canbe introduced into the microscope tube through the tube Tl; the depth of insertion is determined by thepush-rod R, which is adjusted by the micrometer M.

When the tube T3 resting in ball-bearing B2 is rotated about its axis, the tube Tl and the specimen holderfollow it, so that the angle between the plane of the specimen and the horizontal plane also varies. Inprinciple this angle can have any value, but in practice it is limited to about 60°, since at larger angles theelectrons will no longer be able to pass through the opening in the specimen holder. At every position ofthe specimen plane the specimen can be moved in two directions in this plane: in the axial direction ofthespecimen holder by means of the micrometer M, and at right angles to this by rotating the nut N. Rotat-ing the nut causes the tube Tl to rotate around point A, so that in the drawing, for example, E riseswhen LI pushes down the right-hand end of Tl. HD is the objective-diaphragm holder.

Fig. 18. Recording of lens current. The length of the horizontaldouble arrow corresponds to a time interval of two minutes,the length of the vertical one to a relative current variation.of5 x JO-6. The fast current fluctuations and also the drift duringone minute are no greater than about I in 106. During the record-ing the current was varied twice in steps of 5 x JO-6 (correspond-ing defocusing 12.5 nrn).

but the control range is very wide (0 to 100 kV). To-gether, the two circuits are capable of keeping the highvoltage constant to the extent required. For interferingsignals at a freq uency lower than about 10 Hz the con-trol is principally achieved through the supply voltage.As a result of the control method described, the loop

gain is very high at low frequencies (more than 105 be-low 1 Hz).

The accessories

The accessories for an electron microscope are al-most as important as the microscope itself, since theyoften determine whether or not an investigation can becarried out. Besides the accessories we have alreadymentioned, like the cameras, the television cameratube and the image intensifier, the EM 300 has manyothers: we shall just mention a few of them here. Oneimportant accessory is a cooling element (below120 OK), which can be brought close to the specimen inthe tube to reduce contamination of the specimen bycarbon formed by the breakdown of hydrocarbon mol-ecules (see fig. lOb). There are also a large number ofspecimen holders enabling the specimen to be put undertension, cooled (to below 120 OK), heated (to 1250 OK)and so on. A special type of specimen stage, called agoniometer stage, enables the user to put the specimeninto any required orientation (fig. 19). This is arrangedso that the specimen can be tilted about an axis that

385


cuts the microscope axis, and so that any given detailof the specimen under investigation can be made tocoincide with this point of intersection. A change in theposition of the specimen thus causes no change in focusor magnification, nor any change in effective cameralength for diffraction exposures.Among the many specimen holders that can be

fitted in the goniometer stage there is one that permitsthe specimen to be rotated 3600 in its own plane aboutan axis through its centre. This facility allows a freechoice of the azimuth of the tilt we have just mentioned.When the goniometer stage is used, the pole-pieces

of the objective lens illustrated in fig. 9 have to be re-

placed by others with a wider gap since otherwisethere will not be enough room for the tilting of thespecimen. This has the result of giving the objective agreater focallength (4.1 mm). In order to concentratethe field as much as possible in the space below thespecimen, the pole-pieces are unequal. When the gonio-meter stage is used with the objective lens modified inthis way a resolving power better than 15A can beachieved.The very extensive range of accessories and the good

resolving power make the Philips EM 300 electronmicroscope particularly suitable for many widely di-vergent types of investigation.

Summary. The Philips electron microscope type EM 300 dif-fers from its predecessors in that it has an even better resolvingpower. In normal use this is always better than 5 A and in favour-able circumstances the theoretical resolving power can be closelyapproached. Operation has also been further simplified by in-cluding various vacuum locks, a semi-automatic exposure meterand a "(unction switch", etc., and also by the design ofthe instru-ment as a 'Whole. The image can be displayed on a televisionscreen as well as on the usual fluorescent screen. The magnificationcan be varied from 220 times to 500000 times. Nearly all of the

electron-optical sections can be aligned. The electrical circuits areall transistorized. The vacuum system is fully automatic in opera-tion and has a high pumping speed. The excellent resolving poweris attributable to the very short focal length of the objective lens(1.6 mm), to good mechanical stability and to the careful stabili-zation of lens current and high voltage (the relative variations areless than 2.5 X 10-6 and 5 X 10-6 per minute respectively). There isa wide range of accessories, including a goniometer specimenstage, a cryogenic anti-contamination device and a variety ofspecimen holders.

1968, No. 12 387

Recent scientific publicationsThese publications are contributed by staff of laboratories and plants which formpart of or co-operate with enterprises of the Philips group of companies, particularlyby staff of the following research laboratories:

Philips Research Laboratories, Eindhoven, Netherlands EMullard Research Laboratories, RedhilI (Surrey), England MLaboratoires d'Electronique et de Physique Appliquée, Limeil-Brévannes

(Val-de-Marne), France LPhilips Zentrallaboratorium GmbH, Aachen laboratory, Weisshausstrasse,

51 Aachen, Germany . APhilips Zentrallaboratorium GmbH, Hamburg laboratory, Vogt-Kölln-

Strasse 30, 2 Hamburg-Stellingen, Germany HMBLE Laboratoire de Recherches, 2 avenue Van Becelaere, Brussels 17

(Boitsfort), Belgium. B

Reprints of most of these publications will be available in the near future. Requestsfor reprints should be addressed to the respective laboratories (see the code letter) orto Philips Research Laboratories, Eindhoven,. Netherlands.

G. A. Acket: The influence of doping fluctuations onlimited space-charge accumulation in n-type galliumarsenide.Physics Letters 27A, 293-294, 1968 (No. 5). E

F. R. Badcock & D. R. Lamb (Associated Semiconduc-tor Manufacturers Ltd., Wembley, England): Stabilityand surface charge in the MOS system.Int. J. Electronics 24, 1-9, 1968 (No. I).

J. R. A. Beale: Modern developments in semiconductordevices.Brit. J. appl. Phys. (J. Physics D), ser. 2, 1, 1229-1240,1968 (No. 10). M

P. Billard & C. Hily: Les systèmes optiques à champétendu utilisables dans l'infrarouge.Acta electronica 11, 237-325, 1968 (No. 3). L

G. Blasse: Fluorescence of uranium-activated com-pounds with rocksalt lattice.J. Electrochem. Soc. 115, 738-742, 1968 (No. 7). E

G. Blasse: Crystal structure and fluorescence of com-pounds Ln2Me4+Me6+Os.J. inorg. nuel. Chem. 30, 2091-2099, 1968 (No. 8). E

G. Blasse: Fluorescence of compounds with fresnoite(Ba2TiSi20s) structure.J. inorg. nuel. Chem. 30, 2283-2284, 1968 (No. 8). E

P. F. Bongers, C. F. van Bruggen .(RijksuniversiteitGroningen), J. Koopstra (R.U. Gr.), W. P. F. A. M.Omloo (R.U. Gr.), G. A. Wiegers (R.U. Gr.) & F.Jellinek (R.U. Gr.): Structures and magnetic propertiesof some metal (I) chromium (Ill) sulfides and selenides.J. Phys. Chem. Solids 29, 977-984, 1968 (No. 6). E

G. Bouwhuis: Interferometrie met gaslasers.Ned. T. Natuurk. 34, 225-232, 1968 (No. 8).

A. Bril, W. L. Wanmaker (Philips Lighting Division,Eindhoven) & J. W. ter Vrugt (Philips L.D.): Sr3(P04)z-Tb, a bluish-white emitting cathode-ray phosphor witha long decay time.J. Electrochem. Soc. 115, 776-777, 1968 (No. 7). E

K. Bulthuis: Anomalous penetration of Ga andilnimplanted in silicon.Physics Letters 27A, 193-194, 1968 (No. 4). E

K. H. J. Buschow: Crystal structures, magnetic proper-ties and phase relations of erbium-nickel intermetalliccompounds.J. less-common Met. 16,45-53, 1968 (No. I). E

B. L. Cardozo & R. J. Ritsma (Institute for PerceptionResearch, Eindhoven): On the perception of imperfectperiodicity.IEEE Trans. AU-16, 159-164, 1968 (No. 2). E .

H. B. G. Casimir: On Bose-Einstein condensation.Fundamental problems in statistical mechanics Il,editor E.G.D. Cohen, North-Holland Pub!. Co.,Amsterdam 1968, p. 188-196. E

A. Cohen (Institute for Perception Research, Eindho-ven): Errors of speech and their implication for under-standing the strategy of language users.Z. Phonetik, Sprachwiss. &Komm.f. 21,177-181,1968(No. 1/2).

W. F. Druyvesteyn, C. A. A. J. Greebe, A. J. Smets &R. Ruyg: Faraday rotation of Alfvén waves in liquidsodium.Physics Letters 27A, 301-302, 1968 (No. 5). E

Y. Genin: A note on linear minimum variance estima-tion problems.

E IEEE Trans. AC-13, 103, 1968 (No. I). B


Contents of Electronic Applications 28, No. 3, 1968:

J. Rozenboom: Diac triggering of thyristors and triacs (p. 85-94).F. G. Oude Moleman & B. J. M. Overgoor: Preamplifier with FET input for a wide-band oscilloscope (p. 95-98).P. A. Neeteson: Gateless scalers with J-K flip-flops (p. 99-109).A. P. F. Zegers: Radio interference due to television receivers (p. 110-124).

Contents of Mullard Technical Cornmunications 10, No. 94, 1968:e

T. H. Oxley: Germanium microwave diodes for braad band mixer and low level detector applications (p. 122-132).H. F. Dittrich: Advanced technjques in radio frequency heating generator design (p. 133-134).H. F. Dittrich: Dielectric heater using half-wave line at 30 MHz (p. 135-145).D. E. Nightingale: A 400 kHz induction heater of advanced design for powers up to 60 kW (p. 146-149).D. E. Nightingale: A 300 kHz induction heater ofadvanced design for powers up to 120 kW/240 kW (p. 150"156)..

A. A. van der Giessen, J. G. Rensen & J. S. van Wie-ringen: A study of the constitution and freezing beha-viour of iron oxide-hydrate gels by means of theMössbauer effect.J. inorg. nuel. Chem. 30, 1739-1744, 1968 (No. 7). E

L. Heijne: Ionic conductivity in oxides: experimentalproblems, survey of existing data.Mass transport in oxides, Proc. Symp. Gaithersburg,Maryland, 1967 (NBS Special Publ. No. 296, 1968),p. 149-164. E

H. Jonker, C. J. G. F. Janssen & Th. P. G. W. Thijs-sens: PD-Photoplating. A new photographic additiveprinted circuit process.Photographie und Film in Industrie und Technik, Ber.T. int. Kongress, Köln 1966, p. 249-253; 1968. E

A. van Katwijk (Institute for Perception Research,Eindhoven): A functional grammar of Dutch numbernames.Grammars for number names, editor H. Brandt Cors-tius, D. Reidel Publ. Co., Dordrecht 1968, p. 1-8.

D. R. Lamb & F. R. Badcock (Associated Semicon-ductor Manufacturers Ltd., Wembley, England): Theeffect of ambient, temperature and cooling rate, on thesurface charge at the silicon/silicon dioxide interface.Int. J. Electronics 24, 11-16, 1968 (No. I).

R. Memming & G. Schwandt: E1ectrochemical proper-ties of gallium phosphide in aqueous solutions.Electrochim. Acta 13, 1299-1310, 1968 (No. 6). H

J. Neirynck: Predistortion of high-pass filters for thecompensation of parasitic capacitances.Network and switching theory, editor G. Biorci,Academic Press, New York 1968, p. 273-282. B

A. van Oostrom: Field emission and surface structure.A survey of phenomena in ionized gases, Int. AtomicEnergy Agency, Vienna 1968, p. 291-308. E

G. den Ouden: Internal friction of weld metal.Brit. Welding J. 15, 436-447, 1968 (No. 9). E

P. J. L. Reijnen: Sintering behaviour and microstruc-tures of aluminates and ferrites with spinel structurewith regard todeviation from stoicheiometry.Science of Ceramics 4, 169-188, 1968. E

C. J. M. Rooymans: The effect of very high pressureson ceramic systems.Science of Ceramics 4, 133-149, 1968. E

J. M. Stevels: Permitted and forbidden structural unitsin vitreous systems.Mat. Res. Bull. 3, 599-609, 1968 (No. 7). E

G. W. Tichelaar & J. G. Verhagen: Woher kernmendie Paren beim Schutzgasschweiûen unter Mischgas?Der Praktiker 20, 132, 1968 (No. 6). E

J. Tillack: Gravimetrische Analyse van Halogenidenund Oxidhalogeniden nach der "H-Rohr-Methode".Z. anal. Chemie 239, 81-87, 1968 (No. 2). . A

J. A. Weaver: Machines that read.Science Journal4, No. 10,66-71, 1968.

,..

M

K. Weiss: Die Beweglichkeit von Elektronen in AgBr.Z. phys. Chemie Neue Folge 59, 242-250, 1968 (No.5/6). E

,K. Weiss: Zur Umwandlung der "Tieftemperatur-phasen " des AgJ 'in die a-Modifikation.Z. phys. Chemie Neue Folge 59, 318-320, 1968 (No.5/6). E

R. M. G. Wijnhoven: A simple high-speed magneticaccess switch matrix.[EEE Trans. C-17, 542-550, 1968 (No. 6). E

F. J. Witts & P. W. Whipps: An experimental slow-rate crystal puller for seeded growth from a flux.J. sci. Instr. (J. Physics E), ser. 2, 1, 970-971, 1968(No.9). M

~"

Volume 29, 1968, No. 12 Published 28th Februarypages 353-388

the philip,s. electron microscope em 300 phillps technical review volume 29 the philip,s.electron...

Documents