A simplified method for the isostatic hot pressing of ceramicsK. H. Härdtl

The investigations described in this article pave the way for the cheap mass productionof pore-free ceramic material, which is increasingly in demand for technological appli-cations. Pore-free materials are already used in applications that include envelopes forhigh-pressure sodium lamps, cutting tools, acoustic surface-wave filters, and read andwrite heads for magnetic recording.

Philips tech. Rev. 35, 65-72,1975, No. 2/3

Introduetion

In the applications of ceramic materials in electricaltechnology there is an increasing requirement for mater-ials that are completely free of pores. Some exampleshere are the transparent Ah03 for the envelopes ofsodium lamps, (Pb,La)(Zr,Ti)03 - known as PLZT-which is also transparent and used in electro-opticalapplications, Ab03 for cutting tools, piezoelectricceramics based on Pb(Zr,Ti)03 for surface-wave filters,MoSh for electric heating elements, and ferromagneticceramics (ferrites and garnets) for microwave com-ponents and read and write heads in magnetic record-ing systems.The conventional sintering process (i.e. sintering at

atmospheric pressure) will only yield completely dénseproducts in a few particular cases and then only ifspecial measures such as the addition of dopes are.taken. Examples already known include Ah03, towhich a small quantity of Mg is added, Y203 dopedwith Th, ferrites, and PLZT with an excess of PbO [ll.A process which is commonly used to obtain high

densities is 'hot pressing'. Nowadays two methods ofhot pressing are in use: uniaxial hot pressing, both pieceby piece [2land continuous [3l,with a punch and a dieof refractory material, and isostatie hot pressing, inwhich the starting material, in powder form, is coatedwith a thin metallic layer and an inert gas is used totransfer the pressure [4l. Some disadvantages of theseprocesses are that a corrosive reaction often occursbetween the die (or metallic coating) and the ceramicmaterial to be compressed, and that only pieces withsimple shapes can be manufactured. Moreover, theseprocesses are expensive and not very suitable for massproduction.In this article a method of isostatic hot pressing is

described in which no metallic layer is necessary andwhich does not suffer from the previously mentioned

Dr K. H. Härdtl is with Philips Forschungslaboratorium AachenGmbH, Aachen, West Germany.

,65

disadvantages rei, Different technically important cer-amic materials such as (Pb,La)(Zr,Ti)03, BaTi03,SrTi03, Ab03, ferrites and garnets have been used inthe studies we have made to find the optimum condi-tions for the application of this method.

The new method

In the second method of hot pressing mentionedabove, a thin metallic encapsulating layer is necessaryto prevent the working gas from penetrating into theopen pores of the starting material. Increasing thepressure would not then cause the pores to disappear.If the starting material is not a powder, but a ceramic

product with pores which are not connected to the sur-face (closed pores), it can then be pressed without themetallic layer, since the working gas cannot now pene-trate into the pores.The new procedure is therefore performed in two

steps: firstly, a conventional sintering process, fromwhich an 'intermediate product' with closed pores isobtained, followed by a second sintering at high gaspressure.The conventional sintering process which is used for

ceramic materials in the electronics industry, such as

[1) For Ab03 doped with Mg see: R. L. Coble, J. appl. Phys,32, 787 and 793, 1961; for Y203 doped with Th seeR. C. Anderson, U.S. Patent 3,545,987, 8 December 1970;for ferrites see A. L. Stuijts, Proc. Brit. Ceram. Soc. 2, 73,1964; for PLZT with an excess of PbO see G. S. Snow,J. Amer. Ceram. Soc. 56, 91, 1973.

[2) See for example: P. Murray, D. T. Livey and J. Williams,The hot pressing of ceramics, in: W. D. Kingery (ed.),Ceramic fabrication processes, Technology Press M.LT.,Cambridge, Mass., U.S.A., 1958, pp. 147-171.

[3) See for example: G. J. Oudemans, Philips tech. Rev. 29, 45,1968.

[4) See for example: H. Bumm, F. Thümmler and P. Weimar,Ber. Dtsch. Keram. Ges. 45, 406, 1968.

[5) Similar work, parallel to ours but independent, has beencarried out by two other investigators: see E. A. Bush, U.S.Patent 3,562,371, 9 February 1971, and S. E. Isaksson,British Patent 1,300,864, 20 December 1972.

66 K. H. HÄRDTL Philips tech. Rev. 35, No. 2/3

perovskites, ferrites or garnets, normally yields prod-ucts with a density of 95 to 99 % of the theoreticalvalue. The pores of ceramics with a density of morethan 93% are generally closed. This is in agreementwith a theoretical result according to which the poresin all ceramic materials are closed when the totalporosity is reduced to 9% [6J.

To illustrate this point, fig. 1 shows a photomicro-graph of a wafer of (Pb,La)(Zr,Ti)03 ceramic, whichhas been sintered conventionally for four hours in anoxygen atmosphere at a température of 1170 "C. Theresidual porosity lies between 1 and 2%. The grainboundaries and also the residual pores are clearlyvisible after thermal etching of the polished surface. Itcan te seen that the pores are situated exclusively onthe grain boundaries.

At first sight one might suppose that the dark patchesare not pores but holes resulting from grains breakingloose from the surface during the polishing. However,evidence to the contrary is supplied in fig. 2. These twophotographs show the sarne segment of a conven-tionally sintered wafer of PLZT, which has beenpolished on both sides to make it translucent. Whilethe photograph in fig. 2a has been taken with reflectedlight, the one in fig. 2b has been made using a combina-tion of reflected and transmitted light. Only surfacepores are revealed in fig. 2a, all of which can be foundin fig. 2b (two pores are ringed). In fig. 2b many otherpores which lie deeper in the ceramic material are madevisible by the transmitted light. It can be establishedthat the concentration and dimensions of these deeperpores agree with those of the pores on the surfaceshown in fig. 2a.lt can also be clearly seen in fig. I that the residual

pores are not directly connected to the surface. We aretherefore dealing with closed pores.

A second condition which must be satisfied by ourmethod of isostatic hot pressing is that, at the tempera-ture at which hot pressing occurs, the gas diffusionvelocity in the ceramic material must be negligiblysmall, so that no noticeable quantity of gas can pen-etrate the pores. In the literature there are no quanti-tative data on the diffusion constants ofthe usual work-ing gases (N2, noble gases, CO2 etc.) in the ceramicmaterials which are considered here. Research has how-ever shown that, in the perovskite materials which wehave investigated, the diffusion velocity of nitrogen atthe sintering temperature is many orders of magnitudelower than that of oxygen [7J. The absolute magnitudeof the first diffusion velocity is unknown but it appearsto be low enough for our purposes. The same is alsovalid for argon.

In a later phase of our investigations we were able toconfirm that the diffusion velocity of nitrogen in the

Fig. 1. Photomicrograph of a wafer of (Pb,La)(Zr,Ti)03 ceramic(molecular proportion of La/Zr/Ti : 1 J /65/35) which has beensintered conventionally for four hours in an atmosphere of oxygenat a temperature of J J 70 oe. The sample has been polished andthen thermally etched (1050 oe, one hour in air). The residualporosity is between I and 2 %. It can clearly be seen that theresidual pores (darker) are not linked with each other or thesurface. Magnification 800 x.

WJ

(I ..e

,.

®

• -" 4J

I

• a

b

Fig. 2. Photomicrograph in (a) reflected and (b) transmitted lightof the same wafer of PLZT (lJ/65/35), which has been polishedon both sides. The pores on the surface (two of which have beencircled) are visible in both exposures; in (b), moreover, manymore which are situated in the interior of the plate can be seen.Magnification 400 X .

Philips tech. Rev. 35, No. 2/3 ISOSTATIC HOT PRESSING 67

material which we have studied is indeel rather low.This explains not only why nitrogen is a good workinggas, but also why, in some applications, nitrogen shouldnot be present in the atmosphere in which the firstsintering process (leading to closed pores) takes place.This sintering step must not therefore be performed inair, but in oxygen, for example. We shall come back tothis later at the end of the article.

A diagram of the apparatus which we have used forour method of hot pressing is shown in jig. 3. The mostimportant components are a water-cooled pressurevessel, which can withstand a pressure of 200 bars, andan electrical furnace with platinum heating elementssuitable for a maximum temperature of 1400 oe.

The following procedure is used: the conventionallysintered material is placed in the furnace at an atmos-phere of air. After closing the pressure vessel, the work-inggas (nitrogen or argon) is pumped in until a pressureof approximately 60 % of the final value is obtained;the furnace is then brought to the desired temperaturein about 30 minutes. Owing to the heating, the pressurein the vessel rises to the desired value, at which it ismaintained by means of a regulating valve. The cooling-down period is approximately 30 minutes.

Because the pressure vessel is closed in air, there isalways a partial pressure of oxygen present (0.3 to Ibar), depending on the temperature during pressing.The presence of oxygen prevents a possible reduc-tion of the material at the high temperature duringhot pressing. Should it be necessary, in certain cir-cumstances, to have a better control of the oxygenpressure than is described here, a solution is readilyfound.

Preliminary experiments

(Pb,La)(Zr,Ti)03 ceramic

For certain applications Pb(Zr,Ti)03 ceramic is usedin which a proportion of the Pb2+ ions are replaced byLa3+ ions. This increases the dielectric constant, thecoupling factor and the transparency. These changes inthe properties are connected with the defect structurecaused by the substitution of divalent ions by trivalentions. The surplus charge of the trivalent ions is corn-pensated by vacancies in the Pb or the (Zr,Ti)sublattice.

The preparation of this type of ceramic is impairedby the tendency of the volatile PbO to escape during thesintering process. A simple way of reducing the loss ofPbO is to add an extra quantity of PbO to the startingmaterial and to maintain a high PbO partial pressureduring the sintering. It has been shown that a corn-paratively high density is obtained after the first sinter-ing step when these methods are used [11.

In our first experiments, which were carried out with(Pb,La)(Zi,Ti)03 ceramic (PLZT), the starting mater-ials PbO, Ti02 and La203 were weighed out in therequired proportions and carefully mixed by dry-milling. We selected the appropriate proportions in

c

QV

Fig. 3. Schematic cross-section of our apparatus for isostatie hotpressing. V water-cooled pressure vessel with sealing clamps Cand gasket G. I heat insulation. F electric furnace. S ceramicsample. T thermocouple. L supply leads for the heating elements.I V gas inlet valve. M manometer. 0 V gas outlet valve.

accordance with the formula Pbl-ayLayZrxTil-x03, inwhich the parameter ex indicates the amount to whichlead had been substituted [81. The mixed powders werepre-sintered for 10 hours at 800 oe, and then carefullydry-milled and compressed isostatically (4 kbars) with-out a binding agent, to a prismatic or cylindrical shape.The conventional sintering took place in an atmos-phere of oxygen at temperatures between 1100 and1300 De. PbZr03 was added in order to maintain ahigh partial pressure of PbO. A value of the parameterct in the end product of 1.2 to 1.3 was obtained from theweight loss during sintering due to the loss of PbO.

The conventionally sintered intermediate productsexhibit a residual porosity of 1 to 0.1 %. Since a lowporosity had already been obtained at this stage, wewere faced with the difficulty of how to determine thefurther reduction which can be achieved with ourmethod of isostatic hot pressing. We could not useconventional methods to measure the increase in thedensity, such as immersion in water, or the determina-tion of the change in dimensions, since these methodsare not sufficiently accurate. With such low porosities,information concerning the elimination of pores canonly be obtained using optical methods. The best in-formation, even if it is only qualitative, is obtained fromphotomicrographs, taken with either reflected or trans-mitted light. The latter method is only possible if thin

[6] D. W. Budworth, Trans. Brit. Ceram. Soc. 69,29, 1970.[7] T. F. Murray and R. H. Dungan, Ceram. Ind. 83, No. 6, 74,

1964.[S] D. Hennings and K. H. Härdtl, Phys, Stat. sol. (a) 3, 465,

1970.K. H. Härdt1 and D. Hennings, J. Amer. Ceram. Soc. SS, 230,1972.

68 K. H. HÄRDTL Philips tech. Rev. 35, No. 2/3

Fig. 4. Photomicrograph in reflected light of a wafer of the samematerial as in fig. I after isostatic hot pressing for two hours at1170 °C in nitrogen at a pressure of 200 bars. The wafer has beenthermally etched and polished. Magnification 800 x.

wafers of a material which does not strongly absorbare used.

Fig. I shows the result of conventional sintering ofa sample of PLZT (x = 0.65, y = 0.11), photographedwith reflected light. Tn jig. 4 a sample of the samematerial is shown after isostatic hot pressing by themethod described here (2 hours at 1170 oe in N2 at apressure of 200 bars). Tt can be seen that the pores havedisappeared, while the grain size has not changedduring the treatment.

We have also investigated the densification of thismaterial with transmitted light. Fig. 5a shows the resultof conventional sintering and in fig. 5b and c the effectof conventional sintering followed by isostatic hotpressing using our method (respectively after 2 andafter 16 hours at 1170 oe in N2 at a pressure of 200bars) is seen.

The majority of the pores had already disappeared

after 2 hours and the porosity was almost completelyeliminated after 16 hours. We do not know whether thesmall dark specks which are visible in fig. 5c representremaining pores, or whether we are concerned herewith a separated second phase (possibly PbO). Electronmicroprobe studies have left this question unanswered.

The results of similar experiments with related mater-ials are reproduced in Tab/e l. The tabulated theoret-ical densities have been calculated from the molecularweight obtained from the composition, the number ofmolecules per unit cell and the unit-cell dimensions, asdetermined by X-ray diffraction. Since the distributionof the vacancies at Pb or (Zr,Ti) sites is not known withcertainty in PLZT [81, the molecular weight, and there-fore the theoretical density, cannot be accurately given.It is apparent from the table that our method of iso-static hot pressing leads to the theoretical density forall the materials studied, as is confirmed by photo-micrographs.

In order to obtain an impression of the increase intransparency which is obtained with our method, wehave measured the attenuation coefficient for lightwith a He-Ne laser; for PLZT with x = 0.65 andy = 0.11 we found a value of 4.5 cm-I This meansthat in a ceramic wafer with a thickness of 100 l1.m, asis used in electro-optical applications, the light loss dueto absorption and scattering is smaller than 5%.

The influence of pressure, temperature and sin teringtime

BaTi03 and SrTi03 ceramics

In order to obtain a better idea of the effect of pres-sure, temperature and duration of our method of hotpressing on the shrinkage behaviour ofcerarnic objects,we have also studied the ceramic materials barium

Table J. Experirnental data obtained with the isostatic hot-pressing method without an encapsulating metallic layer, carried out withPbl-yLayZrx Til-x03 and other materials with a perovskite structure.

Pressure Time Tempera- Initial Final TheoreticalComposition ture density density density

(bars) (1m) (OC) (g/cm'') (g/cm3) (g/em3)

Pbl-yLayZr xTit-x03x = 0.65 y = 0.025 200 0.5 1250 7.88 7.95 7.91-7.95

x = 0.65 y = 0.06 200 2 1150 7.73 7.86 7.83-7.86

x = 0.65 y = 0.075 200 0.5 1250 7.68 7.79 7.78-7.81

x = 0.65 y = 0.09 200 2 1170 7.77 7.80 7.75-7.79

x = 0.65 y = 0.11 200 0.67 1170 7.63 7.72 7.70-7.74

x = 0.65 y = 0.14 200 0.5 1220 7.57 7.68 7.64-7.69

x = 0.58 y = 0.075 200 0.5 1250 7.65 7.78 7.78-7.81

x = 0.62 y = 0.075 200 0.5 1250 7.69 7.79 7.78-7.81

x = 0.72 y = 0.075 200 0.5 1250 7.67 7.79 7.78-7.81

x = 0.76 y = 0.075 200 0.5 1250 7.67 7.77 7.78-7.81

x = 0.80 y = 0.075 200 0.5 1250 7.63 7.79 7.78-7.81

Pb(Tio.5ZrO.5)O.9B5Feo.01503 200

2 1250 7.98 8.01 8.01PbTio.42Zro.5B03 + 1 mol % MnO 200

2 1250 7.95 8.018.01

Philips tech. Rev. 35, No. 2/3 ISOSTATIC HOT PRESSING 69

a

..b

"

c

Fig. 5. Photomicrograph of a sample of PLZT (11/65/35), takenwith transmitted light. a) Initial condition after conventionalsintering. b) After hot pressing by our method (two hours,1170 °C, in N2 at 200 bars). c) After hot pressing under the sameconditions, but now for 16 hours. Magnification 1SOX.

titanate and strontium titanate. These materials arechosen rather than PLZT for two reasons. Firstly,there is no risk during the hot pressing of a volatilecomponent evaporating, as is the case to a certainextent with PbO in PLZT. Secondly, the theoreticaldensity of BaTi03 and SrTi03 is known to a muchgreater accuracy than that of PLZT.

In the case of BaTi03 we used Bae03 and Ti02 asthe starting materials, weighing them out in accord-

• ••..

'..,.. .,

•

ot'

'",

,.'.•

Fig. 6. Photomicrograph of a conventionally sintered wafer ofBaTi03 ceramic. Magnification 1SOX .

ance with the formula BaTil.Oo503. The small excess oftitanium is necessary to ensure good sintering proper-ties. The powders were mixed in a dry state, pre-sintered for 10 hours at 1100 oe and, after dry-milling,were sintered for 4 hours at 1315 oe in oxygen at apressure of 1 atmosphere. For SrTi03 ceramic thestarting material was cornrnercially available SrTi03powder containing an excess of 1.6 mol % Ti02. Thispowder was carefully dry-milled and then sintered for8 hours at 1340 oe in 02 at a pressure of 1 bar.

After conventional sintering we obtained densities ofbetween 5.83 and 5.86 gjcm3 for BaTi03 ceramic andbetween 5.02 and 5.05 g/cm'' for SrTi03 ceramic. As-suming a perfect crystallattice, we have calculated fromthe lattice parameters that the theoretical density ofBaTi03 is 5.996 g/cm ' and that ofSrTi03 is 5.117 g/cm '.In conventionally sintered samples of BaTi03 andSrTi03 ceramics this corresponds to a residual porosityof approximately 2.5 % and 1.5% respectively, whichis confirmed in the photornicrcgraphs (fig. 6). Ourmethod of isostatic hot pressing has been carried out attem peratures of 1100, 1150, 1200, 1250 and 1300 oe ata nitrogen pressure of 25, 50, 100, 150 and 200 bars.In order to follow the densification process we havemeasured the linear shrinkages of a prism-shapedsample (5x5xI7mm) in its longest direction. Thechange in density is then calculated from the equation:

d.ld, = l- 35 + 352 - 53,

where di and dr are respectively the initial and finaldensities. Since the initial densities amounted to 95%of the theoretical values and the shrinkage was of theorder of 0.0 I, the second- and third-order terms in theabove equation could be ignored.

Fig. 7 and 8 clearly show how the density of BaTi03and SrTi03 ceramic samples asymptotically ap-proaches the theoretical density, as a function of time,temperature and pressure. The samples have become

70 K. H. HÄRDTL

Fig.7. Shrinkage behaviour (density d and relative density dl'plotted against time t) of conventionally sintered BaTi03 whichhas been hot pressed by our method at 1100 -c (a), I] 50°C (b),1200 "C (c) and 1250 "C (d) in N2 with a pressure which variesfrom 50 to 200 bars. The samples have been periodically cooledin order to carry out shrinkage measurements. The theoreticaldensity (dr = lOO %) has been indicated by a dashed line; thepoints plotted on the ordinate are the initial densities (afterconventional sintering).

com pletely dense at temperatures 100 oe below theconventional sintering temperature and at a pressure of200 bars. If the temperature during hot pressing ap-proaches the conventional sintering temperature, com-plete densification can be achieved after only tenminutes.

The densities which are measured after 24 hours ofhot pressing agree well with those determined byshrinkage measurements. Photomicrographs show thatsamples which have attained the theoretical densityasyrnptotically have become pore-free. This may beseen in fig. 9, for example, in which the sarne sampleas in fig. 6 is shown, but after isostatic hot pressing for16 hours at 1200 oe in N2 at 200 bars.

In general we can say that at the pressure underconsideration, doubling the pressure has approx-imately the same effect as increasing the temperatureby 50°C. It is interesting to note that the increase indensity varies logarithmically as a function of time atlow pressures and ternperatures.

Fig. 9. Photomicrograph of the BaTi03 ceramic of fig. 6, afterhot pressing for 16 hours at 1200 oe in a nitrogen pressure of200 bars. Magnification 150 x .

[9] AI203 ceramic has also been studied by E. A. Bush [5]. Hehad to use considerably higher values of the process param-eters (1650 oe, 1000 bars and 1 hour) than we did in order toobtain a dense Ah03 ceramic. We think that the fact that wewere able to use much lower values is connected with ouraddition of a small quantity of MgO to the Ah03 (the defectsthus introduced increase the sintering velocity).

[10] 'Cermets' are combinations ofcerarnics and metals.

Philips tech. Rev. 35, No. 2/3

-- -- - --- -----100%

d

t;;

200 bar _99.5

~150

/ -990

/ / .,,-x 1005.90 / / x/x"'-

o /+ x/c-:/x+5~ ~ ~~~. r liDO oe 975L :-__-:';;--__~;;_;_'

o 10 100h

98.5

- 98.0

g

-f

100%

d;995

- 99.0

98.5

- 98.0

BaTi0311500e - 97.5

-f

6.o0g/cm3 r:: 0 ?-9Q_bgr 100 %- .,,-+-+- 150+d

o /+ a;-995

/+ x-----x1005.95 //x 99.0+

x/x 98.55.90 _. 50__ .

./ 98.0_//

585t

__.....,.Ba Ti0312000e 975

f_0 10 100h

-f

6.o0g/cm3 ~:ggb" 100%

d d;

/ 99.5

5.95 50

x ./

99.0

98.55.90

./ - 98.0

5.85 BaTi03 97512500eQ [_

0 10 100h-f

Philips tech. Rev. 35, No. 2/3 ISOSTATIC HOT PRESSING

----------- 100%

S.IOg/cm3

d

f

...--- 200 bar 99.5

/+ 150+/ 99.0

5.05 . .-l_.x~x 100

t'e-:::-- •• :;

5.00I1S0oC

Q0-- 10 100h

-t

--- -_ - --;-- 200 bar 100%

S.IOg/cm3~150 - 99.5d

/ /" ~XIOOdr

f 0/ 99.0 t5.05 :--x-.--" 50f~ 98.5

5.00 _.-98.0

SrTi0312000C

!l.0 10 100h

-t

~_200bar 100%--;:-:::+ 150

5.IOg/cm3 ...-:::;..-r-- 100~_X. 5099.5

d dr

t S.05x.~99.0 f

500198.5

98.0SrTi0312500C

£0 10 100h

-t

200bar-=~i_.;-a..xISO - 100%r-__.-' 100

S.lOg/cm3./ _ 50r .,.,....... 99.5d /. dr

t 5.05 _/.. •.............__. 25 99.0 t

. 5001.---o

98.5

98.0

10 100h-t

Table n. Experimental data obtained with our method of isostatic hot pressing, carried out with non-perovskite material.

Fig. 8. Shrinkage behaviour of conventionally sintered SrTi03ceramic which has beenhot pressed at 1150 °C (a), 1200 °C (b),1250 °C (c) and 1300 °C (d) in N2 at a pressure which varies from25 to 200 bars.

Other materials

Our method of isostatic hot pressing can be success-fully applied to other ceramic materials. We have usedcommercially available ceramic, materials such asAhOa [91,YaFe5012 (YIG), Mn-Zn ferrite and Ni-Znferrite. The results are presented in Table IL S. E. Isaks-son has also reported good results with 'hard metals'such as WC and TaC (with Co, Ni or Fe as binder),with 'cermets' [lOl based on AhOa-Ag and with MoSi2and V02 [51.Difficulties or failures with the applicationof our method have only arisen with coarse-grainedBaTiOa and with Pb(Zr,Ti)Oa ceramic in which tracesof a liquid second phase appeared: at the sinteringtemperature. Studies are in progress to obtain aninsight into the nature of these difficulties.

The gas atmosphere at the first sintering step

As stated earlier, the first sintering step which isused to obtain closed pores 'should preferably be car-ried out in oxygen and not in air. We came to thisconclusion when we tried to find out whether thediffusion velocity of nitrogen is indeed much lowerthan that of oxygen. We subjected two samples of thesame PLZT ceramic to similar treatments, the onlydifference being that the first sintering step in oneexperiment took place in' nitrogen and in the otherexperiment in oxygen. We assumed that in both casesthe residual pores would contain the gas from theatmosphere In which the sintering occurred. The gaspressure in such pores enclosed by grains is not knownwith certainty, but probably lies somewhere between1 and lObars; this pressure is approximately equal to2 vlr, where r is the surface tension and r is the radius'of the pores. During isostatic hot pressing these pores

Pressure Time Tempe ra- Initial Final TheoreticalComposition ture density density density

(bars) (hrs) eC) (g/cm3) (g/cm3) (g/cm'')

Ai203 + 0.3 wt % MgO 200 22 1300 3.93 3.98 3.99Y3Fes012 (garnet) 200 4 1300 4.82 5.16 ,5.17Ni-Zn-ferrite 200 4 1200 5.31 5.33 5.33Mn-Zn-ferrite 200 4.5 1200 4.94 5.10 5.12

, . .

71

72 ISOSTATIC HOT PRESSING Philips tech. Rev. 35, No. 2/3

Table Ill. Shrinkage and expansion behaviour of samples of PLZT (x = 0.65,y = 0.11) wherethe first sintering process has been carried out in either an oxygen or a nitrogen atmosphere.

di initial density;dr density after two hours of hot pressing at 1170 °C in argon at a pressure of 200 bars;/),.{ shrinkage during the hot pressing. The originallength was 16.6 mm;Ma increase in length after two hours annealing at 1170 °C in air at a pressure of 1 bar;d« density after annealing.

di dr

I/),.{

IMa da

(g/cm3) . (g/cm3) (um) (urn) (g/em3)

In 02 (4 hrs, 1170 0c) 7.60 7.77 116 I 0 7.77

In N2, with a partial pressure of 7.52 I 7.76 179 I 47 I 7.67oxygen of 10-4 bar (4 hrs, 1170 "C)

will shrink to such a small volume that the internalpressure will become equal to the sum of the externalpressure and the pressure due to surface tension. Itwillmake little difference to the shrinkage behaviour ofthese samples whether they are sintered in N2 or 02.However, if the samples are heated again for 2 hours atatmospheric pressure to the sintering temperature, wemay expect that, in the first case, the pores will swell,resulting in a decrease in density, whilst in the second

case the oxygen will just escape because of its higherdiffusion velocity and no swelling of the pores willoccur. As Table IJl shows, the experimental resultscompletely agree with these expectations.For certain applications, such as envelopes for lamps,

which reach a high temperature, it goes without sayingthat such a swelling of the pores must be avoided. Inthese cases the first sintering step must be performed ina gas such as oxygen, and in any case not in air. Thefirst sintering step for Ah03 used for the envelope ofsodium lamps is performed in hydrogen, which alsodiffuses very rapidly.The advantages of our method of isostatic hot press-

ing in comparison with earlier methods can be sum-

marized as follows .. The technique is simpler andappears to lend itself to mass production. Because nodie or metallic encapsulating layer is necessary, thereare no problems of corrosive reactions between thematerials from which these are constructed and theceramic. The ceramic samples can easily be pressedinto any desired shape. Comparatively low pressures(50-200 bars) and temperatures (lOO-200°C below theconventional sintering temperatures) are sufficient.

Summary. In earlier methods for isostatie hot pressing in themanufacture of ceramics a metallic layer has always been used inorder to avoid the penetration of the working gas into the startingmaterial. In the new process described here the starting materialin powder form is first sintered at atmospheric pressure to aceramic body with closed pores, after which hot pressing withoutthe encapsulating metallic layer can be carried out. Suitableworking gases are nitrogen or argon, both of which have asufficiently low diffusion velocity. The sin tering at atmosphericpressure should preferably be performed in oxygen. The newmethod is especially simple to carry out and is suitable for themass production of ceramic components of any desired shape.The process requires a relatively low pressure (50-200 bars) andtemperature (approximately 100°C below the conventional sin-tering temperature). Because no die or encapsulating layer isused, problems relating to the contact between the ceramic andthe materials from which the die or encapsulating layer are madedo not arise. When these methods are used, the ceramic mater-ials Pb(Zr,Ti)03, (Pb,La)(Zr,Ti)03, BaTi03, SrTi03, Ah03,Y3Fes012, Mn-Zn ferrite and Ni-Zn ferrite can be densified toalmost the theoretical density.

Recent scientific publications

Philips tech. Rev. 35, No. 2/3 73

These publications are contributed by staff of laboratories and plants which form part ofor cooperate with enterprises of the Philips group of companies, particularly by staff of •the following research laboratories:

Philips Research Laboratories, Eindhoven, Netherlands EMullard Research Laboratories, RedhilI (Surrey), England MLaboratoires d'Electronique et de Physique Appliquée, 3 avenue Descartes,94450 Lirneil-Brévannes, France L

Philips Forschungslaboratorium Aachen GmbH, WeiBhausstraBe, 51Aachen,Germany A

Philips Forschungslaboratorium Hamburg GmbH, Vogt-Kölln-StraBe 30,2000 Hamburg 54, Germany H

MBLE Laboratoire de Recherches, 2 avenue Van Becelaere, 1170 Brussels(Boitsfort), Belgium B

Philips Laboratories, 345 Scarborough Road, Briarcliff Manor, N.Y. 10510,U.S.A. (by contract with the North American Philips Corp.) N

Reprints of most of these publications will be available in the near future. Requests forreprints should be addressed to the respective laboratories (see the code letter) or to PhilipsResearch Laboratories, Eindhoven, Netherlands.

D. Bäuerle: Optical phonons in small crystals of ABOaperovskites.Phys. Stat. so!. (b) 63,177-182,1974 (No. I). A

J. W. M. Biesterbos: Structure and de electrical prop-erties of a Au-Rh-glass thick-film system.J. app!. Phys. 45, 153-160, 1974 (No. I). E

S. J. Bless: Drop transfer in short-circuit welding.J. Physics D 7, 526-539, 1974 (No. 4). E

P. Blood, G. Dearnaley (AERE, HarweIl) & M. A.Wilkins (AERE, HarweIl) : The depth distribution ofphosphorus ions implanted into silicon crystals.Radiation Effects 21, 245-251, 1974 (No. 4). (Alsopublished in Ion implantation in semiconductors andother materials, editor B. L. Crowder, Plenum, NewYork 1974, pp. 75-85.) M

P. F. Bongers & G. W. Rathenau: E. W. Gorter(1912-1972). Aspects of some fields of magnetism towhich he has contributed.Int. J. Magnetism 4, 282-289, 1973 (No. 4). E

J. Borne, R. Goudin, G. J. Naaijer & J. Ratovonar:Console de visualisation et de dialogue pour enseigne-ment assisté par ordinateur.Acta Electronica 17, 329-342, 1974 (No. 3). L

H. Bouma & Á. H. de Voogd (Institute for PerceptionResearch, Eindhoven): On the control of eye saccadesin reading.Vision Res. 14, 273-284, 1974 (No. 4).

P. Branquart, J. P. Cardinael & J. Lewi: Optimizedtranslation process, application to ALGOL 68.Proc. Int. Computing Symp. 1973,Davos, pp. 101-107;1974. B

J. W. Broer: Comment op. the paper 'Sysrematic tech-niques for solving nonlinear communications'. •IEEE Trans. PC-16, 215, 1973 (No. 4). E

J. Brokken-Zijp & H. van de Bogaert: CIDNP evidencefor radical intermediates in the thermal decompositionof Z-alkylaryldiazo sulphides.Tetrahedron Letters 1974, 249-252·(No. 3). E

H. H. Brongersma: Surface structure analysis by ionscattering.J. Vac. Sci. Techno!. 11, 231-235, 197~ (No. I). . E

E. Bruninx: The accurate deterrnination of Ga inGaxSey by means of neutron activation: a comparisonwith X-ray fluorescence and wet chemical methods.J. radioana!. Chem. 19, 47-53, 1974 (No. I). E

K. H. J. Buschow, Á. M. van Diepen & H. W. de Wijn(State University of Utrecht): Moment reduction inmagnetically ordered samarium intermetallics.Phys. Rev. B 8, 5134-5138, 1973 (No. 11). E

V. Chalmeton: Intensificateur à galette de microcanauxpour neutronographie.La neutrographie, pub!. Kodak-Pathé, Paris 1973,pp.48-50. L

E. Cuppens & C. Ducot: Future education and newtechnical means.Acta Electronica 17, 241-270, 1974 (No. 3). (Also inFrench.) E, L

N. H. Dekkers: A universal deflection unit generatinga field of any order and azimuth.J. Physics D 7, 805-814, 1974 (No. 6). E

J. P. Deschamps: Asynchronous automata and asyn-chronous languages.Information and Control24, 122-143, 1974 (No. 2). B

P. Á. Devijver: On a new class ofbounds on Bayes riskin multihypothesis pattern recognition.IEEE Trans. C-23, 70-80, 1974 (No. I). B

74 RECENT SCIENTIFIC PUBLICATIONS Philips tech. Rev. 35, No. 2/3

H; J. A. van Dijk,' C. M. G. Jochem, G. J. Scholl &P. van der Werf: Diameter control of LEC grownGaP crystals.J. Crystal Growth 21, 310-312, 19,74 (No. 2). E

J. Donjon, G. Marie, B. Monod & R. Le Pape: Utilisa-tion du tube TITUS pour la projection d'images detélévision sur grand écran.Onde électr. 54, 121-127, 1974 (No. 3). L

C. Z. van Doorn & J. L. A. M. Heidens: Angular de-pendent optical transmission of twisted nematic liquid-crystallayers.Physics Letters 47A, 135-136, 1974 (No. 2). E

D. den Engelsen & G. Frens: Ellipsometric investigationof black soap films.J. Chem. Soc., Faraday Trans. 170, 237-248, 1974(No.2). E

L. J. M. Esser: The peristaltic charge coupled device.Proc. Charge Coupled Device Appl. Conf., San Diego1973, pp. 269-277. E

L. J. M. Esser: The peristaltic charge-coupled devicefor high-speed charge transfer.1974 IEEE Int. Solid-State Circuits Conf. Digest tech.Papers, pp. 28-29 & 219. . E

E. Fabre & R. N. Bhargava: Thermally stimulated cur-rent measurements and their correlation with efficiencyand degradation in GaP LED's.Appl. Phys. Letters 24, 322-324, 1974 (No. 7). L, N

A. J. Fox: The grating guide - a component for inte-grated optics.Proc. IEEE 62, 644-645, 1974 (No. 5). M

G. Goutelle & M. Martin: Dispositifs d'apprentissagepour processus opératoires linéaires.Acta Electronica 17, 283-296, 1974 (No. 3). L

W. S. C. Gurney & J. W. Orton: New techniques forthe study of Gunn diode contacts.Solid-State Electronics 17, 743-750, 1974 (No. 7). M

J. Haisma, G. Bartels & W. Tolksdorf: Preparation ofthe magnetic-bubble materials Y3-xSmxFes-yGay012and Y3-xLaxFes-yGay012 by liquid-phase epitaxy, andtheir relevant physical properties.Philips Res. Repts. 29, 493-516, 1974 (No. 6). E, H

J. Haisma, H. J. Prins & K. L. L. van Mierloo: Mag-neto-optic determination of the refractive indices offerric borate.J. Physics D 7, 162-168, 1974 (No. I). E

J. C. M. Henning & J. H. den Boef: Mechanisms offlexural-strain-modulated electron spin resonance.Solid State Corrïm. 14, 993-996, 1974 (No. 10).' E

F. Heutink: Widerange variable-pulse generator ..Electronics Letters 10, 57-58, 1974 (No. 5). E

E. E. Hoefer, H. Grimmert & B. KiesIich: Computer-controlled synthesis of tomograms by means of a TVstorage tube.IEEE Trans. BME-21, 243-244, 1974 (No. 3). H

W. K. Hofker, H. W. Werner, D. P. Oosthoek & H. A.M. de Grefte: Experimental analysis of concentratienprofiles of boron implanted in silicon.Ion implantation in semiconductors and other materi-als, editor B. L. Crowder, Plenum, New York 1974,pp. 133-145. " E

B. B. van Iperen: Efficiency limitation by transverseinstability in Si IMPATT diodes.Proc. IEEE 62, 284-285, 1974 (No. 2). E

R. N. Jackson & K. E. Johnson: Gas discharge displays:a critical review.Adv. in Electronics & Electron Phys. 35, 191-267,1974. MW. H. de Jeu, Th. W. Lathouwers & P. Bordewijk(University of Leiden): Dielectric properties ofdi-n-heptyl azoxybenzene in the nematic and in thesmectic-A phases.Phys. Rev. Letters 32, 40-43, 1974 (No. 2). E

H. Klamet: Computer-aided design of the layout ofintegrated circuits (CADLIC).Proc. Int. Computing Symp. 1973, Davos, pp. 451-458;1974. }{

A. J. R. de Kock, P. J. Severin & P. J. Roksnoer: Onthe relation between growth striations and resistivityvariations in silicon crystals.Phys. Stat. sol. (a) 22, 163-166, 1974 (No. I). E

G. Krekow & J. Schramm: Electrostatic gray-scale fac-simile recording.IEEE Trans. ED-21, 189-192, 1974 (No. 3). }{

F. A. Kuijpers & B. O. Loopstra (Reactor CentrumNederland, Petten): A neutron-diffraction study on thestructural relationships of RCos hydrides.J. Phys. Chem. Solids 35, 301-306, 1974 (No. 3). E

A. van Lamsweerde: Deadlock prevention in real timesystems.Proc. Int. Computing Symp. 1973, Davos, pp. 135-142;1974. BS. Leblanc: Une expérience de recyclage dans l'industrieau moyen de l'enseignement assisté par ordinateur.Acta Electronica 17, 297-318, 1974 (No. 3). L

D. B. Lee (Mullard Southampton Works): The push-out effect in silicon n-p-n diffused transistors.Thesis, Southampton 1973. (Philips Res. Repts. Suppl.1974, No. 5.)

J. Lemmrich: Winkelregelungen von Gleichstrom-maschinen mit frequenzanaloger Signalverarbeitung.Thesis, Darmstadt 1974. (Philips Res. Repts. Suppl.1974. No. 4.) }{

J. Lohstroh: The JFET as a photosensitive cell in imagesensor arrays.1974 IEEE Int. Solid-State Circuits Conf. Digest tech.Papers, pp. 34-35 & 221. E

M. H. van Maaren, H. J. van Daal, K. H. J. Buschow& C. J. Schinkel (University of Amsterdam): Highelectronic specific heat of some cubic UX3 intermetaIIiccompounds.Solid State Comm. 14, 145-147, 1974 (No. 2). E

.-----~----------~-~-- ~-.---- ,-

Philips tech. Rev. 35, No. 2/3 RECENT SCIENTIFIC PUBLICATIONS 75

J. Magarshackt On synchronizing signatures andoscillator optimization.Microwave J. 17, No. 3, pp. 70 & 87, March 1974. L

H. H. van Mal, K. H. J. Buschow & A. R. Miedema:Hydrogen absorption in LaNi5 and related compounds:experimental observations and their explanation.J. less-common Met. 35, 65-76, 1974 (No. I). E

G. Marie: Large area display materials and devices.Electronic materials, editors N. B. Hannay & U.Colombo, Plenum Press, New York 1973,pp. 317-369.

L

D. Meyer-Ebrecht: Schnelle Amplitudenregelung har-monischer Oszillatoren.Thesis, Braunschweig 1974. (Philips Res. Repts. Suppl.1974,No. 6.) H

A. R. Miedema: Superconductivity in transition metalalloys: an alternative to the rigid band modelII.J. Physics F 4, 120-135, 1974 (No. 1). E

J. Monin (Conservatoire National des Arts et Métiers,Paris) & G.-A. Boutry: Optical and photoelectric prop-erties of alkali metals. IPhys. Rev. B 9, 1309-1327, 1974 (No. 4). L

K. Mouthaan & R. M. Snoeren: Characterisation ofavalanche photodiodes having nearly-unilateral-carriermultiplication.Electronics Letters 10, 118-120, 1974 (No. 8). E

J. M. van Nieuwland & H. L. Hagedoorn (EindhovenUniversity of Technology): On the effect of very smallirregularities at the pole faces of a cyclotron magneton the quality of the ion beam.Nucl. Instr. Meth. 116, 13-16, 1974 (No. I). E

J. M. van Nieuwland & N. Hazewindus: Some aspectsof the design of a cyclotron central region.Philips Res. Repts. 29, 528-559, 1974 (No. 6). E

T. G. J. van Oirschot: Liquid-phase epitaxial growthof (AIGa)As on polished and roughened GaP substratesfor transmission photocathodes.Appl. Phys. Letters 24, 211-213, 1974 (No. 5). E

C. van Opdorp, R. C. Peters & M. Klerk: Use ofSchottky-diode collectors for SEM determination ofbulk diffusion lengths.Appl. Phys. Letters 24, 125-126, 1974 (No. 3). E

E. M. van der Ouderaa: Structure and application ofcomputer program ICAN: integrated circuit ac anal-ysis.AGARD Conf. Proc. No. 130,20.1-20.15, 1973. E

R. I. Pedroso & G. A. Domoto (Columbia Universiiy."New York): Technical note on planar solidification'with fixed wall temperature and variable thermal prop-'erties.Trans. ASME C (J. Heat Transfer) 95, 553-555, 1973(No.4). N

J. G. J. Peelen & R. Metselaar: Light scattering bypores in polycrystalline materials: transmission prop-erties of alumina. .._J. appl. Phys. 45, 216-220, 1974 (No. 1~. E

G. Piétri: Progress in PMT for high energy physicsinstrumentation.1973 Int. Conf. on Instrumentation for high energyphysics, Frascati, pp. 586-591. L

T. J. A. Popma, A. M . .van Diepen & P. F. Bongers:Magnetic properties of the garnets

Bio.sCa2xY2.2-2xFe5-x Vx012.J. Phys. Chem. Solids 35, 201-205, 1974 (No. 2). E

W. Puschert: Optical detection of amplitude and phaseof mechanical displacements in the angstrom range.Optics Comm. 10, 357-361, 1974 (No. 4). .A

E. D. Roberts : A modified methacrylate positive elec-tron resist.Appl. Polymer Symp. No. 23, 87-98, 1974. M

P. Schagen: Image tubes with channel electron multi-plication.Adv. in Image Pickup and Display 1, 1-69, 1974. M

C. Schiller & D. Bois: Observation des défauts dans lessemiconducteurs par microscopie à balayage en catho-doluminescence.Rev. Phys. appl. 9, 361-371, 1974 (No. 2). L

E. Schröder: Korrektur der Farbfehler doppelbrechen-der Prismen in digitalen Laserstrahlablenkern.Optik 39, 489-498, 1974 (No. 5). H

A. M. J. H. Seuter (Philips Lighting Division, Eind-hoven): Defect chemistry and electrical transport prop-erties of barium titanate. . .Thesis, Twente 1974. (Philips Res. Repts. Suppl. 1974,No.3.)

P. J. Severin: The influence ofthe phase shift on thick-ness measurements of silicon epitaxial layers with aFourier transform spectrometer.J. Electrochem. Soc. 121, 150-153, 1974 (No. I). E

J. G. Siekman & P. J. van Vugt (Philips Centre forFactory Technologies, Eindhoven): Precision electronbeam welding of vacuum chambers applied to a tita-nium large area proportional counter.DVS-Berichte No. 28, 105-108, 1973.

G. Simpson: The dielectric constants of a ferroelectricceramic.Ferroelectrics 6, 283-288, 1974 (No. 3/4). M

A. J. Smets: The fine solar sensor ofthe AstronomicalNetherlands Satellite.Proc. 1st European Electro-Optics Markets and Tech-nology Conf., Geneva 1972, pp. 57-63; 1973. E

M. Snelling: General context-free parsing in time n2•

Proc. Int. Computing Symp. 1973, Davos, pp. 19-24;1974. B

J. L. Sommerdijk, A. Bril & A: W. de Jager: Two.photon luminescence with ultraviolet excitation oftrivalent praseodymium.J. Luminescence 8, 341-343, 1974 (No. 4). E

J. L. Sommerdijk, .J. M. P. J. Verstegen (PhilipsLightingDivision, Eindhoven) & A. Bril: Lumines-cence of 6s2 ions in magnetoplumbite-like aluminates.Philips Res. Repts. 29, 517-527, 1974 (No. 6).: : E

76 RECENT SCIENTIFIC PUBLICATIONS Philips tech. Rev. 35, No. 2/3

F. A. Sta as, K. Weiss & A. P. Severijns: Surface effi-ciency of various types of heat exchangers in anHe3-He4 dilution refrigerator.Cryogenics 14, 253-263, 1974 (No. 5). E

W. T. Stacy, R. Metselaar, P. K. Larsen, A. Bril &. J. A. Pistorius: Crystal defect mapping via impurity-activated luminescence in gadolinium gallium garnet.Appl. Phys. Letters 24, 254-256, 1974 (No. 6). E

W. T. Stacy, J. A. Pistorius & M. M. Janssen: Helicalgrowth defects in gadolinium gallium garnet.J: Crystal Growth 22, 37-43, 1974 (No. I). E

A. L. Stuijts: Synthesis of materials from powders bysintering.Ann. Rev. Mat. Sci. 3, 363-395, 1973. E

A. Thayse: Differential calculus for functions from(GF(p))n into GF(p).Philips Res. Repts. 29, 560-586, 1974 (No. 6). B

G. E. Thomas & E. E. de Kluizenaar: SIMS with astandard quadrupole residual gas analyser.Rev. sci. Instr. 45, 457-458, 1974 (No. 3). E

H. van Tongeren: Positive column of the Cs-Ar low-pressure discharge.J. appl. Phys. 45, 89-96, 1974 (No. I). E

N. C. de Troye: Integrated injection logic - a newapproach to LSI.1974 IEEE Int. Solid-State Circuits Conf. Digest tech.Papers, pp. 12-13 & 214. E

J. D. B. Veldkamp & W. F. Knippenberg: AnomalousYoung's modulus behaviour of SiC at elevated tem-peratures.J. Physics D 7, 407-411, 1974 (No. 3). E

L. Vriens: Two- and three-electron group models forlow-pressure gas discharges.J. appl. Phys. 45, 1191-1195, 1974 (No. 3). E

H. W. Werner, H. A. M. de Grefte & J. van den Berg:Fingerprint spectra in secondary ion mass spectrom-etry.Adv. in Mass Spectrom. 6, 673-682, 1974. E

A. de Winter & H. Blok (Delft University of Technol-ogy): Fling-off cooling of gear teeth.Trans. ASME B (J. Engng. Ind.) 96, 60-70, 1974(No. I). E

H. Zijlstra: Eigenschaften und Anwendungen vonWasserstoff absorbierenden Verbindungen der Über-gangsmetalIe mit seltenen Erden.Fette - Seifen - Anstrichmittel 76,49-52, 1974 (No. 2).

E

Contents of Electronic Applications Bulletin 32, No. 3, 1974:

L. E. Jansson: Converter circuits for switched-mode power supplies (pp. 83-92).G. van Aller & Th. G. Schut: Combined image-intensifier and 'Plumbicon' tube for studio colour cameras(pp.94-100).J. Mulder: Input network design for a high-frequency wideband power amplifier (pp. 101-109).A. A. Los, D. C. L. Vangheluwe & A. C. Smaal: Ultrasonic delay lines for chrominance signal decoding (pp.111-120).Nomogram for single-layer air-cored coils (p. 124).

Contents of Mullard Technical Communications 13, No. 124, 1974:

Titanium-gold: high-reliability transistor metallisation (pp. 142-152).J. Merrett: Summary of factors affecting power supply choice and design (pp. 153-156).W. Hetterscheid: Base circuit design for high-voltage switching transistors in power converters (pp. 157-169).C. Barrow: Cores for switched-mode power supply transformers (pp. 170-181).CL8960 X-band doppier radar module (pp. 182-184).

Contents ofValvo Berichte 19, No. 1, 1974:

P. Bockelmann & W. Westendorf: Die elektronisch en NF-Einsteller TCA 730 und TCA 740 (pp. 1-18).F. Pötzl: VHF-Breitbandzirkulatoren in konzentrierter Bauweise (pp. 19-28).

Volume 35, 1975, No. 2/3 Published 2nd July 1975pages 29-76