Document GK02ZJMz004bk2XndZNoobLMq

tr !#* "' 25 11 ~i u - 1 j|p.m>|Ly.j,|JiJi 1-- r l U2zO : r1 11. i p i 4j,.J r ji iu j! iu ijjii; [ . m u m jij j p< * , * w m >m LEAD INDUSTRIES ASSOCIATION O CAST 41K> STREET * NEW YORK 17, N. V. December ik, 1959 SUBJECT: "UIJRA LOW LOSS c t r a mt c TmgzcTRlcam to Members of the Lead Industries Association: Enclosed Is s reprint of an article entitled "Ultra Low Loss Ceramic Dielectrics," which appeared la a recent Issue of "The Journal of the American Ceramic Society." This paper vaa presented by Bunas arl Koenig before the Electronics Division of the American Ceramic Society at Asbury Park last fall. As acknowledged on the last page, the work vas made possible by the LIA fellowship grant to Rutgers. Additional ccples are available at no cost in quantities up to 25 and at $9-2C per hundred in larger quantities, as long as our supply lasts. Very truly yours. Secretary ^p > p q p iw ; g t * a T i j *!'!NWIS Me -we " '*U ...ii'." 1 '"w,w rujw-ssi ji,a,ji wypijyy. ryyrete?,|gypj <! am ' ;' ' ' ' I'^'iij* uiuwpjy^w LIA22439 inf>i<rfni iw k imCm 4 W Lm< WmMh i *9 i , Ultra Low Loss Ceramic Dielectrics byM. M. *UNAO mni J. H. KOtNtt Tha tkjKi mt (Mi fuu tu to derate} uhra tow teat cersalc tfitlMtiks. Tto fouiblt fi>Mi W tha dialactrk tow al cruk Min were iK luiyiH Uitorttki&r. li^tfiAMU) fcodiM, 4nl|f4 m th# Wait f tkii usljiis proeida to Hn tow dialactrk tow. o 1 - i - ill tread tohn frnfwrnrr rWrtMotr n|tii|4nrat Tbit m|iurrd Ik nwwmil 4 ultra krw Io n errata* ituU( Him hit Itrru nlcoWie wt*i un4cf a\ ibt tlirw iwrt dunni mrut iriK ilO(li uwh d I hi* wk ha* a*a lrw frtelrd i tar literature.1 IVtr ratlirf data wrtr rrvirvd W* mL| ihr ptaewt Itomril italviti It fMtIUr Cavaaa tf OtototMc Ui i to Cirsult If an i i rln in' told h ai^ml lit a netdrnwf which h uadi i4 rrrlaia dtrlrrlr* wid Ihr rtnirir told ihea w act ai a hr* attHfrd utrt mr oiV, Ihea Ihr dtflrtlrif hat mo dirhvtrtr |>m Hut aiiul drlrtlik* hatv kanr tovt, ill uhwh it* Ihr a^itod rlnlrtr rnrt|t I* (math nc*rtltd In tout t5w t ihr rlrrlru' told Ar *ri i Ihr drlMrr. ihr pbaw an{W town thr told and rlntnr cwrrrul tfeaild be Ina than HO*, r Wl* - 4. tm I U Ihr "ha aaftr."* t'Mially m I (j*w toto) re Ut I (dwipalto total) to uwd tu *\ vrily the km <Wtrkr U dktoctric. Aav |nmu n\trtiB| thr and rlrrtric cati0 <d the dickrtric u(u heal eaerjry cue be a cuusr d darWctrk tot tdectncul ciadwlta *4 a drtotne aad tint U| of the leieutatow of a pminmt dtp>Jc (or id aa iaduccd electric di|4r uiaaret at ultra hi<h frequency) with rrrpccl to the lhaw ui|lr i4 aa apptod rlrrtric told are (uonra cum d dirtrclrir Ion. Thc*e kmr* arr <d aa electrical crifia ta the nurwvtjfw aruae When a dirlrctric riU of a mixture d turn it tu-rr mtrruK thr Maswefl-Wajcnce type cd 4i* hnnihl al |W FaO Martini *4 Ihr lOnrtnaur* Dhbto, Thr AnirTHan Crtamir knHy, Admryr Kit, Nim levy, Oclutoe |l', IV.Vi Krmvrd Octutor S, IMh; nrtod nfjr reoHrad April 77, IMi Thr *rm arr. mpevtKrl*. tnf*b uaiUU aad dtolw, 1hkiil d Cman, Rutfret, Thr $utr t'tmnHy fYfrtM kn-l*. C'Wirari* |)A MinttfAdM, -42577, Saliet, 8rh>a4 <4 Ceramic**, hvtftn, Thr ftutr Vatreralljr, Thr I'atlrd Hair* Arm) tofnal knrtirh Libnl<ry, Fet Mw- aawlh, Nn Jnwy. Ml 1 N 1955.01 September 1959 LIA22440 Ultra Lam Last Ceramic Oieiettria 444 electric km appear*. Thi* km is dur to u accumulate* uf electric charges i Ike {rain kamUrics vilhia (be mistore. rmallr, the Umrrlicil dielectric cesunt tad pien (tclrf <4 ibrt^(tQpt tubiUDCT (a pvkturr *d to Burr phasn, Hie a ceramic body) art in terms <4 the di electric |mqftir*<4 IW cum)Nmt |4isei Seven] iVnir* have been drtf*<i*d rexardmf this, mrii as Wciorf and Lkhtreccker'i (rmU,1 Hmp^maa's tbwey,* and NM'i theory.4 home <4 tine ibnric* Has takes iato acrrwst the km due Ui tbe mechanical interact** betwees |>hari. Fee an ullrs toe km dwWctrir. ibrrr is a pmMity tkat this V- Cimm an tsu|Hrftaat part id tk mechanism. F\e cxamfdr. cs*wdrr a ceramic body nuioia( pirath rV-lric material *uch as crystalline tjuaru UVu aa electric field H with a hn;h frequency is aj^dsrd l* a ceramic, each pebcktHc rn *laJIe in it viB lend to nvilUU srlh the same frequency. TV anqdttudr w<dd be vets. very small, since the crystallite* art sornatitderf by a cUwsfduiw. Ihsrto this oscilUU*. s k u m] m <4 frequency srill be leorfatcd at each crystallite in thr c t t amir bndy. TV dmi{u(i t4 IV wamd wavr energy drjmdinK u q tV elastic |efartirs 4 tV ceramic sriQ c*uw a dirUctfar k*s. Even though a ceramic mil nmtam any |iii>eMric material. IV same kmd 4 mrchamnal Im may occur A siin(4r il1ustrali* 4 this is given in Fig I, so which a di electric c.<ntains a li| cylindrr with dielectric runslant a, embedded in a hterxirmniui sutnUare *4 dielectric ntfHUal M. If aa electric Vld. K*. is a|<pkd, there will be a kace (trcqur) no tV era! If e( > then tV trequ? sriD have a tendency tn decrease If t < *. tV reverse will occur. Thus the i<id srifl tend to make mtatamal tikatim with a frequency of 3 a. generating a wnmd wave If tbe flux urd f*e making a msmir U4y ** not aiflirint, numerous voids wiU be fmnd in IV fired My. Since wane *d tV raw material* uwd in a ceramic Udy are fa* bru-ieg materials, tw will V |si4wtd wbea tV tody is bred. TVtrfiee, evro if suffW'senl flux is uwd. there will be ta tV body Voids may weaken IV binding <4 tV rrystallme phase in tV kdy and wit! facilitate the vihratk* of tV small * Karl Lirbteimkef. **IWtric Oausst i4 Natural and 5>Wthrtk Mixtures," /*,* Z . IT, lift- S* (lJCftl * l A (I hntn'-nun, "TV CilvUila <4 Viria Hiywral C*a^tets<4 llrlrripiieiii SiUtamn: I. iHrtrelric Constant* and Oaduclivriin *4 ltiiure% C*spM*d <4 t**<r>pir fJubi,awnlM A mm Vkw . It, KV\ <4 (|KW) * H' Ninel. ' Ffkvfft JVtnlwlia in Miiiwn *4 Two |M rln inn is aa kWrtrir IVId isd CilvUtiai 4 the 1>m AngV in Oui^liliairr hetw%" Atm Tie* , 12, 4|0 (tf&l) partsdes under the Vftucaor of aa applied cfcctric held, than fmntixi a auuad ant. Them voids cma V miaiauacd by proper technique ia the maple petparmtsun. la thr afurtmentimed cases, the dielectric kwarn arc function* of the elastic properties of the ceramic as vt9 aa thr dielectric properties of thr emipcoat phatra. Ei- lerimnlal tests for examiumc the lelatkct betweta the dielectric luns and thr tLflic property <4 oeranuc bodies arc rerssary. TV result is nut availatilr at present but will be wdcrtika in tV near fstne. W4U*tmite was fiend to be an excellent crystalline phase be ultra k/w P*a kdin in thr earlier work.1 TV dielectric lowcs id mch wutlastunitr tvj* budin are next riainihred ns a funrtb* id tV si4hat<ailr content. Figure ?(>4) baa summed diagram uftVdirlectricloa* of a cries if wi4U*lunite bidri TV }(* dot-dash hae reprt* seats tV k*ss due to the wnUast*tte crystal* and tV matrix phase (cias* and Irad alunxmum whcale). TV ayqmnimatr shape uf tV esnt may be estimated by niiq Niesrfs cqtsati> 4 UVt tV ratio id tV dsrlmnr noUan 4 tV seal* liUioitr cry4als and tV matnx |fue n oturfy I. the curve can be a|*|w<iximatcd by a stxat^ht hue The brokes kar represents IV dielectric Vu csbsd by tV ufd iodudeU a the bide This |m waild herrsv m)adlv srith decrrasioi dux oolcot bemuse tV lack 4 dux srill htndwe many vuidt TV upfwf subd line rr|ee esti thr nsuhant kiss which nu^hl haw a Ruaimwm Value at a certain sdUrtMite conntiatKO Asaumioz that tbe Voy ha% thr |rptr unust (4 dux and has been fviqwriv falrieUd. the dirhvtiir koa due to voids should V \w uua3 TVtef.er. the main part of thr dielectric k*s4 tV ceramic hdy wiQ V doc tn tV ofhiOaitr crystals and tV rnatitx phaw Stare the dielectric baa t4 the sifUMusitr crystal is fixed. pofoUy the tint tray tn decrease tV dielectric M 4 tV body is to se a matrix with a lower dielectric Inu la this case, the roultaol dirlrctrie kaa arid sK V kmered (x shown by tV lart t 4 curves is Pq. 2(,4)). Hut. if tV dielectric k* 4 tV matrix phase wne kmrr than that 4 thr v>fU<oilr crystal*, thru IV dirIre* Irir |mu v . smtlaMtmitr nortstnlkm curve would not have a mimmum. as in Fig. 2\fl) la such case, it is brtter to wr clan itsrtf as tV dielectric, and three would V no nerd to use wiflaMtailr if thr (Kirpmc is tn make (hr kmrrt dirfectric ins bode. Knee tV matrix phase <4 thr body r a (Itn ir a mixture nf a flaws and a small amount *4 little crystals Ike Iborrtkxl dielectric kiss <4 flaw** in rrlatkai tn mm(Nwtio cma V rnerrlated srith thr resultant dielectric km 14 ceramic bodies. M*t classes oattain sihnmsuyzen wtwl as tV main structural units and several kinds 4 |w*ilit kaM 1'suaUy, wane uf tV fwisitive i<o are tnnind to SiO ortreki with rather weak fatrs which may be <4 inosr <r rm-ahwt charartrf Heciuisr tV districts* if SiO nrtwwks in tV ftaaa lacks aa orientational rezularity, there must V many p* initial minima fin tV pmuivf was It run V amimed that WiiHF." i iiu j u ' O ifi|ij u.jwjpifji1.4* UWf < 1.1*22441 s wwwapM panwwwp 444 Jmirwat f Thr American Cmmk S0ri*ty--Bmm*gmm4 Kfnig TtUt L MoriuhW f of Svrol lorn Aceorrflnf fwhf u* Na* K* Rh* Cs* o-- 0 09* 0 17* on l ' 2 42 a an Re** Cs'4 8r** Ba** toot 0 0*4 0 47 om lift * AP* fc** V*' La** 0.009 0 082 on 0u 1.04 C** 0 0013 * 0 010* Tl** a.M Zr** 0.37 Ct** 0.71 Vq I. 42, No. 9 Table 1. at tom AccarrfVf la tabs** f- 25 a o- fil Rr- 110 1- R 4 0*" 30 0 U* Na* * Rb* Cs* Ca* It 7 20.2 <2.3 * I 71.3 27.2 Si 4 0.1 * TVk, ubkr *bnv* Ok hfeb I Ca*. Mf** Ca** h* Ba** fk*' It 4 34.0 40.1 7*4 01.3 *** 22 TH* U 1 IV4, Be'4, IT, tkc pwlm iott arc dutributH ta ausy potential db arrrfdia{ |ti tbt U*i <4 fUtialica) b c Hu mi. TV t",(nl'a' *ner{y a positive km Hh a ebarft i V pm by tk cqaatkm r-r4- Vf 4 (mm hr artilhr) ekrtrir pntrotial ekrtrir kM t<n|lh, at (hr pmfcim 4 Ik ha. pknabikr rrf the ha. tMpar a oar when Na ions <4 Na |ka arc replaced by fit ur Ha hai Since thr electric chant erf ft (oa* Ha) ka it Ivicr that erf Na ana mad the pirfartfability h about 3 te t kart that erf Na ha, the depth erf the potential wait fie R (or Ba) ha miK1 be mute than twice that <rf Na ina. iwftim that 4 aid are rf the same vahr fra the two glam TVrdiet. the deep p><ralial wetH can hr realiied if the nlmr and pnlariJabOrty erf the tom are Urge It wiQ be ward m Tal4r t that fia and Ct have hi^h eketroaic pekar* Ushflitin Robert*1 baa cwlraiatrd pnlariaabOitart erf inm ia tiaplr (tTtuk. The bade r|mlha fnn which the pnlariia* biiilin have been caknlatedia - Z~* - tl\K - UA* 4 2) r rn ihar rf the Mil ceff ia the repeal lattice divided bjr d* Makr rf whVr which it naif. ^hnahike ptr ia ibrak. |hntiUkv rf ibr n ipetiw hat A" ihhrtnr caiUal rf the crystal. t*in( the trfxrrwd data rf A* of a aamber trf crystal*, thr pidanJsMVtin Have been estimated and are tabulated ia Table It The hrrcomg equalha ia based cm the oi.k.a to obtain the prrfariratrfbtic* based tm thr Ci a ait system, thr pirfarirabiktira tabulated alm\t *b"uld he ditided by 4r. It should be anted that the pnlaritalMlitir* trfrtainrd be Rirfwrta1 wwlude at only the rlrdnmir part* but also thr inair part*. The thdnak prlarirahibtir* ran be amu card In hatr af^ifiaimatrlv a ealue im ^erliif rf thr crystal atmrtore in which thr ha exist*, bat the hair p4ar< babdi'irs should chance drpeadiac oa the crystal rfwlat fhrpknl tihrrts "IHrkrlric Ca^UiU and hhnuhket rf b m Smr iJt CrjiU^ aad lUriuui Ttuult," fi^ Ktr, ft, 1214-30 (1M0|. RpS. Vohdkrrfi The atnxtim erf flam is sne otAa thaa aaost crystal* uthat the prlarirabihtkaof km* in the glulnmld bedirfertat fmai tbuse listed ia Tables I aad It. NenrtKkm, it caa Ur aimed that the when <rf atcnilaV are the aam evra thooch the ha ate ia the crystalline or ia the flamy state. Rhea aa rhrtric Wd is applied U the (Urn, dielectric phruaha. which belonging,d rf ehcvnmic aad ioaar polar* iaatkm. w*B appear. r - r+ + f- - 4 abut fsw u dae to the dhtatha uf the eketna doad of thr hue by the ap$*bcd field; is dor to the drifting <4 the ion*; ooe part 4 which dae to the samB shift ia each potential well aad another part erf which (^*m) n dae to thr lomaccacat erf the intributia rf aB the punitive bum ear aB thr pteetid weds aader the tnftaracr erf the applied ehee- trir held Since the pnrfmlrfhty erf escape erf roe powtive ioa from the potential wtfl n pnprtkeul to the exp |-((aMr /ADI (tr4 3). r*mm am*t depend atroaplr ou the depth <4 thi* potential weB aad the trmperatore. If the force between thr puitm km aad the Sirf) arlwtci is suft- cicfttK tac so that the ptmiul weQ fur thr poathr km is eery deep, thca the dkketric bos dae to ^*** appears only at very kick frequency. such * infrared ^vtical frequency, became thr ie*t**iy bare be thr daft *rf thr ina it rtn| ia thr drep ptestu) weft. IWlne 1 on dor to tur ap pear at very in freqnracy bat duffein at bigk fiequeonr (mtermrdUu radio frequency rxige). The teasna it an frrfbfw*: Sore thr rale <4 Irurfcf <4 thr positive km fra (me inaeatial well In swtbrr t* leifmctiuoil to the exp |-(('b m *m thru the rate * mulkr be drrjwr pmeatiel wtIK and k*wt trat|wratnre. If thr rate *4 tramrfer is mil and the fmtartw-y rrf the a|<f4wd brid w Ucb, the positive km do M have aufirwnt liar t* frarh thr state <4 eqai- bhritm dr4hlniik wbkh L rapidly ckingai with thr a.<. drethe field A* a remit, ;brtr is ili an notribt- tkm fnas f*M wbrn thr ImjarUfT is wifbriraUv ki<k ta *arh rasrt, and m oatriUk httlr to the darketfir low ia the rad (rrqarwry ranee Thr f'daing rvam|4r due* thr nay thr drf*th rrf thr po* teatial well sfnl thr darlrrtrir kra at drffvtent freqaearwa. It ha* brew stated that thr Im due to apfwwra at ap proximately the frrqurory p\t* by thr eqnatkm avq (-?ir) where a is *ec. *. Awcmr that a certain Na (iua show*a Inch dwlntrirkoaat |(f w Mabat I Inc) If the Na km n replaced bv Il< e Ha km and thr potential depth become* twice that trf the Na (Um, then lc tktfba Ml 'w^V'L1 - ..uiywri gyp*1:- *U-U*!1 Pff!J il||H!JW1 HI '-l- ,um| Ll *22442 tm --m i- ScvUmbo- ISM Ultra Law Lwu Cmwtit Dirledrici TtWci. > ihoftor nf, Tow I ot Wtrmt Fft^wmciw * Ti *rp lufi neqmncy t Ta "ft -ft i i *+J 10 too *+T5 too 10,000 *+ idooi Low flrequrwy t1 *oT| fi hr, 01 01 0.01 0.001 *+r *+r5i * m CMlnbjtt Ip . when $ (I/r) <** + , Mt vW> (l/Ti) wifi contribute lo lu I rWa m k ipnipjpmty 1/7* hit tiB not pptrflpti ftrtktr flips XfT Cbat 445 mmTt r+*n V 101 tOOs ram to x too tot Vmn ioooi to x too t "* m 0*4 tpl < I Tobla IV. Fnvnr Foden of $mfl Clous According to i. M. Simh ISut ford &0, (SQ %> o oaa Xa*U(SiO 70.4%; AhO,*^; KaiOICO^; CpO Jl^) 0.0006 hafUx(SiO. JO-OCi; MblOOf;; AIA K3: HaO *> 0^) 0 eou rtfU(SiO7.r;; AVOiao^; Ka,OSt^; two a.oS; rtOMO%>o.ooio * liautox. ( Ba fist. k (ivti by the kfioia| <qii(k, iMsini h k tV not, resnhmc in * 10*1 which k a very low frequency. Go<R(a(fll)y. iVrr MU br ihnnri po dielectric kx dot to am tV Iftqwncy nh{t (nMi about 100 c.p %. to ihot I0* c p.t. Ttedun. (Uw* nt4ntfl|( deep potential we0s can V expected to Vw a auxiawa dielectrk u m, Ike km n. frequency curve, at a kw lrt^wq`. At abnut tkc Frequency ranee where Ike dielectric constant berm* to decrease (Fit 4). the phase at f*k delated e*npared to the phaw at the applied held A a remit, hi*b dielectric hm appearv Generally ipnkinc, when the dielectric nUflL . for flow frequency k appreciably hither than the square id the refractive iwki a f< iptical frequency, a dielectric tnu should be expected due to /*, at a certain frequency rup where d* 'k k lar^T. w k tV frequency of the a<. held. The fotkiwiai ittustratnx (Table 111, Fqt 4) abow* the behavior at the dielectric constant and purer factor at dif " * v i *r,` dwWclrir rnuuant a: lay ipoM frequency dwbrtnr oaKUat MtnrduUIr (<0<rwiAf tkr applied held 4Wuk pptUM after the aplwtia id a cu*t hid. rrhuUna thna. TV (orrt| Kputioa k far shr <w rcWutM tine, tVa there k man thu ox itUabaa list * " * + i *fi*+ rrsn + i sfc- r*c Actual fUw cmtain many kinds of potential weflk with various potential depths. The explanation erven above, however, can explain qualitatively the tow float character at glasses molainvnx deep potential velli Examples d dielectric lust for arveral |tum are damn in Table IV. It k dear from these data that glamc* eralotmag ions such nFbnr Ha. which bate bich elcctrraif p4sms> bihties, show appreciably lower dielectric but thar does Na (Ian. Pure laud silica dor* not mnlam fnripi inns an that dielectric Vm due to iir movement skaild be very low. U*ben a flaw eratain* fiwrtm kns. flufher valency, ns wcU as higher eWctnmic pdamability f the kn, k ntrenary for the (Ian to have a he dirkvit floaa (I) Summary a/ BawsHcel Aad^in When a ceramic crouds uf tm or nrr phases, the |4kcipal dielectric |nu k due to the dielectric luw td tV toauponrnt phases and may be due ahn to the Hade floss of the ceramic hdy To decrease the Kx, the falKia( conditions arc mecuary: (1) The diekrtric knars cd the ctnapramt phases (crystal phase and binder phav) skaald be as flow as prusihle A flow dielectric bus of the binder phac can br realised if it eratains sons such as Pb, B< Tt Cd. and Ha, which have fltifb valency and kn{h electrnnic poUrimhalitr. n i a h t r a in . . . i t a ............ 446 Jmtrnnl tj Tkt Amman Crramit Sacitty--3nmo[ and Knenit VoL 42. No. TaU* V. My CoapoiHion* |l %| ant frepartlM U-I RSI KM KM KM KaQtfwilfoirtw 0.0 uo to.o 80 0 56 8 80.0 It No. 4 bal day BariannarbramW Lead hwflieste hirWMtb triwidr JO 0 10.0 too 20 0 15.0 M.O 10.0 15.0 ISO 20.0 M.O 20 0 15.0 10.0 U.O M.O 10 0 5.0 Sikes 2.0 2 0 Uitn. m, (*F.) Water abwvptfo Power factor (*A) fat 1 ne. ssf r ms tasp )* Dirfectric nouataat* Lorn factor ()* JAK-l-10 grade 2100 0 00 0.01* S5 om U7 2100 C.00 0.008 in 0.008 L-8 2100 0.00 0.017 7 83 0 128 L-7 2100 0.00 0 014 8 82 0 101 L-7 2100 0.00 0 012 7.15 0.088 U 021.0000 0.014 7.40 8. MB L-7 These vmfocs are awage malts af at hast 10 spaefoms. - (2) Reduce tbe km due to daKic cam. For Um poac. It n that () a iwpieMkctnc Mtcriil be used m (be principal component *f ibe bodj; (I) the pria* ciprt dielectric constants and*, of the principal crystal- foe |foe rfwold be frtv the tit tad they should be il- toet equal to the ddeevir cooftaat f 7f. bender phase; (r) if the demure d the diekctric nttaati d the phase* a relatively Urp, the paia d the principal phase fkald be tarty ipherical ic chape; (f) die (tun' or binder phase should cenat the crystals tight enough to prevent a ms- version d electric energy to mechanical eaergy; and (e) the <3) Cod the ccraat .dj- very slowly, the law* hint the same as for the annealing d a glass. During the a--pram, rearrangement d molecular wtaorti r was in the fim take* place imhiog la aveaubhodpntioa d the artworts and the n m. 1c this case, the potential crib far the was can be expected to be deeper than those d the manorsWd glass. (J) hdllsfisd Dielectric constants and power (acton were measured at 1 me and mom temperature using a Booutou Q meter. Sihtr paint was need as the electrodes to provide a better contact with the materiel because it caa flow into the minute dtpnfoum on the surface of tbr specimen Cue eras taken to paint on enough silver so that there wm no rriiiteiM on the surface, measured between the opposite edges of the specimen. The sOvcied specimens were dried slowly on a foH plate for JO minutes to about 500*F. and held them fen 1 hour to make sure that aU the organic solvents of the surer snipemi on were evaporated. The spreimem were cooled to room temperature ia a desiccator. During the entire electrical test period, the humidity and temperature must be fairtv constant to obtain consilient result*. Tbe ermditino* (temperature and humidity) can be awurt to be constant when tbe capacitance and Q readings with air as the dielectric remain the same. All uf the matured forties, including those sbosm in Tables V and VI. were analysed by X-ray and prtnrapbie methods. IV. Dtsewsslan sf In suds ,, A WUftgm batch d the cwpoitin eras weighed, and _ Sun cc. d water were added; the hatch was wet ball foiled Referring to Table V. bodies ESI. ES2, and ES-3 were fivxed with lead and barium, bodies ES-4 and ESA were far 24 hours and then dried completely. The dried hatch was pawed through a foempuharim. mixed la a Lancaster mixer for 10 minutes while adding C% water by weight, and then mixed for an additional 20 fonutr* la the Laa* caster. The material thus prepared was prejsrd into disks 2 in. la diameter to irmidr close owlart d particles and to TobU vi Body Composition M. %) ofo fropwtws fotiate prrhmsnary ihetmral rr*cti><ns during a peefiring to |fcd*F. The ptesed disks were fivefired to cfcsniaatc the voids cnnxrd bt' {svw rvnhtiia (nw the oaitpuonti and to aasttmsr inteinal stresses, which would otherwise de tehp in the grain boundaries, in the final test specimens. The pcefired dnkt were crushed and dry bail milled for about 15 hours to |a a 200-mesh screen. Sis cc. a SuperM binder (li^ sdutinu) wv*e added per 100gm. of powder and mixed in a Lancaster mixer for 30 minutes. The p<wdcr NV 424* mdUatmite Frit JAI Krit JS2 MilaWMfCri Water itMCptaw h>*f U<ii< (Cr) (at 1 me. and rasa temp )* Dwkctnr pmUst* Lma lacVW 1* JANIIO grade IK 1 80 0 40 0 2000 o on 0 0724 8 04 o L-5 J* * 80 0 40 0 900 0 00 o am 7 5R 0 1| L-T was pressed ratn specuneni (disks iptrutiauldy 2 in. la dfoarUv and 0 1 in Ourt) at TCOnlh. prr tq is., fired to the Farr Coumamowt maturing temperature in 4 twain, soaked at that tesnpcratafv 1*1 1** fr 20 minutes, and oulrd very aforty to won tempermtnrc no I fo dining appmamutefy 19 boors. (2) fitpnhfodlsdN KO. MO, NaO M> Several experimental forties wm designed bawd on the UO M0 17 5 20 55 80 54 0 57 5 20 85 previous theoretical diiruMwn. The cumpositinus cf some d these forties are grim in Table* V and VI. *TW values srr average rrsnhs 4 at kraU 10 qiehnu ! i r l i puwswyw .l s j u I n i i.w mmf i Sl ww 1IJIH Ml . ';, v- ;. . jiumtt *- ^ . . . i:.--^ . '* ;s* .- . . f \.U _- 1 September 1950 VUf Lme Lm Ctrtmic Dieketria 447 nrit ITS in. the ternary sj-ricn PtiO At/VSOi (Pig. S). The calculated batch lurapuatiua of melt ttt was fabricated ialo a ceramic hdy The body hat a firing range of 19UU* to I4G0*P. and abnws a adf-glaring jeipcrty. Bodies ES-4 and ES-S, which were fluxed with bssmalh instead of lend, show the same peeks as the other ES bodies. - No report eras found in the literature of nay baamth com pounds that cnindde with these peaks. A empoatioa riaular to melt 128, but wring BiiOi of PbC> sms fabricated and fired to maturity at 1600*F. The fired body was analysed by X-ray and showed the aamr peaks as melt 128. Probably biimath reacts the aame way as lead aad forms bismuth aluminum silicate similar ia struciarr to lend fluxed with bismuth ud banuu, Md b>dy E$4 vs fhoed with lead. bismuth aad barium. All id the** bodies were fiatod to lute lin luv diekctric low (L-7 aad L4, JAN-I- 10); they prmnird hrary muddying cation* swch a* Pb". Bi**, and Bi***. which have high vakocy (ia ctenpartvai with the alkalis) and high ekrtrooic pnlariiahiHtie*. This conbrms to the pminn Ihnrrtkal uphaitioa d deep potential vefii. Table M jnti Iwn body ooopuritiuns ia which all td the {Um phases were,preffitted. The craapowUrm* r4 the frits rr aim fh eti is Table VI. These two bdict provsdr a dear campsrima at the effect of heavy cat*on* with high tlectnwic polarirahibly no dieketrie kat The {Uv phase id JC I ooetaim Na0 aad K/l and the dielectric law ts relatively high (laa I 7.2 X 10 *<). The presence o? alkali inns increases the power km. Where BaO replaces the alkabs. as so JK2. the dirketrit craretaat b somewhat increased. The pnrer factor bnweter. b sub stantially reduced (tae I 5.7 X 10 * for JK-J). It ap pears that lead aad bamm UftlkT the best flozac system id that tried lodate h* ultra low lnuteraair bndin The X-ray patterns nine that aQ id (hr Isdb cimtain flwidUstanitc crystal* Bodies ES I. ES-2 ES S. and ESA ihw a few other peak* identified as kad aluminum whrale. Thin section* of the bodies were ouuained petropaphiaHy and were found to contain crystals, glam, aad samO tiamiaf cd raids. The cryeuk, which were eamfl aad id irregular shape, were identified as lend alaniuiuwi wlicate by X my. The wuQastomtc crystals cannot be saea without croamdKirub because the refractive index of wnflitonite is rery doa to that of the mneraua medium. t'aag maned Nicola some of the peedoeninantly larger woUastuoitc crystal* are very dearly lets. The remaiaing tiny crrsuK are a mixture cd widUxtnche crystals aad knd ahmnaum silicate crystals. It can be noted that although the fired bodies contain woflastonilc crystals which have aredktike ihqes. they have dim bar dielectric losa. These may be explained by the following: (1) Tbe diekctric constants of tbe osnpocet phases are reUUvely riose ( id wnflaslonilr crystal ill; of bodies T. 14; cmwt|umtfy tbe of tbr binder phase is approximately KS). (2) Tbe nU*t<*ute crvAaH are wefl bonded by tbr tiny crystallites, which ha\e aw-ibme paperty. and tbe glass phase Tbrrefine, tbe clastic fcna due , vibnUna of tbe wtdlasttnnie r> ul* may be rery small V. Condualau* I (I) ExperimcaUl Unties, srhich were designed on tbe basis of theoretical consideration*, ptoridi lev sdtra low (2) Tbe nhra V*w Im rceairic body should br oapuad td a W>w has priarifKtl crystalline phase stach as wolUxirvitc with a binder phase cetsiiwifi brsn* aodtfvr ( cwthmt with hirh ekctric pidariiabihtirs such as Pb* *, Ha**, and Ui * * * It appears that Pb and Ha together pcoeidc the best fluxing system id those tried to date for nitre low di electric to** ceramic* The th><r want lo arlowwledge the advice sod ima u s o td Etna* Jwwmptrbi d thr hstim Srh*f rf Ceramics The pi> eat wtek was yamd hr tbr Lead Industries Aswciatson Earbn nil which provided tmckgoauMl he this iavrsligath* wvt ifsosnd bm by the Signal Gaps Bcmirti aad Omifmeat Uhiawin ; y .U- p, --ujwwnoi. 'M.is.PTy ' <? rr