il)-137 358 A DIELECTRIC STUDY OF THE /TIIE-TEMPERATURE-TRANSFORMATION (TTT) DIAGRAM 0-U)UMASSACHUSETTS INST OF TECH CAMBRIDGE

UCASIFIED N F SHEPPARD ET AL. 14 DEC 83 TR-7 F/G il/ NLmNLS OET hEEEohmiEEKED

- i.t .. . . .. tL .. .

1140

1"25 1.4 11.6

ICROCOP RESOLUTION TE HAT '

WATNLBURAnO inWRD932 A

OFFICE OF NAVAL RESEARCH

Contract N00014-78-C-0591

Task No. NR 356-691

TECHNICAL REPORT NO. 7

A DIELECTRIC STUDY OF THE TIME-TEMPERATURE-TRANSFORMATION (TTT)

tDIAGRAM OF DGEBA EPOXY RESINS CURED WITH DDS

by

Norman F. Sheppard, Jr., Michael C. W. Colnand Stephen D. Senturia

Article prepared for presentatLon at

The 1984 SAMPE Meeting, Reno, Nev April 1984

MASSACHUSETTS INSTITUTE OF TECHNOLOGYDepartment of Electrical Engineering and Computer Science

and Center for Materials Science and EngineeringCambridge MA 02139

December 14, 1983

oReproduction in whole or in part is permitted for any purposeof the United States Government.

LAME This document has been approved for public release and sale.j its distribution is unlimited.

84 01 30U49" ' ' '-.% * ;%,a,, %. -.. ' -. '. ... '.. ' -: . .... '.I*,.. - ....-.- ,'.. .,-:.. ..-.. ,. . ,

UNCLASSIFIED4tC'tITY CLASSIFICATION OF THIS PAGE 'W1,n Veto Enited)

READ INSTRUCTIONSREPOR T DOCUMENTATION PAGE BEFORE COMPLETING FORM1. REPORT N4UMBEA 12. GOVT ACCESSION NO.j 3. RECIPIENT'S CATALOG NIJMUER

4. TITLE (and Subtitle) S.TYPE OF REPORT & PERIOD COVERED

A DIELECTRIC STUDY OF THE TIME-TEMPERATURE- Technical ReportTRANSFORMATION (TTT) DIAGRAM OF DGEBA EPOXY 6/83 - 12/83RESINS CURED WITH DDS -6. PERFORMING ON*. REPORT NiJMUER

___________________________________Technical Report No. 77. AUTi4ORfs) J . CONTRACT OR GRANT NUMBER(S)

Norman F. Sheppard, Jr., Michael C. W. Coln, N00014-78-C-0591and Stephen D. Senturia/

S. PERFORMING ORGANIZATION NAME AND ADDRESS 10. PROGRAM ELEMEN T, PROJECT. TASKC

Massachusetts Institute of Technology AE OKUI UBR

Department of Electrical Engineering andComputer NR 356-691Science, Cambridge MA 02139

It. CONTROLING OFFICE NAME AND ADDRESS -12. REPORT DATE

Department of the Navy, Office of Naval Research, December 14, 1983800 N. Quincy Street, Arlington VA 22217, Code427 13. NUMBERO0P PAGES

81.MONITORING AGENCY NAME & AOORESS4'lf different tvam Contmioihag Office) 11. SECURITY CLASS. (of this reortf)

UNCLASSIFIED

I &. DECL ASSSI1CATIONa/ DOWN GRADINGSCHEDULE

16. DISTRIBUTION STATEMENT (*I f ti Report)

This document has been approved for public release and sale; its distributionis unlimited

17. DISTRIBUTION STATEMENT (of t"e abstract 81199,0d to Block 20. It Elftersnt 10m R011048)

10. SUPPLEMENTARY NOTES

19. IY WORDS (Cmrtho on roers side it moceceep 4114 Idm1Ii& by 610ok unbor)

Nicrodielectrometry, dielectric cure monitoring, time-temperature-transforma-tion diagram, epoxy resins, DGEBA, DDS, condictivity, permittivity, lossfactor.

20. ABSTRACT (Coagi an Fevers* sie Itgo*01 ME J41011110 17 6100* 0101060f)

-lcrodielectrometry and differential scanning calorimetry have been used tostudy the isothermal cure of EPON 825 with diaminodiphenylsulf one (DDS). Theresults are compared with published torsional braid analysis (TBA) studies ofthe same system using a time-temperature-transformation diagram. The dielec-tric measurements were made at temperatures between 100 and 220 C using mecas-urement frequencies between 0.1 Hz and 10,000 Hz. The DSC was operated in

(continued on reverse)

D 1473 toITION oF I NOV 65 IS OUSOLETEt

5/4 102-F41A6601SECURITY CLASSIFICATION OF ThIS PAGE (W~gObef DS EtsmE1)

..................................................... .-. .. ~ . . . .

P. P. i9 T. 11 P. . .sg . * .- . . .

USCLASIFIEDSIRCUmITY Ct.ASSIFICATION OF THIS PAGI (When Daa EntmrO

(continuation of Abstract)

an isothermal mode to determine extent of conversion at temperatures between137 and 177 C. Two mechanisms of charge transport are responsible for theobserved dielectric response; an ionic conductivity early in cure that de-creases as the reaction proceeds, and a Debye-type dipole relaxation later incure. The time to reach 60% conversion determined by DSC correlates -withthe TA gelation loss peak, as would be expected from the Flory theory ofgelation, but there is no clear dielectric "event" at 60%. conversion. Thepeak in dielectric loss factor identified with the dipole relaxation :orrelatewith the TBA vitrification loss peak<

Aocesslon Fop

NTIS GRA&IDTIC TABUnannounced flJustificatlo

By

Distribution/

Availability Codes

IAvail and/or

Dit Special

C,,/4

SECURITY C€ASSIVICATION OF THIS P&GEt(What Da* Enewe)

A DIELEC!XIXC STUDY OF TE TIMi-TUPERATURE-TRANsFoRMhTION (T)DIAARAM OF DOEBA EPOXY RESINS CURED WITH DDS

Norman F. Sheppard, Ir., Michael C.W. Colnand Stephen D. Senturia

Department of Electrical Engineering and Computer Scienceand Center for Materials Science and Engineering

Masachusetts Institute of Technology. Cambridge, MA 02139

Abstract Keywords: Microdielectrcmetry,dielectric oce monitoring, time-

Mierodilootrcaetzy a"d dif- temperature-transformat ion (TTT)ferential scanning calorimetry have diagram, DGERA/DDSbeem used to study the isothermalente of lIN 83 with dianimodi-phenylsulfone (00). The resultsare eompared with published tor- 1. INTRODUCTIONsional braid analysis (TRk) studiesof the same system Using a time- Dielectric methods for moni-tompetature-tramsformation diagram. torial the cure of epoxy resinsThe dieleetrie measurements were have been established for nearly 25made at temperatures between 10o9C Years, dating back to the pioneer-asnd 2200C sing measurement fro- iag work of Delmonte (11 and War-quencies between 0.1 Is and 10,000 field and Petre* (2]. DramaticNs. The NSC was operated in as changes in the dielectric proper-isoithermal mode to determine extent ties of the material accompany theof conversion at temperatures be- transformation of the resin from atween 1370C end 1779C. Two machan- viscous liquid to a brittle solid.iane of eharge transport ar* ro- The simplicity of dielectric secas-spoasible for the observed dielec- urements have led to wide use bothtrio response; an ionic condutiv- in materials analysis and processity early In cure that decreases as control [3,4,51. A recent develop-the reaction proceeds, and a Debyo meat in the field has been the

*type dipole relaxation later in technique called Microdielectro-@wre. The time to reach 60% Gon- metry, which utilizes a miniatureversion determined by DSC corre- integrated circuit sensor to per-lates with the INA gelation loss form the dielectric measurementsoak, as would be epeted from the [6.71. The mall site of theFley theory of gelation, but there probe, the on-chip temperature sen-is so eloea dieleetric ,evento at sor, and the wide measurement fro-605 onversion. The peak in queasy range make licrodielectro-dieleetrie loe factor identified metry a promising alternative towith the dipole relaxation Gorre- conventional parallel-plate dielec-lates with the IMk vitrification trio measurements for process con-lose peak. trol applications.

F.4..4 -.W.'. . f*W \ 4

This pae: addresses the in-Ssiesoratio of dieleetrie data GELED R8ER

Obtanedduring tie auto of epoxy TM----resins in temus of the physicalchanges taking place daring caor. $GELIED GLASSe nature of the changes san be gdation

described with the aid of the time-temperature-trantfozzation (or =T) UCIJOdiagram fox themosets, developed 101T, .by Siliham C$I. The Ti? diagramdescribes the state of a resin T GLASSundergoing isothermal cure as afunction of eurs time Mad the iso- LOG TMEthermal our. temperature. The Ln-eralized TlT diagram in Figure 1shows four identifiable states of Fig. 1. Generalized Time-Tempera-the curing material: liquid, gelled tuse-Transforuation diasram showingrubber. ungelled 8lass. and gelled states of thermosettin$ resins as aglass. At eure temperatures be- function of aute time at isothermaltween the glass transition te mpers- cure temperatures. After Gilihamcue at Selation T ° and the 8i.ultimate Sias$ tranzition tempera-ture of the full cured material.T # ouring proceeds from thelf.uid state to a geled rubber, 2. E ERLMTALand finally to a gelled glass.selatiom is the transformation of 2.1 lIua.Lathe liquid to a rubbery Sel. Fur-ther eurti in the rubbery sel The EPON 825 resin was ob-state Increases the crossliak den- tamed from New England Resins andsity of the network until the glass Pigments, the diaminodiphenylsul-transition temperature of the fone curing agent from Dr. Garygelled rubber reaches the curing eagnau: at the Amy Materials andtemperature, at which point vitri- mechanics Research Center in water-fistics, i.e., the formation of a town. NA. The resin mixture waselled slag*. occurs, prepared in a S gram batch by celt-

ing together th* resin and theThe objective of this work is Curing agent until the curing agent

to enamsie "ielectrie measurements was dissolved. The mixture was keptin the cntext of the TT diagram refrigerated until use. No evi-with at goal of obtaining a better dee of reproecipitation of the DDSunderstaing of the physical meek- was observed.ains underlying the dielectricresponse. The system chosen forstudy Was a low molecular weight 2.2 Xietediela@tiometr-DNMA resin. Shell mx 825, czredwith diamiaodiphenylsulfou (DOS). Two types of Xicrodielectro-Tw TT dtsam for this system has metry sensors Were used in thisbeen published by Ians and GLlla a work. Nost of the results were(P1 based on torsional braid analy- obtained using the sensor describedsis (UA). The isothermal cure was in Reference 7. which incorporates.studied using Niorodieleotrometry an on-ehip semiconductor diodeever the temperature raune 100C to thermometer for combined dielec-IM0Ce and the kinetics of the trio/temperature measurements.chemical reaction were studied us- Some measurements were made withnLag differential scanning ealori- modified sensor in a flat Kaptoywmotay (WC) ever the range 1376C to Oribbons package., A photomicro-177eC. graph of the modified sensor

mounted in its package is shown inFigure 2. The packaged sensor is0.5 ma thick, 5 -8 wide and So cm

*- ' • .i'' "." " "'- ". . ' • "~ ~ " ." . '- .. . .. - .

Loss factors as low as 0.01 can nowt be measured. corresponding in typi-

.rm- cal cured resins to tan deltavalues on the order of 0.003.

.,;;" In the isothermal aure experi-ments reported here, the sensor andsooket assembly were placed into an

oven preheated to the cure tempera-*i ture. The sample temperature, as

measured by the on-chip temperature. sensor. came to equilibrium in

approximately 3 minutes. Thedielectric permittivity, loss fac-tor, and temperature were thenmonitored for the duration of thecure. An inert atmosphere was

-5 maintained by flowing dry nitrogenthrough the socket assembly.

2.3 Differential Scannint

* A Perkin-Elmer DSC-II differ-ential scanning calorimeter (DSC)was operated in an isothermal mode

to measure extent of conversicnFig. 2. Photomicrograph of micro- versus cure time at cure tempera-dielestrometer chip showing close- tures in the range 1370C to 1770C.upA sensox chip in flexible Zap- The heat flow output signal wastodW 'ribbom* package. measured by a digital voltmeter

interfaced to an INS calculator,which stored the data on cassette

long. This new sensor/package con- for later processing. The tempera-bination has been used for neat ture scale was calibrated using theresin studies and for cure monitor- melting points of indium and lead,in$ in laminates. The sensors weo and the enthalpy was calibratedprepared by heating them on a hot using an indium standard. Hermet-plate to approximately 509C and ically sealed aluminum sample pansplacing a small sample of the resin were used. The sample sizes werenixture (typically 20 mg) on the in the range of 5 to 15 mg. Thesensor electrodes. Rapid melting sample was placed in the cell atand flow produced good contact 77eC and then heated to the desiredbetween the resin mixture and the temperature at 160CC/mn. The re-electrodes. action exotherm was followed to

apparent completion. The areaThe instrumentation used for under the exotherm was taken to be

the Nierodielectrometry measure- the heat of conversion at thatmUtS has been significantly i- temperature.proved over that described previ-ously [,7]. The Rewlett-Packard The ouring of an epoxzy resin3575k Gais/Phase Moter used for at temperatures below the ultimateanalyzing the sensor response has T does not insure complete reac-bees replaced by a Fourier Trans- ron. To obtain extent of oouver-torm Digital Correlator of out own sion from an isothermal cure belowdesign. ith the correlator, the the ultimate T , the total heat ofeessible frequency range of the reaction of thl fully cured resinmeaurement is extended to 0.005 - must be known. This is obtained by10,000 3s and the sensitivity is dynamically scanning the tempera-ssifisatly improved compared to ture at a slow rate to insure com-that achievable with the 3P3575A. plete conversion. The total heat

of reaction was obtained from aramped oure at a rate of 2.$*C/-mL 3. RESULTSfrom 809C to 3000C. This ezperi-mest yielded a total heat of reac-tion of 99.5 eal/g, which is 99.2 3.1 MicrodielectrometrvhJ/mol (23.7 keal/mol), consistentwith results reported in the liter- FiSures 3a, 3b and 3c show theature for similar epoxy systems dielectric loss factor, a , the[111. Degree of conversion as a dielectric permittivity. s', andfunction of tine at each isothermal the loss tangent (or tan delta),cure temperature was obtained from which is the ratio s'/l , for thepartial integration of the reaction cure of EPON 825/DDS at 1000C meas-ezotheza, sezualized by the total ured at frequencies of 0.1, 1, 10,heat of reaction. 100, 1000 and 10,000 Hz.

The frequency dependence of000 - -the loss factor versus time indi-

Scates the superposition of two00 components, an ionic conductivity

0OHe and a dipole relaxation (7]. Early10 Min oure, a" is inversely propor-

tional to frequency, indicating aconductivity. The conductivitydecreases approximately ezponenti-ally with cure time, due to the

QCM1 Increasing viscosity of the resinas cure proceeds. This decrease isfollowed by a peak in a" with anamplitude of about 2. which occurs

>. earliest at the highest frequency.Tkis loss peak is due to the re-

112 striction of molecular dipoles inK= Oj the crosslinking resin.I_ The two processes of ionic

14 conduction and dipole relaxation0__ are also evident in the permittiv-

ity versus time plot of figure 3b.Early in cure, when the loss facto:

_ _0_ is large )10), the apparent per-

C. mittivity is extraordinarily largo(>20) due to charge accumulation atthe blocking sensor electrodes. Asthe loss factor drops, the blocking

OJ - electrodes are no longer important2 ;MtGHk and the apparent perittivity

Q(0- levels off at about 14, a valueconsistent with the polarizability

___________ _ of the resin. As curing proceeds,0 1o0 200 300 400 the permittivity then drops to a

TMOA T final value of approximately 4.Associated with this drop in a' isthe peak in a*, clearly evident inFigure 3a.

ig. S. Nierodieleetrometry results The data of Figures 3a and 3bof ewe of lM #I1/DDS at looeC. can be combined to yield tan deltaNeasuemest frequeseies 0.1, 1, 10, versus tine, as in Figure 3c. The100, 1000, 10000 as. characteristic features of the tan8) lee fasetor, as b) perittiv- delta versus time are a first peakIt?, e e) tan delta s"1I early in the cureo related to the

,, , . ,.. ..'. , .'.,-, . . . . .:..-.,.,.

I ionic conductivity and the blocking Woocelectrodes, sad a second peak lacerIa aure d"e to dipole relazation. )100 le

Me use of a wide ra zge of 1 -measurement frequencies helps il-lustatoe the extent to which theorientational, bility of polar 90-groups in the resin is being re-striated as aute proceeds. At each 0.01suceessive e peak, the reciprocalof the frequency for that peak 20gives a measure Of the earacteris-tic time requized for a dipole to Fl1 -Overcome the viscous drag of thesurrounding molecules and orient 12with the electzic field. From thetime at which the peak is fi rstobserved at 10Hz to the tine thepeak is observed at 0.1 Hz, this ,6characteristic time increases by 0_ _ _I _I

five orders of magnitude, fromhundreds of microseconds to teas ofseconds. 10 C.

Curia experiments similar to I Ithose Illustrated in Figure 3 wereperformed over a wide Caso of curetemperatures. Figures 4a. 4b and 1 "

4o show the cure temperature depe -dences of the loss factor, permit-tivity and tan delta versus curetime at a measurement frequency of Q001

10 Is. The characteristic features 0 100 200 300 400resulting from the icati coaductiv- TEM I -TESity snd the dipole relazatioa neck-anisms discussed above are apparent Fit. 4. Kicrodielectroetry resultsat all cure temperatures. As er- of cute of EION 82/DDS at 10 Rz.poeted, the ionic conductivity d- Care temperatures of 137, 147, 157,creases more rapidly with time sad 17, 177, 1870C.the dipole relazation oscur sooner a) loss factor. s" b) pemmittivity,at higher cure temperatures. due to s' c) tan delta s/elthe thermal activation of the cur-I & $ r e a c t i o .0

3.2 Diff erantia soa in

e extent of conversion ver-sue time results obtained from DSC I c

are show I igure S for cure 20temperatures ranging from 137C to o177C. There are three things to 0 60 120 180 240sate about these curves: (1) As 0,expeeted, the rate of reaction T (IJTES)Increases wit increasing cure te-Pen tue, due to the thermal aestiv- Fig. S. Extent of conversion basedsties of the esing reaction. (2) on heaot of reaction versus cureThee is as infleetion in the 02- time determined using isothermalteat of oversions versus time DMC measurement. Cure temperaturesewsvo refleotiag the autoatalytie 137 - 1770C.

nature of the epoxy-amine curing Figure 6 shows a TrT diagram of thereaction (11]. (3) The ultimate time to reach fractional conver-extent of conversion reached do- sions of 20%, 40% , 60% and 80%,creases with decreasing ce ten- taken from Figure 5, superimposedperature. These results are re- on the gelation and vitrificationlated to the TIT diagram and to the curves determined by Ens and Gill-dielectric data in the following ham (9]. The time to reach 60vesection. conversion coincides with the time

to reach gelation. The agreement isexceptionally good considering the

4. DISCUSSION assumptions made, and confirms thatthe systems are nominally identical

The objective of this study and valid comparisons can be made

was to relate dielectric measure- to the TBA measurements.

meats to the time-temperature-transformation (TT) diagram, so as

to obtain a better understanding of 250 L- Rthe physical processes underlyingthe dielectric measurement. The 200 (TTT diagram foe the EPON 825/DDS - a 0

system daterakLed using torsional 150 Abraid analysis (TBA) was reported A 20% A 3

by ins and Gillham (9]. 7BA mess- 0 40%uresa the free .)scillatory decay of lop-0 60%

a resin-impregnated lass-fiber o 0%braid in a torsion pendulum. Ma- 50 1 Iina in the damwing (logarithmic I0 100 1000 O00decrement) of the braid are inter- iMIE(MINTES)preted to iandi:ate the liquid torubber-transitLon associated with Fig. 6. Time-Tempeature-Transfor-

gelation and tie rubber to glass nation diagram for EPON 825/DDStransition assi:cated with vitrifi- system showing time to reach con-

cation of the esia. Because this versions of 20%. 40%, 60 and 80%.

is a dynamic masuremeat. the time Liquid to Rubber (gelation) andof occureace of the loss peak is Rubber to GlAss (vitrification)frequency depeadet. The TT? dia- transformations from Torsional

grasm indicates the time to reach Braid Analysis result4 of Enns andthe selation meventf and the tine Gillam (9].to reach the vitrification "e eat"as a function of the isothermal 50-cure temperature. L-R R --* G

The DSC results provide a wayOf verifying that a valid compari- - io'Do0son can be made between the dielec-trio measurements made in our labto the TBA measurements of Inns and 100- 0 10Hz

Giliham, which were done on nomin- a Q0H

ally the same material. According 50 I I

to the Flory theory of gelation 1 10 100 1000 0000E121o the gel point of a stoichio- TIME (AM TIES)metric system of difunctional mole-cules reacting with tetrafu ntional Fig. 7. Time-Temperatuce-Transfor-erosslinkers will sel at an extent ation diagram for EPON 825/DDSof reaction of 57.7%. If one as- system showing time to reach 10,000sums that the heat liberated by Iz and 10 Rz dipole relaxation lossthe epoxy ewing is due onay to the peaks. Liquid to Rubber (gelation)reaction of epozides with amines, and Rubber to Glass (vitrification)then, the time to reach a DSC son- transformations from Torsionalversion of 57.7% should coincide Braid Analysis results of Enns andwith the time to reach gelation. Gillham E9].

-V.. s- .0 1140-.,.i,.'..-'.' -",.-•."."- . .' . .,-''''" -.- ,,,,- .- "3",. ."."- -'-"- - " -

As discussed earlier, the peak evidenced in the a" values of Fig-in the dielectric loss factor at a ures 3a and 4a. However, the cross-fixed frequency indicates that the ing of the 60% conversion point,average dipole relaxation time has corresponding to divergence of thereached a value equal to the recip- macroscopic Shear viscosity at gel-rocal of the frequency. As curing ation, does not produce a singular-proceeds, the peak is observed at ity in conduction properties. Thesuccessively lower frequencies, so macroscopic shear viscosity re-the dipole relaxation time is in- fleets the ability of the moleculescreasing. We can plot these to undergo rearrangement over*evets",. the times to reach given macroscopic distances, while thedipole relaxation time, on the TTT transport of ions through thediagram. Figure 7 shows a TTT material requires moleculardiagram of the time to reach the rearrangement only on the atomicdipole loss peak at 10,000 Hz and scale. While the ionic conductiv-10 Ha. Slain superimposed on the ity decreases with increasingtorsional braid gelation and vitri- crosslink density because of anfication events [91. The countours increase in local or atomic-scaleof constant dipole relaxation time "microviscosity", the conductivityappear to parallel the vitrifica- does not cease at selation.tion curve of the TBA measurement.The vitrification of the resinoccurs when the mobility of the 5. CONQLUSIONSreactive groups in the resin be-.comes so small that the reaction This study has examined theessentially stops. The dipole re- curing of a typical epoxy/aminelaxation time is a measure of the system with dielectric measurementsobility of the polar groups of the made over the frequency range 0.1

matrix. It is not surprising then Hz to 10,000 Hz using Microdielec-that the vitrification event corre- trometry. Two mechanisms of chargelates with the rapid increase in transport are evident in the mess-the dipole relazation time. ured dielectric loss factor, an,

and permittivity, a'. An a" in-In contrast with the corres- versely proportional to frequency

pond&nee between the dipole peak and an abnormally large a' suggestad ThA vitrification, there is no that an ionic conductivity domi-clear "eveato in the dielectric nates the response early in cure.response corresponding to TBA Sel&- Later in cure, a peak in a" concur-ties. In a Previous paper E10] rent with a fall in a' haraeteris-studying the eure of DSUA with m- tic of a Debye type relaxationphoylene dismine (aUDI), we re- indicates that the mobility ofported that a pre-gelation dielec- polar groups in the matrix istric relaxation time approached rapidly decreasing. The dielectric"infi ite" values at salatio, i.e. measurements were compared with thevalues beyond the measurement capa- mechanical properties of the resinbility of the system. In Reference determined by torsional braid anal-10. that dieleetric relaxation time ysis [91, by plotting the time towas attributed to dipole orienta- reach the dipole relaxation losstion. We now uadorstand that it peaks as a function of cure temper-arlses from the ionic conduction aturoe on a time-toperature-trans-and blecking electrodes. Further- formation diagram. The cure ten-more, our improved instrumentation perature dependence of the time tonew permits even longer relaxation peak correlates with the "vitrifi-times to be observed. Tht data in cation event" observed by TBA,this Paper do n suppo- che as- because both events depend on thesig omet of a d in the mobility of the resin natriz.9reelation elsz:ios time to There is no evidence of a featureSolatie. The crossliaking leading in the dielectric response whichto the fomatioa of a rubbery Sl correlates with the "selation&es doorease the eoaduetivity, as event", because at that stage of

l ... . w ~~ ~~ ,, .. .-,.,-..-. ..-.. .... ,_,.,..-:, --,,,, -

cute the dominant Rode of charge 7. S.D. Senturia, N.F. Sheppard,transport is ionic conduction, H.L. Lee and S.B. Marshall,which is a measure of "microviscos- SANPE Journal, 19(4), 22ity' and not the macroscopic shear (1983)viscosity sensed by TEA.

8. 3.1. Giliham, "Torsional BraidAnalysis (TBA) of Polymers",in Develouments in Polymu

6. ACINOWLEDGENENTS Characterization - 3, J.V.Dawkins, Ed., Applied Science,

This work was supported in London, 1982, Chap. 5.part by the Office of Naval Re-search, and by the National Science 9. 3.B. Eans and 3.1. Gillh-im, 3.Foundation, under Contract ECS- Appl. Polym. Sci., 21, 2.5678114781. DSC measurements were (1983)made in the Polymer Characteriza-tion Central Facility of the MIT 10. S.D. Senturia, N.F. Sheppard,Center for Materials Science and H.L. Lee and D.R. Day, 3.Engineering (C01S), which is sup- Adhesion, L. 69 (1982)ported in part by the NationalScience Foundation under contract 11. 7.M. Barton, Polymer. _U, 603DI-81-19295. Device packaging was (1980)dome in the MIT MicroelectronicsLaboratory, which is also part of 12. P.3. Flory, Prinje._ 3_ i.C3E. We are indebted to Mr. L. Polymer Chemistry, CornellButter of Texas Instruments, University Press, London,Dallas. TX. and to Dr. D. Day of 1971.Niormet Instruments, Cambridge, MAfor their assistance in supplyingthe sensors used in this work.

8. BIOGRAPHIES

Norman F. Sheppard, Yr. is pros-7. REFEENCS ently a graduate student in the MIT

Department of Electrical Engineer-1. Y. Delmont, 3. Appl. Polyn. ing and Computer Science, having

Sol., (4,. 108 (1959) received an S.B. (1978) and an S.M.(1979) in Chemical Engineering from

2. 1.1. Varfield and N.C. Petree, NIT, and an S.M. (1981) and an E.E.7. Polym. Sci., 17, 305 (1959) (1981) in Electrical Engineering

from IT.

3. C.A. May, Proc. 2ls t AMfmI

Symposium, p. 803, Los Michael C.W. Coln is presently aAngeles, 1976 graduate student in the MIT Depart-

mnt of Electrical Engineering and4. L.D. DraSatakis and Z.N. San- Computer Science. He holds an S.B.

Jans. InsulationlCircuits, (1976) in Engineering Science andp. 27, Jan. 1978 Chemistry from the California

Institute of Technology, and an S.M5. V.I. Daugartmer and T. (1979) in Electrical Engineering

licker, SAMiIE ournal, 19(4), from MIT.S(1983)

Stephen D. Senturia is a Professor6. N.F. Sheppard, S.L. garyerick, of Electrical Engineering at MIT,

D.R. Dayjad S.D. Seaturia, having received an S.. in PhysicsProc. 26 S AM Symposium, p. from Harvard (1961). and a Ph.D. in6S, Los Angeles, 1981 Physics from MIT (1966).

* a --a . . ** :. . . .. ; .. ..- . ,.-.-. .-.. . ". . " . .'. - . - . . -

5P472-3/Al 472:GAN:716:dde1983 78u472-608

T5C1NICAL RPORT DISTRIBUTION LIST, GEN

No. No.Copies Copies

Office of Naval Research (2) Naval Ocean Systems Center (1)Attn: Code 413 Attn: Technical Library800 North Quincy Street San Diego, CA 92152Arlington, VA 22217

CN Pasadena Detachment (1) Naval Weapons Center (1)Attn: Dr. 1. 3. Marcus Attn: Dr. A. B. Amster,1030 East Green Street Chemistry DivisionPasadena, CA 91106 China Lake, CA 93555

Commnder, Naval Air Systems Command (1) Naval Civil Engineering Laboratory (1)Atta: Code 310C (I. Rosenwasser) Attn: Dr. R. V. DriskoDepartment of the Navy Port Hueneme, CA 93401Washington, DC 20360

Defense Technical Information Center (12) Dean William Tolles (1)Building S. Cameron Station Naval Postgraduate SchoolAlexandria, VA 22314 Monterey, CA 93940

Superintendent (1) Mr. John Boyle (1)Chemistry Division, Code .6100 Materials BranchNaval Research Laboratory Naval Ship Engineering CenterWashington, DC 20375 Philadelphia, PA 19112

U.S. Army Research Office (1) Scientific Advisor (1)Attn: CID-AA-IP Commandant of the Marine CorpsP. 0. Do 12211 (Code RD-1)Research Triangle Park, NC 27709 Washington, DC 20380

Mr. Vincont Schaper (1) DTNSRDC (1)DXNSUC Code 2830 Attn: Dr. G. Bosmajian,Ammapolis, RD 21402 Applied Chemistry Division

Annapolis, HD 21401

Naval Ocean Systems Center (1) Mr. A. M. Anzalone (1)Attn: Dr. S. Tememoto, Administrative LibrarianMarine sciences Division PLASTEC/ARRADCOMSam Diego, CA 91232 Bldg 3401

Dover, NY 07801

2 '7

1963 DL/413/83/01356A/413-2

UCHNICAL RIPORT DISTRIBUTION LIST, 356A

No. No.Copies Copies

Dr. £. Baer (1) Picatinny Arsenal (1)Department of Macramolocular Science Attn: A. K. Anzalone, Bldg. 3401Case Western Reserve University SIUPA-FR-M-DCleveland. Of 44106 Dover, NT 07801

Dr. M. koadhurst (2) Dr. 3. K. Gillhan (1)Bulk Properties Section Department of ChemistryNational Bureau of Standards Princeton UniversityU.S. Department of Commerce Princeton, NT 08540Washington, DC 20234

Dr. K. D. Pas (1) NASA-Lewis Research Center (1)Department of Mechanios and Attn: Dr. T. T. Serofini, MS-49-1Materials Scaries 2100 Brookpark RoadRutgers University Cleveland, OH 44135New Brunsawiek. M 0^003

Naval Surfae Weapons Center (1) Dr. Charles H. Sherman (1)Attn: Dr. 3. M. Augl, Dr. B. Nartman Code TD 121White Oak Naval Underwater Systems CenterSilver Spring, MD 20910 New London, CT 06320

Dr. 6. Goodman (1) Mr. Robert W. JonesGlobe Union Incorporated Advanced Projects Manager57S7 North Green Bay Avenue Hughes Aircraft CompanyMilwaukee, Wisconsin 53201 Mail Station D 132

Culver City. CA 90230

Prof. Eatsuo Ishida (1) Capt j. 3. Auborn, USNRDepartment of Maeromolecular Science ATST Bell LaboratoriesCase-Western Reserve University Room 6F-211Cleveland. OR 44106 600 Mountain Avenue

Murray Hill. NT 07974

Dr. John Lundberg (1) Dr. T. 3. Reinhart, Yr., Chief (1)School of Textile Engineering Composite and Fibrous Materials BranchGeorgia Institute of Technology Nonmetallic Materials DivisionAtlanta, GA 30332 Department of the Air Force

Air Force Materials Laboratory (AFSC)Wright-Patterson AFB, ON 45433

"r~ " ' %,°., ,.-. /.. .. '..'_.' .' " .. .. * q 5 .' . *j* *4 ; *%-'y. 9 . : . q

Dr. C. Giori (1) Dr. 3. Lando(111T Research Institate Department Of Macromolecular Science10 West 35 street Case Western Reserve UniversityChicago, IL 60616 Cleveland, OR 44106

Dr. JR. S. Roo (1) Dr. 3. A. Manson(1Department of materials Science Materials Research Centerand Metallurgical Engineering Lehigh UniversityUniversity of Cincinati Bethlehem, PA 18015Cincinnati, ON 45221

Dr. Robert a. cohens (1) Dr. R. S. Porter()chemical Engineering Department Department of Polymer ScienceMassachstts Institute of Technology and EngineeringCambridge. MA 02139 University of Massachusetts

Amherst, MA 01002

Dr. T. P. Conlon. Jr.. Code 3622 (1) Prof. Garth Wilkes(1Sandia Laboratories Department of Chemical EngineeringSandia corporation Virginia Polytechnic Institute andAlbuquerque. MM State University

Dlackbarg, VA 24061

Dr. Martin Kaufma 1 Dr. David Soong(1code 36506 Department of Chemical EngineeringNaval Weapons Center University of CaliforniaChina Lake. C& 93555 Berkeley, CA 94720

Prof. C. S. Paik Suag () Prof. Brian Newman(1Departmt of Materials science and Department of Mechanics andEngineering. loom 3-109 Materials ScienceMassachusetts Institute of Technology Rutgers. The State UniversityCambridge. MA 02139 Piscataway, NJ 08854

Dr. Curtis W. Frank (1) mr. Tor= S. PapirDepartment of Chemical Engineering Chevron Research CompanyStanford University 576 Standard AvenueStanford. CA 94035 Richmond, CA 94802

or. D. a. Uhlmams Honeywell Power Sources CenterDepartment of materials Science Defense Systems DivisionMassachusetts Institute of Technology 104 Rock RoadCambridge. MA 02139 Horsham, PA 19044

Professor A. MeegerDepartment of ChemistryUniversity of Californiaseats Barbara, CA 93106