The Significance of Glass Transition Temperature on Epoxy Resins for Fiber Optic Applications World Leader in Epoxy Systems for Fiber Optics 14 Fortune Drive + P.O. Box 567 * Billerica, Mass. 01821 + Tel 617.667<3805 * Telex: 94-7140 ©Epory Technology Inc, 1980SHE SIGHIPICANCE OF GLASS TRANSITION TEMPERATURE (O8 EPORY RESINS POR PIBER OPTIC APPLICATIONS ‘Thoms K, Tanslevich and Vanessa &. Moore pony Technology, ine. 1é Fortune Orive aastaact Since the development of fiber optics, epoxy resins have been red exclusively fer bending and potting op- tical fibers into a myriad of connectors and cast. geo- metries. Along with describing sone of these fiber ‘Optic epoxy applications, his paper will illustrate ore of the outstanding characteristics of this family ff bonding agents, the glass transition temperature: ‘hie thermodynamic ané thersonechanical characteristic Of epony resins Ls defines ax the midpoint of the ten perature region wherein the properties of epoxy resins Change feon those of @ glassy state to those of @ rub- arristatae Furthermore, this paper will elininate sone of the nystique surrounding 7g, a2 the glass transition ten perature is uaually abbreviated, ag wall as highlight Row the 79 19 « useful engineering tool for cpoxy resin aystens for optical fiber bonding. serpoouerroN Ae the use of thermoset plastic#, of which epoxy res~ Ing are a eesbery continues to encompass an ever-viden~ {ng range of applications in fiber optics, the need for reliable test methods in therefore intensifier Order to better control manufacturing and ageist in tthe selection of the proper epoxy for a given fiber optic design. ‘me glass transition temperature, or Tg as it is com ronly designated, is a phenomenon of thermoset epoxy Terine which can be measured indicating © cure, effecr Civenees of curing agents and softening point at ele- vated temperatures. ALL anorphous polymers, such as epoxy resins, exhibit a ty. Ae a polyser is heated, it undergoes a trans- ition similar co a classical thermodynamic second order ‘transition. This dnvelves a discontinuous change in a ‘Secondary thermodynanie quantity auch as the expansion coefficient or heat capacity. Therefore the 7g is of ten referzed to as a second order transition since It Ss characterized by a change in the expansion cosffi- cient and heat capacity. Moze simply the Tg is often referred to as the temperature region in which a poly- ser goes from a glassy state to a rubbery state. There fre several factore which influence the magnitude of the Cenperature region where the Tg occurs. ‘They are 1. Composition of the resin molecule 2. Grose-link deneity, 3. The polar nature of the resin molecule’s functional groves 4. Curing agent or catalyse 5. Cure tine and tesperature Emons oF DETERMINING 13 Historically one of the earliest methods for deter~ ‘ining 7g was by noting the heat deflection teapers ture of a cast epony bar. Such cest bar was placed pon a saddle vith a knom force inpinging on the ax 01822 center of the bar. This assembly vas imersed in of] Which vas heated at a Known rate. hen the epoxy bar Dont due to the tanpecatare of the ofl bath reach ‘the Tg of the epoxy, this tenperature was noted as”, the Heat Deflection Tenpersture or HDT cf he epoxy However, since the 19 {8 & change in secondary thermo- Gymanie quantities rich ae expansion coefficient or heat capecicy, equipment sch au the PerkinElmer if- ferentiai scanning Calorineters consoniy referred 20a Shee; or a Porkineriner Thermonechantean eysteny ferred to ae a tis, are Sdesily sulted for nearucing these thecmodynenic and thermomechanteal pelieer transitions ur aim in using a Perkin-Elmer Model Dsc-2, 1s to mea sure the change in heat capacity for determining the 79 of our epoxy materials. A sedel Dsc-2 block diagran is portrayed in Figure 1. Tn using the DSC-?, a transition such as softening, boiling, polymerization or crystallization occurs in ‘the sample material, an endothermic or exothermic Feaction takes place. ‘the change in power required to maintain the sanpie holder at the same temperature fs the reference holder (i.e., its programmed tempera fuse) during the transition is recorded aso peat: ‘the chart abecisea indicates the transition venpert ‘ture and the peak area indicates the total enersy transfer to oF fron the sample. FIOURE 1. Mode 02 Blok Pigure 2 shows a typical DSc-2 curve for EPO-TEX 260. with ite glass eransition temperature being defined ‘the themogran as measured fron the extensions of Pre land post transition baselines. More simply, the heat Capacity change assunes half the value of this change ‘upon going through the transition.Figure 4 is @ TS curve of EPOWTEK 301-2. Note that tthe 19 is defined as the midpoint of the pre and post ‘transition thermal expansions. Also note the marked change in slope after the Tq, indicating an increase in the coefficient of thermal expansion, mcausec af t 26 8 Se UO WO TO HO oT HO reve co) rae pore meee nts Etec ‘the slope reversal at the high temperature portion of the BS¢-2 thernogram in Figure 2 is indicative of ze- sidval cure or that additional curing. (polynerization) can take place. This means that the Tg can be in- creased with additional curing of the epoxy. The jerkin-Elaer Thermonechanical system (TMS), on the other hand, senses the change in the coefficient of ‘thermal expansion that occurs at the Tg or softening point of the epoxy under study. Figure 3 is a schenatic of the TNS analyzer QURE 3 Ani Schama reweo) Host analysts prefer the 05¢-2 in determining the 7g of epoxy resins since it is felt that measurenant of a Chermodinante change Ls more accurate than sensing & physical change a5 1s the case with the TMS. TZ AS A FUNCTION OF CURE ScHEDULE [As we have discussed, the Tg can be @ feacurenent of tthe degree of cure of an epoxy resin system. There~ fore, the cure tine and temperature can have 2 very, Granatic effect upon the 73 of any given epoxy 8 Table 1 and Pigure 5 {llustrate the relationship of (Gure schedule and 7g for EPOWTEK 301-2, a two compo- rent optical epoxy. FQURE Stave Cure Tine @ 8%C for EPOTEK 3012‘TABLE x ‘te Effect of Cure Schedule @ 80° on 119 of Ep0-7EK 201-2 cure Tine nin.) 6 90 120 18 ne. 24 he ‘Tables II, 111 and IV plus Figures 6 and 7 ‘liusteate the relationship of the Ty and cure schedule for EPO” FEC 330, 53ND and 377 which are popularly used for adhesion of fiber optic bundies and single Fibers into fond ferrules or connectors. These epoxies are more Righly cross-linked than EFOWTEK 301-7 and therefore achieve a greater Tg in a shorter period of time at a Given cure temperature. The Effect of Cure Schedule 1 Co 5 n 10 % 15 90 20 100 30 108 5 uo Py ne 30 no 16 ne. no ‘TABLE IIT The Effect of Cure Schedule 9 100% on the Ta of ERO-TEK 25380 ure Time (nin.) 3 (Pe) s 60 10 7 as ” 20 n 30 9 6 106 16 he. as ‘mane Vv ime Effect of Cure Schedule © 100% on the 7g of BrO-TeK 377 cure Tine (min) s a 10 a 13 53 20 se 30 7 “5 33 60 100 18 he 206 7 rt . | Figure 7 is a composite 0SC-2 thermogram depicting the ‘glass traneition temperatures of EPOMTEX 330, 3538D ‘and 377 when cured at 100°C for 18 hours to show that the maximum 7g was achieved in accordance with Figure 6 fang thereby no Fesidual cure is evident. of ar tgswre | reMPCO) able V and Figure 8 show the effect of cure schedule fon the 19 of EPOOTEK 360, 36057 and 360T. Here, the Basic epoxy EFOATEK 360 1s modified with an inert thix otropic agent in vatying anounts resulting in the pro- ucts, EPOSTER 360ST (slightly thixotropic) and EPO- EK 360F (Ehixotropic). it is observed that the Ty is only slightly affected <¢ at all by the anount of in (ert thixotropie agent added to the fundamental EPOWTEK 360 product. A einilar observation is made of silver ‘and gold filled EPO-TEX products. Therefore, inert Baditives have Little or no effect upon the Ty char- acteristics of an epoxy retin system.TABLE 5 Effect of Cure Schedule 2 100% on EFO-TEK 360, 3608F, and 3607 Tne (min.) —fg.360 36087. 36or 5 n a 78 10 %6 7 4 1 90 92 a1 20 100 38 90 30 109 106 90 s no 103 90 60 uo no 90 90 no no 90 18 ne ne uo 90 Lt INFLUENCE OF 1g ON PHYSICAL PROPERTIES ‘The selection of an epoxy system for bonding oF pot~ ting fiber optic assemblies must take into account: 1 Temperature environment the final assembly must tolerate 2 Mechanical performance i.e., flexing, vibration, Impact ete. 3. Chenical resistance required Im all applications, the physical significance of the ‘tq for optical epoxies io that, at temperatures above the Ty, the values of the physical properties such as Tensile Strengeh, Lap Shear and Adhesive bond strength and modulus are substantially reduced. ‘The coeffi ‘cient of thermal expansion (a), on the other hand, greatly increases. Figure 9 graphically illustrates how the lap shear strength of BPO-TEK 301 and 377 are affected by the T. oo x Hw © G1 wm WO 0 WO 1S WoT omens tt For applications where the epoxy bond will be subjectes to severe stresses, the Ty of the epoxy system selected should be above the projected use tenperature. in addition, at tenperatures above the tg for = par- tleular epoxy systen, the coefficient of thermal ex- pansion is 2 to ) tines greater than at teaperatures Below the 73. Figures 10 and 11 are THS plots of the coefficient of thermal expansion above and below the Tg for EPOWTER Sol and 377 | 01 a sa er ot Footie et PP A : | | 3 ale Tee co) ve a wo19) The discussion relating to Ty has concentrated on ‘epoxy systens that, upen curing, are hard and “elassy at room temperature. However, flexible epory systens fare available and have 79's below room temperature. ‘An example of this clage of optical epoxies is Ei mec. Bq AS A PumeR opric be Rnoving the sensitivity of the Ty to specific epoxy resin formulations and cure schedules allows for i= creased avareness on the part of the fiber optic Gesign and/or production personnel to evaluate the proper epoxy material and! processing schedule for their product. For example, consider the following cease which is cbserved frequently in fiber optic performance. Problem #1 A fsber optic bundle is potted with a low Ty epoxy, round and polished. Te will be used in conjunction With a 150# {Liunination source. After several hours the output of the Light guide is considerably iain ished. Upon examination of the input end of the Light guide, it is found that the efoxy has "cozed” out and partially covered the input face of the fiber optic bundle. Diagnosis ¢ solution Sufficient heat is generated by the i1lunination source that heats the input end of the fiber optic bundle toa tenperature greater than the Tg of the ‘epory. Bear in wind that at temperatures above 79, the coefficient of thermal expansion (a) of the epoxy is 2 to 3 tines greater above than below the 79. iso, at temperatures above the 7g the epoxy becomes soft and rubbery. In addition, the more significant effect of @ is above the Tg which is calculsted a @ x AT. ‘The combination of these property trans= Sesone caused the epoxy to expand beyond its original Ginension and obscures the optical fibers. The 6o- lution is simply to select an epoxy sysven with a 7g closer to or greater than the use temperature: ‘This is further exemplified by the following calculations: Asbient Temperature = 20% Assume 2 ute tenperature of = 120% original epoxy selection, EPO-TEK 301 at a cure eched- tg + 48%, @ below T9 = 50 x 10 9 = 125 in/in/e ‘Total calculated epoxy expansion EPO-TEX 301 in/in/%C, @ above (2) Betow 7g = (48% ~ 20%) x 50 x 106 kn/in/Pe 214 x 2074 bavi (2) Roove 1g = (120% ~ 46°C) x 125 x 10° snvin/e = 90x 1074 Safin Toral calculated - 1.6 2= 10.4 x 10°? invin Alternate epoxy selection EPO-TEK 377 at a cuze ached~ vie of 120% for 1 hour 1g ~ 106°C, a below tg = 62.5 x 1076 in/in®e, a above infin? tq = 125 x 10 ‘Total calculated epoxy expansion EPO-TEK 377 (3) Below 19 = (106% ~ 20%) x 62.5 x 107% = 53.7 x 10" infin (4) hove 13 = (120% = 106%) x 125 x 10° 217.5 x 10" infin Total calevlated - 2.642 71x 10 tnvin AAs can be seen there is significantly les expansion with the EpO-72 377 system. Furthermore, Figure 8 shovs the effect of 75 upon the Lap shear of adhesive bond strength of EFO-ER 301 and 373.” Notice that at the use temperature which in ig- Aificantly higher than the Tg ax in the case of EFO~ ‘EK 301 the Lap shear is greatly reduced. lowever, The Lap shear strength of SPOWTEK 377 at the proposed tse temperature ig still high, ‘This means that the EEOSTEK 377 should not “oose". The physical ‘yrovth” (ooae) of EPO-TEK 301 indicates a combination of ad~ hesive bond failure and’ expansion. susane ‘To sumarize, ve have seen the following: 1. The tg of an epoxy has a profound impact on coefficient of thermal expansion (@), lap shear Strength and modulus of optical epoxies. 2. The value of the 15 is strongly affected by the cure schedule and epoxy formulation chosen. 3 The 7g provider an excellent indication of the completeness of cure. 4. the 1g can be used as a valuable tool in design- ing fiber optic systens.