Polymer Journal, Vol. 9, No. I, pp 95-100 (1977)

Solid State Properties of Poly(spiro[2,4]hepta-4,6-diene)

Hiroaki SUGIMOTO,* Motowo TAKAYANAGI, and Toyoki KUNITAKE**

Department of Applied Chemistry, Faculty of Engineering, Kyushu University, Fukuoka 812, Japan.

(Received October 6, 1976)

ABSTRACT Poly(spiro[2,4]hepta-4,6-diene) (PSHD) is considered to behave like a rigid rod on the basis of the unusually large index (a=l.7) in the viscosity-molecular weight relationship. Thus, some solid state properties were investigated with PSHD samples which contained different ratios of the 1,2- and 1,4-addition units. The thermo-mechanical curve of solvent-cast films showed that there were two tan ii peaks below 300°C. The /3 peak was located at about 130°C (110 Hz), and affected little by the polymer composition. On the other hand, the position of the a peak shifted linearly from 200°C to 300°C with increasing fraction of the 1,4-unit, and the peak height decreased simultaneously. A loss peak associated with thermal decomposition and/or crosslinking was found at temperatures above 300°C. The TGA curve revealed that the thermal decomposition of PSHD takes place at about 300°C in air and at about 400°C under nitrogen. An exothermic peak was found in the same temperature range in the DSC curve. The crosslinked specimen showed no loss peak in dynamic measurements, which reflected immobilization of the molecule by crosslinks. The PSHD film did not show any indication of crystallinity. KEY WORDS Poly(spiro[2,4]hepta-4,6-diene) / Rigid Polymer / Thermogravimetric Analysis / Thermomechanical Analysis / Visco­ elastic Property /

One of the authors has reported that spiro­ characteristically with counteranions when tri­ [2,4]hepta-4,6-diene, a one-step derivative of phenylmethyl salts were employed as initiator. 2 cyclopentadiene, undergoes facile polymerization The relation between intrinsic viscosity and with a variety of cationic initiators. 1 •2 The number-average molecular weight is represented polymer consisted of the 1,2- and 1 ,4-addition by eq 2. evi­ structures, as inferred from spectroscopic [7J]:~;uene =4.5 X 10-s ._M,,1.7 (d//g) ( 2) dence. The cyclopropyl ring did not open during the course of the cationic propagation, It is remarkable that the power index of M,, in contrast with radical copolymerization. 3 in the Mark-Houwink viscosity equation was 1. 7, which is much greater than those of the usual flexible polymers with the random coil

Ph3CBF, conformation. High index values such as this 0 -76"C ( 1 ) have been reported only for rodlike polymers 4 1.2-unit like collagen and helical polypeptides. In fact, SHD 1,1,-unit PSHD in the space-filling molecular model the spiro­ cyclopropyl ring exerts large steric hindrance The content of the two structural units varied toward the conformational mobility. * On leave from the Central Research Laboratory In this paper, some solid state properties of of Sumitomo Chemical Industries Co., Ltd. this peculiar polymer are reported for the poly­ ** Department of Organic Synthesis, Kyushu mer samples of different structures. The effect University. of the polymer composition on the tan o peak in

95 H. SUGIMOTO, M. TAKAYANAGI, and T. KUN!TAKE the thermo-mechanical analysis was studied, and as follows. After flaming and cooling under thermally-resistant properties produced by the reduced pressure, the polymerization flask was molecular stiffness were measured by the TGA filled with dry nitrogen gas. Solvent and a and DSC methods. catalyst solution were added through serum caps and the flask was cooled to about - 76°C EXPERIMENTAL in a dry ice-methanol bath. Then a monomer solution was slowly added with stirring. The Materials polymerization was stopped by adding a pyridine SHD was prepared by a modification of the solution, and the polymer was recovered by published procedure. 3 A typical procedure is pouring the reaction mixture into methanol. as follows. In a 500 ml three-necked flask PSHD films were cast from benzene solution equipped with a mechanical stirrer, a thermo­ and kept under reduced pressure for a week. meter, and a graduated dropping funnel were placed 198 g (2 mol) of 1,2-dichloroethane, 320 g Instruments of 50-wt % aqueous sodium hydroxide (4 mol), Gas chromatograms were obtained with a and 4 g of trioctylmethylammonium chloride Shimadzu GC-3AH instrument: Silicone GE (phase transfer catalyst). The mixture was SE31, 70°C, carrier gas H2. Infrared and NMR stirred at temperatures below 10°c, and cyclo­ spectra were obtained by using Perkin-Elmer pentadiene (160 g, 2.4 mol) was added dropwise Model 337 and JEOL Model JNM-MH60 over 2 hr. Then the dropping funnel was re­ spectrometers, respectively. Thermomechanical placed by a reflux condenser and the mixture measurements were carried out with a Rheovi­ was heated at about 60°C for 1 hr. The organic bron dynamic viscoelastometer Model DDV-IIB; layer was separated and the aqueous layer ex­ the frequencies employed were 3.5, 11, and tracted with ether. The combined organic layers 110 Hz. The heating rate was 1-2 °C/min. were dried over anhydrous sodium sulfate over­ Thermogravimetric analysis was carried out on night and distilled under reduced pressure. The a Cahn-RG-electrobalance, with a heating rate synthetic yield of SHD was about 75%, as of 5°Cjmin. Differential scanning calorimetry estimated by gas chromatography. The distillate was performed by a Rigaku Denki Thermoflex 5°C/min. containing unreacted 1,2-dichloroethane was instrument. The heating rate was a Rigaku purified by repeated distillations: bp 65°C X-ray diffraction was obtained by Denki Rota Unit RU-3. (144mm) [lit. 5 57°C (lOOmmHg)]. The purity of the final fraction was better than 99%, as RESULTS AND DISCUSSION confirmed by gas chromatography. Polymerization initiators and solvents, as listed Polymerization in Table I, were prepared and/or purified by Table I gives the polymerization results. The 1 2 the previous methods. • polymerization was fast in all cases. The poly­ Procedures mers obtained were soluble in many solvents, The polymerization of SHD was carried out such as chloroform, benzene, toluene, 1,2-di-

Table I. Polymerization of SHD at - 76°C Con- 1,4-unit Initiator, Monomer, Time, version, ['l ]:~; uene • in polymer, No. mol// Solvent min mmol// % dl/g % BF2O(C2Hs)2 3.0 0.25 CH2Ch 15 90 1.2 42 2 SnCJ4-TCA• 0.25 0.25 CH2Cl2 10 90 3.0 38 3 (CsHs)sC+BFr 0.45 0.45 CH2Cl2 180 95 4.0 47 4 (CsHs)3C+BF4- 0.13 0.30 CH2Ch-toluene 10 90 4.8 70 (1 : 1) 5 (CsHs)sC+BF4- 0.07 0.32 CH2Ch-toluene 10 92 4.4 84 (3: 7) • Trichloroacetic acid.

96 Polymer J., Vol. 9, No. I, 1977 Solid State Property of Poly(spiroheptadiene)

chloroethane, tetrachloroethylene, and carbon u tetrachloride, and were white powders of fibrous 0: ;-, materials. Intrinsic viscosities of 1.2 and 4.0 dl/g, lJJ ; \ :r: / \ ol­ , \ for example, correspond to Mn=22,000 and X I \ , \ 47,000, respectively, based on eq. 2. The content / / I / \ of the 1,2- and 1,4-addition units was determined r / \ /· \ from the relative peak area of the olefin protons / / \ in NMR spectra, as described previously.2 It is ~/ ul clear that the polymer structure can be varied ::,= ------ffi extensively by changing the polymerization :r: § conditions. z lJJ X-Ray Diffraction 0 JOO 200 300 400 The X-ray diffraction pattern of PSHD cast TEMPERATURE ( ° C) films (samples: No. 1, 3, and 5 in Table I) Figure 1. DSC curves of PSHD in N2 atmosphere showed halos in the cases of both the perpendic­ (solid line) and in air (broken line): Sample, No. ular and parallel irradiation of the X-ray beam 1 in Table I. to the surface of a film. Neither orientation nor crystallinity was detected for the roll-extended 100 ...... -=""-'--__:',, \ PSHD (room temperature, by a factor of two) \ \ and for the sample uniaxially drawn (220°C in \ 80 \ a silicone oil bath, by a factor of five). The l \ PSHD molecule is composed of two different lJJ 5 60 structures. Presumably, the 1,2-unit will prevent in lJJ straightening of rigid segments composed of the 0: !;; 40 1,4-unit sequence. The probable existence of <.'.) cis and trans linkages between the structural 20 units makes the situation more complex. These I \ I irregularities must interfere with the crystalli­ \ zation of PHSD. 0 200 400 600 800 1000 TEMPERATURE ( °C) Thermal Properties Differential scanning calorimetry (DSC) curves Figure 2. TGA curves of PSHD in N2 atmosphere (solid line) and in air (broken line): Sample, No. of the polymer (sample: Table I, No. 1) are 1 in Table I. shown in Figure 1. The curve obtained under nitrogen atmosphere showed a slight shift of the base line to the endothermic side (downward) lioH band (3,580 cm-1) were newly observed in at about 190°C, while that in air (broken line) IR spectra of the sample annealed at 200°C. showed a deviation upward from the base line at The decomposition was extensive above 300°C. the same temperature range. Consequently, the The organic residue remained even above 500°C PSHD molecule appears to gain some molecular under nitrogen atmosphere, whereas no residue motions at temperatures above 190°C. The ex­ was found in air at the same temperature. othermic peaks at 351 °C under nitrogen and at Dynamic Viscoelasticity 362°C in air might be due to the decomposition Figure 3 shows the temperature dependence reaction and/or crosslinking. of the mechanical loss tangent, tan o, of PSHD Figure 2 shows thermogravimetric analysis (Sample: Table I, No. 1 and No. 5) measured (TGA) curves under nitrogen gas (solid line) at 110 Hz. There was no peak from -160 to and in air (broken line). The oxidation of the +100°C. The peaks were observed at 133°C polymer seems to take place at about l50°C, as (/3 peak), 246°C (a peak), and 335°C (decom­ reflected by a slight increase in the sample position peak). The PSHD sample with a higher weight. The lic=o band (1,700 cm-1) and the 1,4-unit content (84%, Sample No. 5) gave a

Polymer J., Vol. 9, No. 1, 1977 97 H. SUGIMOTO, M. TAKAYANAGI, and T. KUNITAKE

First the value of activation energy of the a 80 peak is estimated based on the WLF equation, 0 0 03 0 which has been shown to be widely applicable 0 0 0 to the glass-forming flexible molecules. The 0 0 WLF equation is given by6 0 0 0.2 0 0 0 logar=-C1g(T-Tg)/(c2g+T-Tg) ( 3) 0 (.0 0 C: 0 where ar is the shift factor when T g is taken !!l 0.1 I",O:J~0:J:i§l% as the reference temperature and c1g and c2g are • ~--.J. constants characteristic of the polymer species.

In the conventional expression, c1g= 17.44 and

0 100 200 300 c2g=51.6 are employed as the general values TM'perature (°C) applicable to all kinds of polymers forming Figure 3. Tan i5 vs. temperature for PSHD samples: glassy states. (0) 42.2% of the 1,4-units (sample, No. l); Ce) The activation energy for relaxation at any 84.35'0 of the 1,4-units (sample, No. 5). temperature T above Tg, ilHa, is evaluated by differentiating eq 3 with respect to the recipro­

10:}'T ('K-'l cal temperature as follows: 2.44 2.52 2.60 2.68 ilHa=2.303Rc1gc2gT2/(c2 g+T-Tg)2 ( 4)

100 If we assume that Tg of PSHD corresponds to the temperature at which the tan i'J curve deviates -;;- r from the background, ilHa of the a peak at its :5 peak temperature is evaluated by eq 4. The (J ...... 10 method for determining T g was described first _g"' by Hideshima et al.,7 and is conveniently used for PSHD samples with various copolymer com­ positions. Its principle is based on the assump­ 1.96 2.04 2.12 tion that the beginning of any thermal motion !Oo/r ('K-1) associated with the micro-Brownian movement Figure 4. Transition map of a and f3 dispersions is reflected in the rise of the tan i'J curve of the (log/max vs 1/T): Sample, No. 2 (e, C,); No. 3 a peak. For example, the a peak temperature co, e). and the Tg value are read from the tan i'J curve of Sample No. 5 in Figure 3 as 270°C and smaller tan i'J value above 170°C, and the a peak 200°C, respectively. Substitution of these values shifted to the higher temperature side when into eq 4. results in ilHa(250°C)=70 kcal/mol. compared with the sample with lower 1,4-unit This value is smaller than the observed value content (42%, sample No. 1). This difference of 86 kcal/mol but not inconceivable according may be explained by assuming that the sequence to the WLF equation, especially when Tg is of the 1,4-unit is more rigid than that of the 1,2- determined by the above method. unit and that the increase in the former unit The basic assumption of the WLF equation suppresses the motion of molecular chains to a is that the increase in the free volume fraction large extent. It is suspected that the stiffness and the conformational change of the main of the PSHD molecule is mainly governed by chain take place by thermal activation. In the the 1,4-structure. case of the PSHD chain, however, such con­ Figure 4 shows the transition map of the a formational change is difficult to assume, as far and {1 peaks, which was prepared by changing as space-filling model can suggest. the measuring frequency. The mean activation Another possible interpretation of the molecu­ energy of the a peak is 86 kcal/mol and that lar process of the a peak is the torsional oscil­ of the f1 peak is 30 kcal/mol for both samples. latory motion around the molecular axis. Such

98 Polymer J., Vol. 9, No. 1, 1977 Solid State Property of Poly(spiroheptadiene)

of PSHD is located at a temperature higher 300 than those of other polymers. On the other hand, the nature of the (3 peak seems to be

0 0 attributable to the local relaxation mode of the E main chain associated with sequence shorter x 250 than that associated with the a peak. A more Ill after E definite conclusion can only be presented l,Q detailed studies on the stereoisomerism are C Ill the following facts are i:: performed. At present 200 available for discussing the nature of the (3 peak. As seen in the thermomechanical curves of Figure 3, the (3 peak temperature and its relax­ 0 20 40 60 80 100 ation magnitude are not sensitive to the copoly­ 1,4-COMPONEN T ( 0/o) mer composition or the difference in the main­ Figure 5. Temperature for tan i5 maximum (110 chain configuration. When the sample was Hz) vs. fraction of the 1,4-unit. exposed to benzene vapor, there was no change in the magnitude of the (3 peak. Therefore, a mechanism was assigned by Yamafuji, et al.. 8 this peak is conceivably governed by the intra­ to the (3 relaxation of the polymers having no side molecular factor. Soaking of the sample in chains such as poly(vinyl chloride), poly(ethylene sulfuric acid produced crosslinking, but the (3 terephthalate), etc. There will be various modes peak was observed at almost the same temper­ of the oscillatory motion around the chain axis. ature. This suggests that the molecular motion The lower modes of motion are considered to associated with the (3 peak is confined within a be influential along some lengths of the main very limited range of the molecule. chain which include several monomeric units. The PSHD film (No. 1) turned bright black, In other words, the a peak temperature and its probably due to development of the conjugated relaxation magnitude are affected by the copoly­ olefin system, when it was subjected to heat mer composition. This expectation is supported treatment up to 360°C. The a and (3 peaks by the results in Figure 5. disappeared almost completely after the heat Figure 5 shows the relationship between the treatment. a peak temperature in the tan a curve and the Figure 6 shows the storage modulus (E') curve 11 fraction of the 1,4-units in the polymer. The and the loss modulus (£ ) curve for PSHD linear relation determined by the least squares method is given by eq 5:

Ttan o, rnax( 0 C)= l.03F + 197 ( 5) where F denotes the percent fraction of the 1,4-units. Using this equation, the temperature of the tan a peak for PSHD with 100% 1,4-units is evaluated as 300°C for the a peak. The poly­ mer with 100% 1,4-units might crystallize and become more heat-resistant. The decrease in UJ the intensity of the a-peak at higher 1,4-unit LL.I contents, as seen in Figure 3, will also produce better heat resistance. The activation energy of the (3 peak (30 kcal/ JOO 200 300 400 mol) is higher than those of the subpeaks of Temperature (°C) the usual vinyl polymers (10-20 kcal/mol).9 Figure 6. Temperature dependence of E' and E" This difference is not unreasonable when it is after heat treatment up to 360°c Ce) and for the taken into account that the (3 peak temperature original sample (Q).

Polymer J., Vol. 9, No. I, 1977 99 H. SUGIMOTO, M. TAKAYANAGI, and T. KUNITAKE samples before and after the heat treatment. The 3. 0. Ohara, C. Aso, and T. Kunitake, Nippon storage modulus varied little above 300°C for Kagaku Kaishi (J. Chem. Soc. Jpn., Chem. Ind. the heat-treated sample. This means that the Chem.), 1973 602. thermal oxidation and intermolecular crosslink­ 4. M. Kurata, M. Iwata, and K. Kameda, in ing inhibit every molecular motion. Conse­ "Polymer Handbook," Vol. IV-I, J. Brandrup and E. H. Immergut, Ed., lnterscience Publishers, quently, a peak at 335°C in the tan a-temper­ Inc., New York, N. Y., 1965. ature curve of the original sample is assigned 5. K. Alder, H.J. Ache, and F. H. Flock, Chem. to the initiation of such a crosslinking reaction. Ber., 93, 1888 (1960). This temperature corresponds to the peak located 6. J. D. Ferry, "Viscoelastic Properties of Poly­ at 350°C in the DSC curve. The progress of mers," John Wiley & Sons, Inc., New York, the crosslinking reaction was reflected in the N. Y., 1960. rise of the modulus curve above 300°C, as seen 7. N. Saito, K. Okano, S. Iwayanagi, and T. in Figure 6. Hideshima, "Molecular Motion in Solid State Polymers," Solid State Physics, Vol. 14, Aca­ demic Press, New York, N. Y., 1963, p 494. REFERENCES 8. K. Yamafuji and Y. Ishida, Kolloid-Z. Z. Polym., 183, 15 (1962). 1. 0. Ohara, C. Aso, and T. Kunitake, J. Polym. 9. N. G. McCrum, B. E. Read, and G. Williams, Sci., 11, 1917 (1973). "Anelastic and Dielectric Effects in Polymeric 2. T. Kunitake, T. Ochiai, and 0. Ohara, J. Polym. Solids," John Wiley & Sons, Inc., London, Sci., Polym. Chem. Ed., 13, 2581 (1975). 1967.

100 Polymer J., Vol. 9, No. 1, 1977