Spectroscopic Investigation of Local Dynamics in Polybutadienes

Chapter 5

Spectroscopic Investigation of Local Dynamics in Polybutadienes

Downloaded by MONASH UNIV on November 10, 2015 | http://pubs.acs.org Publication Date: November 30, 1987 | doi: 10.1021/bk-1987-0358.ch005

L. Monnerie, J . L. Viovy, R. Dejean de la Batie, and F. Lauprêtre Laboratoire de Physicochimie Structurale et Macromoleculaire, associe au Centre National de la Recherche Scientifique, ESPCI, 10 rue Vauquelin, 75231 Paris, Cedex 05, France The fluorescence anisotropy decay technique and C spin-lattice magnetic relaxation have been used to investigate the local dynamics of bulk polybutadienes at temperatures at least 60K higher than the glass-rubber transition temperature. The orientation autocorrelation function required to account for the experimental data agrees with the Hall-Helfand expression proposed for the local dynamics of polymer chains. In addition, the elementary motions, observed via the considered spectroscopic techniques, have a temperature dependence of their correlation times which is close to the prediction of the William, Landel, Ferry equation, proving that they are involved in the glass-rubber transition phenomenon of the polymer. 13

The l o c a l dynamics of polymers i n solution have been extensively studied during the l a s t decade. On the other hand, for polymer melts many questions are s t i l l unanswered, such as, for example, the nature of the orientation autocorrelation function (OACF) which i s involved, and the relationship of the segmental motions occuring at high frequency i n the melt with the elementary processes responsible for the glass-rubber t r a n s i t i o n . Spectroscopic techniques such as fluorescence anisotropy decay (FAD), and C s p i n - l a t t i c e magnetic relaxation (T^ NMR) are well suited to investigation of the l o c a l dynamics of polymer melts. In t h i s paper we present results recently obtained i n our laboratory on bulk polybutadienes. MATERIALS AND METHODS Polybutadiene Firestone "Diene 45 NF" with the following microsjjructure, % c i g = 37, % trans=59, % 1-2 = 12, and Mn = 1.7.10 , Mw = 4.1.10 was used as a matrix for FAD experiments. An anthracene labeled polybutadiene with the same microstructure was synthetized by anionic polymerization as 0097-6156/87/0358-0046$06.00/0 © 1987 American Chemical Society

In Photophysics of Polymers; Hoyle, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1987.


5.

Local Dynamics in Polybutadienes

MONNERIE ET AL.

47

previously reported for polystyrene ( J[_ ). Monof unctional " l i v i n g " chains of molecular weight 10 were prepared and deactivated by 9,10-bis (bjromomethyl) anthracene. The resultant chains of molecular weight 2.10 contain a dimethylene anthracene fluorescent group in t h e i r middle, as shown i n Figure 1. Because the fluorescence t r a n s i t i o n moment (represented by a double arrow i n Figure 1) l i e s along the chain axis, fluorescence anisotropy w i l l be i n s e n s i t i v e to the rotation of the label around the 9,10 axis of the anthracene moiety and r e f l e c t only the motions of the backbone. The labeled polybutadiene (1 % by weight) and Diene 45 NF (99 %) were mixed i n solution, then the solvent was removed by evaporation. The o p t i c a l density of the films was less than 0.1 to avoid energy transfer and reabsorption. The polymer films were placed i n a c e l l s p e c i a l l y designed for elastomers ( 2 ). A sequence of molding and stoving operations i n an argon atmosphere was carried out to avoid bubbles, to ensure perfect adhesion at the interfaces and to relax stresses. Polybutadiene Bayer Uran the microstructure of which i s % c i s = %trans - 1, % 1-2 - 1 and Mn=323000, Mw=1096000 was used for C NMR experiments. FAD measurements were performed on the cyclosynchrotron LURE-ACO at Orsay (France). The apparatus i s described elsewhere ( 3. ). The continuous spectrum from the synchrotron allowed for the matching of the l a s t absorption peak of the dye (401 nm). This excitation wavelength was selected by a double holographic grating with 2-nm s l i t s , and the most intense emission peak (435 nm) was selected by a single holographic grating with 2-nm s l i t s . This procedure greatly improves the rejection of spurious fluorescence, which i s one of the major problems of fluorescence studies i n bulk polymers. The negligible l e v e l of spurious fluorescence was checked by using blank samples of the unlabeled matrix. The transmission of the emission monochromator was calibrated i n both polarization directions with a 0.5% precision. The polarized emission spectra and the apparatus response (recorded at emission wavelength) were sampled with a 0.12 ns channel width by the single-photon counting technique. Thanks to the s t a b i l i t y of the pulses, the short-time l i m i t of the experimental window i s about 0.1 ns. The upper l i m i t , imposed by the r e p e t i t i o n rate of the pulses and the l i f e t i m e of the dye, i s ^ ns. S p i n - l a t t i c e magnetic relaxation times T, on C nuclei were measured at 25.15 MHz and 62.5 MHz using Jeol PS 100 and Bruker WP 250 spectrometers respectively. DMS0-d6 was used as an external lock. T was obtained from the 180°-t-90° sequence with an accuracy of 7 %. 1

FLUORESCENCE ANISOTROPY DECAY Under the action of a suitable electromagnetic f i e l d , polarized along the Ρ direction as shown i n Figure 2, the absorption of l i g h t i s proportional to the scalar product of the incident e l e c t r i c f i e l d and the t r a n s i t i o n moment. In the same way, the emission of l i g h t i s proportional to the scalar product of the direction of the analyzer and the t r a n s i t i o n moment. Thus, excitation of an i s o t r o p i c population of fluorescent species by polarized l i g h t generally creates a temporary anisotropic population of excited molecules. Molecular motions | ^ ρ ^ Ϊ ς ^ ^ (^filMlësf ^èâ$t$° °^ ^ ρΓ0

tT

a

Library 1155 16th St., N.W. In Photophysics of Polymers; Hoyle, C., et al.; Washington, D.C.Society: 20036 ACS Symposium Series; American Chemical Washington, DC, 1987.

e

c

t


PHOTOPHYSICS OF POLYMERS

Figure 1. Polybutadiene with anthracene i n the middle of the chain. The t r a n s i t i o n moment of anthracene i s represented by a double arrow.

Figure 2 . Polarized absorption and fluorescence emission polarizer; A, analyzer.


: P,

5.


MONNERIE ET AL.

49

the polarization of the reemitted fluorescence l i g h t . The quantity of interest, the fluorescence anisotropy, r , i s defined as : r = (I - I )/(I v

h

v

+

2 I )

(,)

h

where I and I, correspond to fluorescence i n t e n s i t i e s f o r analyzer d i r e c t i o n p a r a l l e l and perpendicular respectively to the v e r t i c a l polarization of the incident beam. In t h i s expression ( I + 2 1^) represents the t o t a l fluorescence i n t e n s i t y . The fluorescence anisotropy emitted at time t w i l l progressively decrease as a function of time and f i n a l l y reach a zero value. A complete analysis of the phenomenon ( 4^ ) shows that the evolution of r ( t ) as a function of time i s d i r e c t l y proportional to the second moment of the OACF of the t r a n s i t i o n moment : y


y

r ( t ) = M ( t ) =< 3 c o s 2

2

0 ( t ) - 1 >/2

(2)

where Θ(t) i s the angle through which the vector under consideration (the t r a n s i t i o n moment i n FAD) rotates during time t , the brackets mean an ensemble average. The time i n t e r v a l t during which the evolution of I ^ C t ) can_^>g recordgd i s d i r e c t l y related to the fluorescence l i f e t i m e (10 to 10 s ) ; e x p e r i m e n t a l l y M2(t) can be obtained u n t i l approximately 10 s (100 ns). It should be pointed out that FAD i s v i r t u a l l y unique i n i t s a b i l i t y to obtain the OACF. A detailed study of the FAD of anthracene labeled polybutadiene i n a matrix of Diene 45 NF has been performed i n the temperature range 240K-353K ( 2 ). At 335.7K, the FAD curve shown i n Figure 3 i s p a r t i c u l a r l y suitable f o r comparison with various molecular models of l o c a l chain dynamics. Indeed, at t h i s temperature the decay i s not too f a s t , i n such a way that an accurate comparison with the models can be made over the f u l l time window available, but nevertheless the FAD curve goes close to zero. I t appears that the i s o t r o p i c model (single exponential f o r the OACF) does not f i t the data and that OACF expressions s p e c i f i c a l l y proposed f o r polymer dynamics (for a review of them, see Ref. _3 and _5 ) are required. The s p e c i f i c behavior of polymer chains i s due to the chain connectivity requirement and the description of l o c a l dynamics requires consideration of the following features: - "elementary motions" with a c h a r a c t e r i s t i c time, τ^ι which diffuse along the chain sequence and lead to a non-exponential short time term i n the OACF. - a damping or a truncation of t h i s d i f f u s i o n , which yields i n the OACF an exponential loss, with a c h a r a c t e r i s t i c time τ. (τ > τ ) For bulk polybutadiene, the best f i t , shown i n Figure 2r, i s reached i n using the OACF expression proposed by HALL and HELFAND ( 6 ): M (t) 2

where I

= e x p ( - t / T ) exp( - tl τ )I (t/T ) 2

χ

Q

1

(3)

represents the modified Bessel function of order 0.




50

100

200

300

400

CHANNEL

Figure 3. Comparison of the best f i t obtained from the Hall-Helfand expression reconvoluted by the measured instrumental function (excitation pulse) (continuous l i n e ) with the experimental anisotropy (dots) of labeled polybutadiene at 335.7K. The excitation pulse i s plotted as a dash-dot l i n e (arbitrary scaled). The upper graph represents the weighted residuals.


51


5. MONNERIE ET AL.

It i s worth noting that t h i s expression also gives the best f i t for FAD curves of labeled polystyrene i n solution ( 5_ )· In bulk polybutadiene, the r a t i o ^2^ \ ^ Y s i g n i f i c a n t l y with temperature and remains rather high (=30). This implies that the processes responsible for the damping i n polybutadiene are slow compared to the d i f f u s i v e ones. The temperature dependence of has been studied i n the range 240K- 353K. I t i s interesting to compare t h i s dependence with the prediction of the well-known William, Landel, Ferry time-temperature superposition equation ( 7. ) which can be written as : τ


log τ

cj/ ( T - T J

= Cte +

Τ

Too

=

T

g

o e s

n o t

v a r

(4)

- c:

where Tg i s the glass-rubber t r a n s i t i o n temperature (172K for Diene 45 NF) and C ^ , C 2 are phenomenological parameters taken from low frequency v i s c o e l a s t i c measurements (For Diene 45 NF, ϋ = 11.2 and C = 60.5K ( ]_ ) ) . In Figure 4, the c o r r e l a t i o n time i s plotted versus Ι/ζΤ-Τ^) and the dash-line corresponds to the W.L.F. equation. The experimental data satisfactory f i t a linear dependence with a slope f a i r l y close to the predicted one. This means that the l o c a l reorientation processes observed by FAD are involved i n the glass-rubber t r a n s i t i o n phenomenon. It should be pointed out that a comparison of the FAD curves of labeled polybutadiene to the ones obtained for free 9,10 ( d i a l k y l ) anthracene probes i n a Diene 45 NF matrix led to the conclusion ( 8 ) that the l o c a l motions observed by the FAD experiments performed on labeled polybutadiene involve about 6 butadiene units. 8

8

g

χ

2

CARBON 13 N.M.R. 13 The l o c a l dynamics of polymer chains can be studied by C NMR through measurements of J^e s p i n - l a t t i c e magnetic relaxation time, T^. The spin of a given ^C relaxes by dipolar relaxation mechanism with the bonded Η so that the corresponding T^ i s related to the reorientation motion of the involved CH internuclear vector through the following expression ( 9. ) : 9 J. = 1 4

IT

2

2

Ύ Ύ _9_H 10

1[ r CH

J ( u )

H

ω

C

j

+

3

j(

w

) C

+

6

J ( ( i )

+ uO]

(5)

oo

J (ω) = /

ο

M„(t) e " I

i u ) t

dt

where h i s the Planck c o n s t a n t , and γ„ are the gyromagnetic r a t i o s of C and ''H respecγjvely and ^ are the Larmor resonance frequencies of C and ^H, J(o)) i s the spectral density at the frequency ω of the OACF of the CH motion, and r „ i s the length of the CH bond. r



Θ Diene 45 NF

Θ


UR A N

4

5

6

7

8

10 /(T-TOO) 3

Figure 4 . - Logarithmic plot of the correlation time τ- vs ( T-T r . a/ determined from FAD f o r labeled polybutadiene i n a Diene 45 NF matrix. ^ b/ τ determined from C T measurements f o r polyButadiene Uran. Dashed l i n e s correspond to Equation 4 using i n each case the c o e f f i c i e n t s appropriate to the considered matrix. 1

m

1


5.

MONNERIE ET AL.

53

Local Dynamics in Poly butadienes


A J

The main interest of C NMR spectroscopy i s that i t allows one to investigate the l o c a l motions performed by the various groups of the polymer chain. Thus, i t y i e l d s more detailed information than other techniques. On the other hand, to derive the correlation times of the motions from the T, measurements, i t i s necessary to use an "a p r i o r i " expression 01 the OACF, M2(t). In the case of polymer melts at temperatures higher than (Tg+60K), the segmental mobility i s high enough to get narrow resonance l i n e s i n the NMR spectrum and thus to perform the T. measurements on the same spectrometers which are commonly used for high resolution NMR i n l i q u i d s . Polybutadiene Uran has been studied at 25.15 MHz and 62.5 MHz i n the temperature range 224 Κ - 358 Κ. The nT, values obtained for CH and 0Η« groups of Uran, wh^re η = 1 tor CH and η = 2 for CH are shown as a function of 10 /T i n Figures 5 and 6. On the same figures are plotted the best f i t s obtained by using the Hall-Helfand OACF expression. It c l e a r l y appears that the experimental values of T.. at the minimum are much higher than those expected from t h i s OACF expression. I t should be noticed that the comparison would be even worse with the single exponential OACF corresponding to an i s o t r o p i c motional model. Secondly i n the whole temperature range, the r a t i o T (CH)/T (CH ) i s d i f f e r e n t from the expected value of 2, i t s value l i e s around 1.5 at both observation frequencies. These two discrepancies can be accounted for by considering, i n addition to the damped d i f f u s i o n of elementary motions along the chain sequence (which leads to the Hall-Helfand expression for the OACF), either a Brownian motion of the CH bonds inside a cone of half angle Θ, or a s p e c i f i c anisotropic motion of the c i s 1,4 sequences of the polybutadiene chain ( £ ). In both cases, the r e s u l t i n g OACF can be written as : 2>

1

1

2

M ( t ) = (1 - a) exp( -

exp( - t / T ^ I ^ t / ^ )

2

+

a exp(

-

t/τ

) exp(

-

t/τ

)I (t/T

}

)

Q

χ

(6)

where the parameter a describes the contribution of the anisotropic motion with correlation time τ . The best f i t s (shown on Figures 5 and 6) are obtained with τ /τ > 150, τ /τ > 600 and a(CH) = 0.27, a(CH" ) = 0Î46? These values indicate that the additional anisotropic motion (characterized by τ ) i s very fast as compared to the elementary segmental motions (described by T- ). In the explored temperature range, τ should be snorter than J§ s at 224K and 10 s at §58K. Recent C T, measurements performed on various bulk polymers with low Tg, such as polyisoprene, polyisobutylene, polyvinylmethylether and polypropyleneoxide, have shown that the experimental value of T^ at the minimum i s always much higher than the prediction from Hall-Helfand expression. Thus, t h i s discrepancy cannot be assigned to a s p e c i f i c motional anisotropy of the c i s 1,4 sequences of the polybutadiene chain. I t seems more l i k e l y to assign i t to a fast anisotropic motion of the CH bonds inside a cone. In the case of polybutadiene, the corresponding half angle of the cones should be 26° and 36° for CH and CH 2

?



54


H

« 4.5

« 4

35

« 3

ô-

~~

10 /Τ 3

13 Figure 5. Temperature dependence of C T^ at 25.15 MHz and 62.5 MHz for CH of Uran polybutadiene. Dots are experimental data. The dash-dot l i n e s correspond to the Hall-Helfand f i t (Equation 3). The dashed l i n e s are the best f i t obtained by using Equation 6. nT, (s)

4.5

4

3.5

3

10 /τ 3

13 Figure 6. Temperature dependence of C T^ at 25.15 MHz and 62.5 MHz for C ^ of Uran polybutadiene (same representation as i n Figure 5 ) .


5.

MONNERIE ET AL.

Local Dynamics in Poly butadienes

55

respectively. This result i s quite satisfactory, the constraints due to the size and the r i g i d i t y of the double bond decreasing the amplitude of the l i b r a t i o n motion of the CH bonds r e l a t i v e to that of the CH groups which are linked to the chain backbone through single bonds only. Another interesting feature deals with the high value obtained for the ratioT2/T^. This means that i n high c i s content polybutadiene, the damping of the d i f f u s i o n of the elementary motions along the chain sequence i s very weak. The correlation^time derived from T^ values of Uran i s plotted vs (T-T ) on Figure 4. A satisfactory agreement i s found with the preïiction of Equation 4, using the appropriate c o e f f i c i e n t s ( T = 101K, C j = 11.35, C = 59.6K ( J_ ) ) . Thus, the elementary motions of the polybutadiene chain which are investigated by C NMR are involved i n the glass-rubber t r a n s i t i o n phenomenon. 2

8

8


œ

COMPARISON OF THE CORRELATION TIMES The spectroscopic techniques which have been used to investigate the segmental motions of bulk polybutadiene i n a temperature range f a r enough from the glass-rubber t r a n s i t i o n show that these motions are involved i n the glass-rubber t r a n s i t i o n phenomenon. Thus, i t i s interesting to go further i n comparing the absolute values of the obtained correlation times. At a given (T-T ) ^ i n t e r v a l , for example 192.3K corresponding to a value of 5.2 for lO^/CT-TJ, the correlation time of the anthracene l a b e l i n the middle of the chain i s 6.5x10 s. At the same temperature the value of , ^ corresponding to the_çjementary motions responsible ror C relaxation i s 4.4x10 s. Thus, there i s about 2 decades of difference i n the correlation times. It i s worthnoting that such a difference i n correlation time values does not originate from the difference of microstructure between Diene 45 NF and Uran. Indeed, preliminary measurements have shown that for the same (T-T^) i n t e r v a l , similar T^ values are found for bot^polymers. °° As the C T^ measurements do not imply any labeling, the corresponding correlation time has to be considered as the actual c h a r a c t e r i s t i c time of the elementary motions of the polybutadiene chain. The longer correlation time observed with FAD for anthracene labeled polybutadiene might originate either from an i n e r t i a l e f f e c t of the anthracene group or, more l i k e l y , from the larger volume which i s required to rotate s i g n i f i c a n t l y the anthracene t r a n s i t i o n moment lying along the l o c a l chain axis, compared to the volume which i s involved i n the conformational changes leading to C s p i n - l a t t i c e relaxation. Such a statement i s consistent with the result ( 8 ) that about 6 monomer units are involved i n the motions observed by FAD for labeled polybutadiene. œ

τ

CONCLUSION In t h i s paper we have investigated the segmental motions of bulk polybutadiene, i n a temperature range h i g ^ r than (Tg + 60K), by using fluorescence anisotropy decay, and C spin-lattice magnetic relaxation time, T . 1



56


The FAD technique has shown that the orientation autocorrelation function of l o c a l motions i n a polymer chain does not correspond to the single exponential associated with i s o t r o p i c rotation. On the other hand, the Hall-Helfand expression which accounts for the chain connectivity requirement leads to a very satisfactory agreement. Thus the segmental dynamics of polymer melts should be described by elementary motions undergoing a damped d i f f u s i o n along the chain sequence, i n a similar way to that i s found i n polymer solutions. This indicates that the same types of conformational changes are involved i n both cases, the slowing down of these jumps i n polymer melts, compared to the case of polymer solutions, arises from the intermolecular c o n s t r a i n ^ . In addition to these elementary motions, the C T^ measurements have shown that a much faster anisotropic motion occurs, corresponding to a l i b r a t i o n motion of the CH bonds inside a cone. Recent C T^ measurements ( 10 ) performed on bulk polybutadiene containing both c i s 1,4 and trans 1,4 sequences have shown a lower mobility of the trans sequences r e l a t i y ^ to the c i s units. This result i l l u s t r a t e s the main interest of C studies, allowing one to reach more detailed information on the dynamics of each group i n the polymer chain and thus yielding a way of setting up a relationship between the chemical structure and the dynamic behavior of bulk polymers. Another important result deals with the temperature dependence of the correlation times of the elementary motions, which agrees f a i r l y well with the prediction of the William, Landel, Ferry equation, using the phenomenological c o e f f i c i e n t s obtained from low frequency v i s c o e l a s t i c measurements. T M s means that the elementary motions which are observed by FAD and C T^ are involved i n the glass-rubber t r a n s i t i o n phenomenon. In t h i s paper, the considered spectroscopic techniques have been applied to bulk polybutadiene. Similar studies on other polymer chains are under progress.

LITERATURE CITED 1. 2. 3. 4. 5. 6. 7. 8. 9. 10.

Valeur, B.; Monnerie, L. J . Polym. Sci. , Polym. Phys. Ed. , 1976, 14 , 11. Viovy, J.L.; Monnerie, L . ; Merola, F. Macromolecules , 1985, 18 , 1130. Brochon, J.C. In Protein Dynamics and Energy Transduction ; Shin'ishi Ishiwata,Ed.; Taniguchi Foundation: Japan, 1980. Monnerie, L. In Static and Dynamic Properties of the Polymeric Solid State ; Pethrick,R.A.; Richards, R.W., Ed.; NASI Series: Reidel Publ., 1982. Viovy, J.L.; Monnerie, L . ; Brochon, J.C. Macromolecules , 1983, 16 , 1845. Hall, C.K.; Helfand, E. J.Chem.Phys. , 1982, 77 , 3275. Ferry,J.D. Viscoelastic Properties of Polymers , 3rd ed.; Wiley: New-York, 1980. Viovy, J.L.; Frank, C.W.; Monnerie, L. Macromolecules , 1985, 18 , 2606. Gronski,W. Makromol. Chem. , 1977, 178 , 2949. Dejean de la Batie, R. Dr.Sc. Thesis, Université Pierre et Marie Curie, Paris, 1986.

RECEIVED March 13, 1987 In Photophysics of Polymers; Hoyle, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1987.

Spectroscopic Investigation of Local Dynamics in Polybutadienes

Recommend Documents