Structure-Property Relations in Polymers - American Chemical Society

6

Downloaded by UNIV LAVAL on July 11, 2016 | http://pubs.acs.org Publication Date: May 5, 1993 | doi: 10.1021/ba-1993-0236.ch006

Fourier Transform IR Spectroscopic Analysis of the Molecular Structure of Compatible Polymer Blends M . Sargent and J. L . Koenig* 1

Department of Macromolecular Science, Case Western Reserve University, 10900 Euclid Avenue, Cleveland, O H 44106-7202

Compatible polymer blends are formed by the combination of two or more polymers to produce a homogeneous, single phase mixture. The ability of polymers to form compatible blends requires that the interaction between the unlike polymer chains be at least as favorable as the self-association of each of the component polymers. Compatible polymer blends can therefore be characterized by the formation of intermolecular interactions between specific chemical groups of the component polymers. Infrared spectroscopy has been used to study polymer blend compatibility on the molecular level. The presence of molecular interactions is determined by examining the differences between the blend spectrum and the spectra of the component polymers. These spectral differences include shifts in the absorption frequency, increases in the band width, and changes in the absorptivity of the bands. This chapter reviews the application of spectral data processing techniques, such as factor analysis, difference spectroscopy, and least squares curve fitting, that characterize these interactions.

TTHE USE OF POLYMER

BLENDS i n a variety o f scientific a n d industrial

applications has b e e n clearly established over the last several decades. N e w p o l y m e r i c materials w i t h superior c h e m i c a l a n d physical properties m a y b e Current address: Miami Valley Laboratories, Proctor and Gamble Company, P.O. Box 397707, Cincinnati, OH 45239-8707. * Corresponding author

1

0065-2393/93/0236-0191$08.25/0 © 1993 American Chemical Society

Urban and Craver; Structure-Property Relations in Polymers Advances in Chemistry; American Chemical Society: Washington, DC, 1993.

192

STRUCTURE-PROPERTY

RELATIONS IN

POLYMERS


developed b y selectively c o m b i n i n g two or m o r e homopolymers to p r o d u c e a compatible b l e n d . T h e ability o f polymers to f o r m compatible blends requires that the interaction between the u n l i k e p o l y m e r chains b e at least as favorable as the self-association of each of the component polymers. Reviews o f the initial studies o f p o l y m e r compatibility w e r e p u b l i s h e d b y F l o r y ( J ) a n d T o m p a (2). M o r e current reviews o f compatible p o l y m e r blends i n c l u d e those b y B o h n (3), R o s e n (4), Krause (5), B a r l o w and P a u l (6), O l a b i s i et al. (7), a n d P a u l a n d N e w m a n (8). Incompatible p o l y m e r blends can be characterized as having components that exist i n isolated phases or domains. I n contrast, the components o f a compatible b l e n d are intimately m i x e d to f o r m a single homogeneous phase. C o m p a t i b i l i t y o f a p o l y m e r b l e n d is dependent o n many variables, i n c l u d i n g b l e n d composition, temperature, a n d m e t h o d o f mixing. Phase separation o f compatible blends m a y also occur over a p e r i o d o f time. Relatively few techniques have b e e n developed that are capable o f examining intermolecular interactions between p o l y m e r components o n the molecular level. Infrared spectroscopy has b e e n successfully a p p l i e d to detect such molecular interactions t h r o u g h analysis o f changes i n specific bands o f the b l e n d spectrum. Reviews o f the application o f infrared spectroscopy to the study o f p o l y m e r b l e n d compatibility have b e e n p u b l i s h e d b y C o l e m a n a n d Painter ( 9 , 10). A variety o f spectral data processing techniques have b e e n developed to assist i n the characterization o f the interactions ( I I ) . T h e presence o f intermolecular interactions can be c o n f i r m e d f r o m factor analysis results. T h e spectrum o f the interaction can be isolated using difference spectroscopy and the degree o f interaction i n the b l e n d can be quantified b y least squares curve fitting. T h e specific type o f interaction may also be identified f r o m differences between the b l e n d spectrum a n d those o f the pure components. S u c h changes i n c l u d e shifts i n the absorption frequency, increases i n the b a n d w i d t h , a n d changes i n the absorptivity o f the bands.

Thermodynamics of Polymer Blends T h e degree o f miseibility o f a mixture is d e t e r m i n e d by the G i b b s free energy o f mixing, A G , according to the equation m i x

AG · = AH

. -

mix

TAS

mix

. ^mix

w h e r e AH is the enthalpy o f mixing, AS is the entropy o f mixing, a n d Τ is the temperature o f the mixture. AG can vary w i t h the composition o f the overall mixture i n several ways, as shown i n F i g u r e 1. T o achieve complete miseibility over a l l compositions, two conditions must be satisfied: AG must be less than zero a n d the second derivative o f the free energy w i t h respect to the two components, δ AG /b$ , must be greater than zero ( φ is the v o l u m e fraction o f each component i n the mixture). B l e n d s that can be mix

mix

mix

mix

2

mix

2



6.

SARGENT A N D KOENIG

193

Compatible Polymer Blends

Figure 1. Possible free energy of mixing diagrams for binary mixtures. (Reproduced with permission from reference 6. Copyright 1991 Society of Plastics Engineers, Inc.) described b y curve A are completely i m m i s c i b l e because they have a free energy o f mixing that is always positive. B l e n d s that f o l l o w curve Β satisfy b o t h conditions and are therefore miscible over all compositions. C u r v e C describes blends that are partially miscible. W h e n the b l e n d is c o m p o s e d o f compositions i n the central range, the free energy o f the system may be r e d u c e d b y separating into two phases whose compositions are given b y the two m i n i m a to the left a n d right o f center. T h e thermodynamic m o d e l most frequently used to describe the mixing o f polymers is the F l o r y - H u g g i n s theory (12), w h i c h assumes a lattice o n w h i c h the p o l y m e r molecules can be arranged. Scott (13) a p p l i e d the F l o r y - H u g g i n s theory to mixtures o f dissimilar polymers a n d obtained the f o l l o w i n g equation for the G i b b s free energy o f mixing: RTV

ΦΒ

ΦΑ In

Φ

Α

+

|

—

I In

φ

Β

+

ΧΑΒΦΑΦΕ

H e r e V is the total v o l u m e , V is a reference v o l u m e taken as close as possible to the molar v o l u m e o f the smallest p o l y m e r repeat unit, φ and φ are the v o l u m e fractions o f polymers A a n d B , respectively, X a n d X are the degrees o f polymerization o f polymers A a n d Β i n terms o f the reference volume. χ is the interaction parameter that is related to the enthalpy o f interaction o f the p o l y m e r repeat units, each o f molar v o l u m e V , R is the universal gas constant, a n d Τ is the temperature o f the mixture. r

Α

A

Β

B

Α Β

r

Because a m i x i n g process increases the randomness or disorder o f the system, the change i n entropy is always positive. H o w e v e r , the entropy o f m i x i n g is a f u n c t i o n o f the molecular sizes o f the component polymers a n d approaches zero as the degree o f polymerization increases. Therefore,


194

STRUCTURE-PROPERTY

RELATIONS IN POLYMERS


because the entropy o f mixing is very small for p o l y m e r mixtures, the free energy o f mixing is essentially d e t e r m i n e d b y the sign a n d magnitude o f the enthalpy o f mixing. T h i s enthalpy o f mixing depends o n the energy change associated w i t h nearest neighbor contacts d u r i n g mixing. T h e free energy o f m i x i n g w i l l b e negative i f the enthalpy o f m i x i n g is either negative o r zero o r i f it is positive b u t less than the entropy t e r m . A negative enthalpy o f mixing indicates that heat is evolved d u r i n g m i x i n g a n d occurs i f the component polymers attract each other more than they attract their o w n k i n d . S u c h a situation is encountered very infrequently i n p o l y m e r blends because the intermolecular energy is d e t e r m i n e d mostly b y dispersive interaction i n w h i c h the energy o f a contact p a i r o f dissimilar polymers is approximated b y the geometric mean o f the self-association energies o f mixing. Therefore, negative enthalpy can only occur i f strong, nondispersive interactions are f o r m e d between the different p o l y m e r c o m p o nents.

Determination of the Molecular Structure Using IR F r o m the previous discussion o f the thermodynamics o f p o l y m e r blends, it is clear that p o l y m e r compatibility c a n only o c c u r w h e n a strong molecular interaction occurs between the t w o components. This intermolecular interac tion must be greater than the h o m o p o l y m e r intramolecular interactions o f the components. O n e o f the aspects o f i n f r a r e d spectroscopy that is w i d e l y k n o w n is the ability to detect differences i n molecular structure a n d interactions. It is o n this basis that i n f r a r e d is used to study p o l y m e r compatibility i n blends (14). F o r such studies, it is necessary to generate a n d interpret the " i n t e r a c t i o n " spectrum. This interaction spectrum is the difference between the spectrum o f the b l e n d a n d the spectra o f the component polymers, a n d it reflects the difference i n the molecular interactions constituting the b l e n d . F a c t o r analy sis methods can b e used to verify the interaction spectrum. Interpretation o f the interaction spectrum i n terms o f the molecular structure a n d interactions depends o n the system under examination. F a c t o r analysis is a mathematical procedure that determines the n u m b e r o f spectroscopically identifiable, linearly independent components i n a series o f mixtures. O n e o f the first applications o f factor analysis to the i n f r a r e d spectra o f mixtures was c o n d u c t e d b y A n t o o n et al. (15). A n excellent s u m m a r y o f the mathematical principles involving this procedure c a n b e f o u n d i n a review b y G i l l e t t e et a l . (11). Essentially, the n u m b e r o f p u r e components is f o u n d b y d e t e r m i n i n g the rank o f a covariance matrix [ C ] , w h i c h is the product o f the data matrix o f the spectra o f mixtures [ M ] m u l t i p l i e d b y its transpose [Mf: [C]

=

[M][MÎ


6.



195

T h e rank o f this covariance matrix is d e t e r m i n e d b y solving the eigenvalue problem


[C][E]

=

[E][X]

w h e r e [E] is the eigenvector matrix a n d [ λ ] is the diagonal eigenvalue matrix. Ideally, the n u m b e r o f p u r e components corresponds to the n u m b e r o f nonzero eigenvalues. H o w e v e r , experimental r a n d o m noise i n the data w i l l also produce nonzero eigenvalues. A theory o f error i n factor analysis was therefore developed b y M a l i n o w s k i ( 1 6 ) to determine the correct n u m b e r o f nonzero eigenvalues resulting f r o m t h e p u r e components. T h e difference b e t w e e n the b u i l t error-free data a n d the actual experimental results is expressed as the real error ( R E ) 1/2 RE

n(rap)

w h e r e m is the n u m b e r o f spectra, ρ is the n u m b e r o f p u r e components, η is the n u m b e r o f points p e r spectrum, a n d X is the eigenvalue. A n indicator function ( I N D ) c a n then b e calculated f r o m the equation RE IND

=

(m — p)

2

T h i s indicator f u n c t i o n attains its m i n i m u m value w h e n the correct n u m b e r o f nonzero eigenvalues has b e e n selected. F a c t o r analysis c a n b e a p p l i e d to a p o l y m e r b l e n d system to determine w h e t h e r a compatible or an incompatible mixture has b e e n f o r m e d . F o r a binary mixture that is incompatible, the results o f factor analysis w i l l indicate the presence o f only two components i n the b l e n d . H o w e v e r , for a compati b l e mixture an interaction w i l l occur between the two component polymers and its presence w i l l be i n d i c a t e d i n the factor analysis results as a t h i r d component i n the b l e n d . O n c e factor analysis has b e e n used to positively determine the compati bility o f a p o l y m e r b l e n d , difference spectroscopy (or spectral subtraction) can b e used to isolate the infrared spectral changes resulting f r o m the interaction between the component polymers. A detailed description o f the application o f digital subtraction to i n f r a r e d spectra has b e e n p u b l i s h e d b y K o e n i g (17). T h e infrared spectrum o f a compatible p o l y m e r b l e n d is actually c o m p o s e d o f contributions f r o m the component polymers plus an additional contribution resulting f r o m the intermolecular interactions f o r m e d between the components. T h e spectral contributions f r o m these interactions c a n b e identified using digital subtraction. This technique involves subtracting the


196

STRUCTURE-PROPERTY

R E L A T I O N S IN

POLYMERS

sealed spectra o f each o f the component polymers f r o m the spectrum o f the b l e n d . T h i s interaction spectrum results f r o m frequency shifts, b a n d broaden ing, a n d changes i n peak intensity. A f t e r the interaction spectrum o f a compatible b l e n d has b e e n isolated b y difference spectroscopy, least squares curve fitting can be a p p l i e d to deter m i n e the concentration of the components present i n the mixture. B l a c k b u r n (18) has developed a least squares m e t h o d that uses the p u r e component spectra to determine the relative amounts o f each component present i n the mixture spectrum. T h e fitting equation presented b y B l a c k b u r n is as follows: Ν

M

ER = Σ


y

i=l

i=l

j=l

M Si =

Σ xjRij J=l

w h e r e Ν is the n u m b e r o f data points i n each spectrum, M is the n u m b e r o f c o m p o n e n t spectra used i n the

fitting

spectral range o f the mixture, W

is a statistical w e i g h t i n g factor equal to the

{

inverse o f S , R {

ik

procedure, S

i

is the data for the

is the absorbance data for the i t h spectral element o f the

feth component spectrum, a n d Xj is the m u l t i p l i e r used b y the least squares p r o c e d u r e that gives the best fit o f the standard spectra to the mixture spectrum. It is f r o m the Xj values that the v o l u m e fractions o f the c o m p o nents i n the mixture are d e t e r m i n e d . A c c u r a c y o f the least squares curve fitting procedure can b e measured b y the m u l t i p l e correlation coefficient,

R, w i t h 1.0 corresponding to perfect

correlation. This coefficient can be calculated f r o m the following equation:

R

2

S ( M

C

- M

m

)

2(M -MJ 0

2

2

w h e r e M is the observed spectrum, M is the calculated spectrum, a n d M is the m e a n spectrum. T h e methods o f factor analysis, difference spectroscopy, a n d least squares curve fitting have b e e n a p p l i e d b y K o e n i g a n d R o d r i q u e z ( J 9 ) to the study o f compatible p o l y ( p h e n y l e n e oxide) ( Ρ Ρ Ο ) a n d polystyrene (PS) blends. T h e indicator function f r o m factor analysis reached a m i n i m u m value w h e n the n u m b e r o f components equaled three. These three independent components w e r e p r o p o s e d to be P S a n d two different P P O conformations, w h i c h is i n agreement w i t h results of a study b y W e l l i n g h o f f et al. (20). W e l l i n g h o f f et al. d e t e r m i n e d f r o m spectroscopic analysis that a strong interaction be tween the p h e n y l r i n g o f P S a n d the phenylene r i n g o f P P O was responsible Q

c


m

6.



197


for P P O conformational changes that o c c u r r e d u p o n b l e n d i n g w i t h P S . T h e interaction spectra o f P P O - P S blends o f various compositions were obtained b y difference spectroscopy a n d least squares curve fitting was t h e n a p p l i e d to determine the concentration o f the interaction spectrum i n the b l e n d spec t r u m : T h e interaction c o n t r i b u t i o n reached a m a x i m u m value i n blends having a composition of 30:70 P P O - P S . F a c t o r analysis a n d least squares curve fitting w e r e also a p p l i e d to determine that partial d e m i x i n g o f the blends begins w h e n the P P O - P S b l e n d are heated to 200 °C. Partial demixing was c o n c l u d e d because, although the indicator function still at tained a m i n i m u m value corresponding to three independent components, there was an overall decrease i n the contribution o f the interaction spectrum to that o f the b l e n d spectrum. A shift o f the carbonyl stretching peak to lower frequencies occurs for compatible p o l y m e r blends. L e o n a r d et al. (21) demonstrated that this shift i n the carbonyl peak i n a compatible b l e n d results f r o m localized interactions between the component polymers. O n e percent solutions o f the m o d e l c o m p o u n d m e t h y l acetate ( M A ) (a simple m o d e l for the ester group) were p r e p a r e d i n the m i x e d solvents h e x a n e - b e n z e n e a n d h e x a n e - o r t h o d i c h l o r o benzene ( O D B ) . T h e carbonyl stretching region o f the spectra o f the solu tions indicates only one b a n d , as seen i n F i g u r e 2. H o w e v e r , the second a n d fourth derivatives, as w e l l as the F o u r i e r self-deconvoluted spectrum, clearly show that the carbonyl peak actually is c o m p o s e d o f two bands. Similar results were f o u n d for M A dissolved i n various compositions o f h e x a n e - O D B . I n contrast, w h e n M A is dissolved i n a single solvent, a doublet i n the carbonyl peak cannot be detected. T h i s observation suggests, therefore, that the shift of the carbonyl stretching peak is caused b y localized solvent-solute interactions rather than b y a b u l k property o f the m e d i u m , such as the refractive index or dielectric effects. T h e strength o f the intermolecular interactions between M A a n d each o f the solvents varies w i t h each system a n d is reflected b y the different magnitudes i n the frequency shifts o f the carbonyl peak. B l e n d compatibility m a y also be studied through examination o f changes i n the w i d t h at half-height o f the carbonyl stretching peaks. A study o f various polyesters ( P E ) b l e n d e d w i t h poly (vinyl halide) ( P V X ) was c o n d u c t e d b y C o u s i n and P r u d ' h o m m e (22). T h e w i d t h at half-height o f the P E carbonyl b a n d increased i n the compatible P E - P V X blends, whereas no changes i n the w i d t h were detected for the incompatible blends. T h i s increase i n w i d t h for the compatible blends was attributed to the rigidity a n d r a n d o m c o i l confor mation o f the P E molecule. N o t all the P E carbonyl groups are favorably disposed to interact w i t h the P V X a n d w i l l therefore experience no change i n their vibrational frequency. B a n d broadening, therefore, results f r o m a distri b u t i o n i n the strength o f the interactions, ranging f r o m strong hydrogen bonds that p r o d u c e the greatest shift i n frequency to an absence o f any interactions, w h i c h results i n no frequency shifts.


STRUCTURE-PROPERTY

RELATIONS IN


198


POLYMERS


199



6.

1790

1770

1750

1710

1730

WAVENUVBERS ( c m * ) 1

(B) Figure 2. Continued. Key: Curves represent the FTIR spectrum (a); second derivative of the spectrum (b); fourth derivative of the spectrum (c); and self-deconvoluted spectrum (d). (Reproduced with permission from reference 21. Copyright 1985 Butterworth.)



200

STRUCTURE-PROPERTY


A F o u r i e r transform ( F T ) I R study o f b l e n d compatibility was c o n d u c t e d b y C o l e m a n a n d Z a r i a n (23) o n blends o f poly(e-caprolactone) ( P C L ) a n d poly (vinyl chloride) ( P V C ) . F i g u r e 3 shows the carbonyl stretching region o f the spectra o f the blends r e c o r d e d at 75 °C. T w o changes i n this peak occur as the concentration o f P V C is increased: a shift to lower frequency a n d a n increase i n the w i d t h at half-height. T h e change i n w i d t h at half-height as a function o f P V C concentration is p l o t t e d i n F i g u r e 4. T h e resulting S-shaped curve indicates that the magnitude o f the interactions o f the carbonyl w i t h P V C approaches saturation at a concentration o f approximately 4:1 P V C - P C L molar ratio, w h i c h corresponds to approximately 6 0 - w t % P V C . This satura tion effect was explained b y considering the relative lengths o f the structural repeat units o f P C L a n d P V C . A s s u m i n g a planar zigzag conformation, the - ( C H ) C O O - unit o f P C L is approximately 3.4 times as large as the - C H C H C l - unit o f P V C . F r o m this approximation it was d e t e r m i n e d that a molar excess o f about 4:1 P V C - P C L is necessary for saturation to occur. 2

5

2

F u r t h e r evidence that a change occurs i n the carbonyl structure o f P C L can b e seen i n F i g u r e 5. T h e 1 1 6 1 - c m peak i n the P C L spectrum has b e e n assigned b y K i r k p a t r i c k (24) as the result o f contributions f r o m C - O stretch i n g a n d O - C - H b e n d i n g vibrations. H o w e v e r , after b l e n d i n g P V C a n d P C L i n a 5:1 molar ratio, this peak shifts to 1165 c m . - 1

- 1

R o o m temperature studies w e r e also c o n d u c t e d o n the P C L - P V C blends, w i t h emphasis p l a c e d o n the presence o f a crystalline P C L component i n conjunction w i t h the amorphous components o f P C L a n d P V C . I n the carbonyl stretching region o f the blends, t w o peaks occur f o r the semicrystalline P C L . A peak at 1724 c m " is assigned to the crystalhne P C L c o m p o n e n t a n d a peak at 1737 c m results f r o m the P C L amorphous 1

-

1

Figure 3. FTIR spectra of PVC-PCL blends recorded at 75 °C for pure PCL (A), 1:1 (B), 3:1 (C), and 5:1 (D) molar PVC:PCL, resfectively. (Reproduced with permission from reference 23. Copyright 1979 Wiley.)



6.


201


19.00 W1%PVC Figure 4. Plot of the width at half-height of the carbonyl stretching frequency as a function of PVC concentration for PVC-PCL blends recorded at 75 °C. (Reproduced with permission from reference 23. Copyright 1979 Wiley.)

component.

F o r blends having P V C concentrations

P V C - P C L or greater, the peak at 1724 c m "

1

o f 3:1

molar ratio

cannot be detected, w h i c h

indicates that blends o f these ratios exist i n an essentially amorphous state. A slight shift o f the amorphous peak to a l o w e r frequency occurs as

the

concentration o f P V C is increased. These shifts are comparable to the shifts detected at 75 °C. Additionally, for the semicrystalline blends, the crystalline peak at 1724 c m

- 1

shifts to higher frequencies as the P V C concentration is

increased. B o t h frequency shifts for the crystalhne a n d amorphous carbonyl peaks indicate the

existence o f a specific interaction between

the

two

polymers that involves the carbonyl group o f P C L . A n o t h e r significant occurrence i n the carbonyl region o f the infrared spectra o f the blends r e c o r d e d at r o o m temperature is that for blends that contain a P V C - P C L molar ratio o f 3:1 or greater, the w i d t h at half-height o f the b a n d is identical w i t h i n experimental error to that o f the blends studied at 75 °C. This result indicates that b a n d w i d t h is not a f u n c t i o n o f temperature. A l t h o u g h C o l e m a n a n d Z a r i a n (23)

established that the interaction i n the

P C L - P V C blends involves the carbonyl group o f the P C L , the specific type o f interaction between

these two polymers was unclear. T w o types

of

interactions may be occurring: an interaction between the P C L carbonyl group w i t h either the α-hydrogen or the c a r b o n - c h l o r i n e b o n d o f P V C .




6.



203

V a r n e l l et al. ( 2 5 ) undertook further investigations of this system to deter m i n e w h i c h o f these two interactions is o c c u r r i n g .


N o shifts o f the C - C l peaks o f P V C w e r e observed i n the b l e n d spectra. B a s e d o n studies o f l o w molecular weight analogues c o n d u c t e d b y V a r n e l l et al. (25), initial conclusions are that the interaction does not involve the C - C l b o n d o f P V C . H o w e v e r , such a conclusion cannot be made because not all the C - C l groups can interact w i t h the P C L carbonyl groups a n d the unreacted groups may, therefore, hide any slight shifting. T o find conclusive evidence o f a specific interaction involving the P V C α-hydrogen, α-deuterated P V C was b l e n d e d w i t h P C L . S u c h an interaction c o u l d b e detected i n this system because the C - D stretching peak is w e l l separated f r o m the C - H peak. B a s e d o n results obtained f r o m studies o f the m o d e l c o m p o u n d system m e t h y l acetate a n d deuterated c h l o r o f o r m , the C - D stretching vibration is expected to shift to a higher frequency i f there is an interaction b e t w e e n the P C L carbonyl group a n d the d e u t e r i u m atom o f the α-deuterated P V C . T h e C ~ D stretching peak does, i n fact, shift to higher frequencies as the P C L concentration is increased. T h u s it can be c o n c l u d e d that the specific interaction i n the P C L - P V C blends is that o f a hydrogen b o n d between the P C L carbonyl group a n d the α-hydrogen o f P V C . I n the foregoing studies o f p o l y m e r blends, the occurrence o f a carbonyl b a n d shift to l o w e r frequencies or an increase i n the w i d t h at half-height o f the b a n d was considered to be evidence o f a specific interaction between the two component polymers. T o support this conclusion, it should be p r o v e d that no such changes i n the carbonyl peak occur for incompatible systems. T h i s , i n fact, was demonstrated i n the study o f the p o l y ^ - p r o p r o l a c t o n e ) ( P P L ) a n d P V C b l e n d system c o n d u c t e d b y C o l e m a n a n d V a r n e l l (26). T h e spectra o f P P L - P V C blends w e r e r e c o r d e d at 80 °C, w h i c h is above the m e l t i n g point o f P P L and, therefore, ensures that b o t h systems are i n the amorphous state. U n l i k e the previously studied compatible blends, the P P L carbonyl b a n d appears w i t h i n experimental error to r e m a i n unchanged i n frequency or shape, despite changes i n the P V C concentration. F u r t h e r evidence can be seen i n the plot o f the w i d t h at half-height o f the carbonyl b a n d as a f u n c t i o n o f P V C concentration. U n l i k e the S - s h a p e d curve f o u n d for the compatible blends, P P L - P V C blends show virtually a straight h o r i z o n tal fine. Similar behavior was f o u n d for the P P L - P V C blends studied i n the solid state at r o o m temperature. T h u s the absence o f either a shifting o f the carbonyl peak to l o w e r frequencies or an increase i n the w i d t h at half-height may b e considered evidence that a p o l y m e r b l e n d system is incompatible a n d that the chains o f one c o m p o n e n t do not recognize the existence o f the second component. A study was c o n d u c t e d b y G a r t o n (27) to determine i f the specific nature o f the intermolecular interaction can be d e t e r m i n e d b y the degree o f frequency shift o f the carbonyl stretching peak. P o l y e s t e r - c h l o r i n a t e d poly m e r blends can experience three different types o f interactions: a h y d r o g e n



204

STRUCTURE-PROPERTY

RELATIONS IN

POLYMERS

b o n d i n g interaction between the carbonyl group o f the polyester a n d either the α-hydrogen o r the β-hydrogen o f the chlorinated p o l y m e r or a d i p o l e - d i pole interaction between the polyester carbonyl group a n d the C ~ C l group o f the chlorinated polymer. A s shown b y V a r n e l l et al. (25), P C L - P V C blends produce a shift i n the carbonyl peak as a result o f a hydrogen-bonding interaction between the P C L carbonyl group a n d the α-hydrogen o f P V C . I n G a r t o n s study, polyester was b l e n d e d w i t h a r a n d o m copolymer o f 80:20 P V C - p o l y ( a c r y l o n i t r i l e ) ( P A N ) . Because this copolymer does not possess any α-hydrogen, it is evident that the type o f intermolecular interaction f o r m e d w o u l d b e between the polyester carbonyl a n d either the β-hydrogen or the C - C l group o f the copolymer. T h e shift o f the carbonyl peak to lower frequencies i n the polyester P V C - P A N b l e n d is nearly identical to the shift that occurs i n the P C L - P V C blends. T h u s the degree o f shifting experienced b y the carbonyl stretching peak o f a compatible b l e n d cannot be used to determine the specific type o f interaction that has b e e n f o r m e d . C o l e m a n et al. (28) next examined the carbonyl stretching region o f the infrared spectrum o f a b l e n d c o m p o s e d o f two erystallizable components to determine the presence or absence of intermolecular interactions; specifi cally, blends o f p o l y ( b i s p h e n o l A carbonate) ( P C ) a n d P C L were studied. This system is u n i q u e because b o t h polymers are erystallizable, although large differences exist i n their crystalline m e l t i n g points (approximately 230 °C a n d 70 °C, respectively) as w e l l as i n their glass transition temperatures (ap proximately 149 a n d — 71 °C, respectively). C h e m i c a l interactions similar to those f o u n d previously i n the P C L - P V C system (25) were identified i n this b l e n d . Specifically, i n the spectra r e c o r d e d at 75 °C (above the crystalline m e l t i n g point o f P C L ) , the P C L amorphous carbonyl b a n d shifts to lower frequencies u p o n addition of P C . T h i s b a n d shift indicates that a specific c h e m i c a l interaction is o c c u r r i n g between the two polymers a n d it involves the carbonyl group of P C L . U t i l i z a t i o n o f the carbonyl region i n an infrared spectrum to determine b l e n d compatibility for a group o f p o l y (vinyl phenol) ( P V P h ) blends was demonstrated b y M o s k a l a et al. (29). T h e carbonyl region for various b l e n d compositions o f P C L - P V P h , P P L - P V P h , a n d p o l y ( v i n y l pyrrolidone) ( P V P r ) a n d P V P h were obtained. T h e P C L - P V P h blends were cast f r o m tetrahydrofuran ( T H F ) a n d r e c o r d e d at 75 °C, w h i c h is above the P C L m e l t i n g point. A s the composition o f P V P h increases, the intensity o f the 1 7 0 8 - c m band increases, w h i l e that o f the 1 7 3 4 - c m b a n d decreases. T h e 1 7 0 8 - c m b a n d results f r o m the h y d r o g e n - b o n d i n g o f P C L carbonyl groups to P V P h phenolic hydroxyl groups, whereas the 1 7 3 4 - c m b a n d is assigned to the self-associ ated carbonyl groups i n the amorphous P C L . T h u s , as the P V P h composition i n the b a n d increases, there is a corresponding increase i n the degree o f interaction between the P V P h a n d P C L . B l e n d s o f P P L - P V P h were cast f r o m T H F a n d recorded at 89 °C, w h i c h is above the m e l t i n g point o f P P L . T h e spectrum o f amorphous P P L i n the carbonyl region is c o m p o s e d o f a - 1

- 1

- 1

- 1


6.


205


broad b a n d located at 1741 c m . U p o n b l e n d i n g w i t h P V P h , the P P L carbonyl groups f o r m hydrogen bonds w i t h the hydroxyl groups of P V P h . This b o n d i n g is demonstrated b y the formation o f a n e w b a n d located at 1722 c m , w h i c h increases i n intensity with the amount o f P V P h i n the b l e n d . Blends o f P V P r - P V P h were cast f r o m T H F a n d r e c o r d e d at r o o m tempera ture. P u r e P V P h contains a b r o a d b a n d located at 1682 c m i n the carbonyl region o f the spectrum. H o w e v e r , this b a n d actually results f r o m a c o m b i n a tion of carbonyl stretching and N - C stretching vibrations. A second b a n d located at 1658 c m appears i n the b l e n d spectrum a n d is attributed to P V P r carbonyl groups that have f o r m e d hydrogen bonds w i t h the hydroxyl groups of P V P h . - 1

- 1

- 1


- 1

Benedetti et al. (30) c o n d u c t e d an experiment to determine the relative strengths of various types o f interactions b y examining the amount o f shift i n the carbonyl peak. F u n c t i o n a l i z e d e t h y l e n e - p r o p y l e n e copolymers ( F E P ) were p r e p a r e d b y reacting the m o l t e n p o l y m e r w i t h diethylamylate ( D E M ) i n the presence o f dicumylperoxide ( D C P ) . Solutions o f F E P were dissolved i n η-heptane ( n = C H ) , tetrahydrofuran ( T H F ) , carbon tetrachloride ( C C l ) , 1,1,1 -trichloroethane ( C C l - C H ) , a n d c h l o r o f o r m ( C H C l ) , a n d the result i n g carbonyl stretching regions o f their spectra were recorded. T h e strength o f the interaction between the carbonyl group i n the D E M unit o f F E P a n d the various solvents was d e t e r m i n e d f r o m the amount of shift o f the carbonyl peak to lower frequencies. T h e solvents were then r a n k e d i n the following order according to decreasing strength: n - C H > THF > CCl -CH > C C l > C H C 1 . These results indicate that the strongest interaction occurs between the D E M carbonyl group a n d a methine hydrogen. 7

1 6

4

3

3

3

7

4

1 6

3

3

3

F u r t h e r studies o n the shift o f the carbonyl peak experienced b y compat ible blends were c o n d u c t e d b y G a r t o n (31). P C L or its l o w molecular weight m o d e l c o m p o u n d m e t h y l acetate was dissolved i n a mixture o f two solvents that duplicate the possible interacting centers i n chlorinated polymers. These two solvents, A a n d B , w i l l b o t h interact w i t h the P C L carbonyl group. It should then be possible to resolve the carbonyl stretching b a n d into the two components that result f r o m the two types o f interactions a n d then calculate the area of each component. A n e q u i l i b r i u m constant, K, may t h e n be calculated according to the following equation: [C=Q-A][B] [C=0-B][A] T h u s , a comparison o f the strength of several possible interacting centers may be established f r o m the e q u i l i b r i u m constants. A comparison o f the interaction strength between the m o d e l c o m p o u n d M A a n d several α-hydrogenated chlorocarbons a n d heptane is shown i n F i g u r e 6, w h e r e the slope o f the Une is equal to the e q u i l i b r i u m constant



206

STRUCTURE-PROPERTY


W/IB] Figure 6. Association behavior of methyl acetate in mixed solvents. Solvent A = 1122TCE ( • ), dichloromethane ( Δ ) chloroform (O), carbon tetrabromide (%). Solvent Β = heptane. The dashed line corresponds to no preferential association of methyl acetate (i.e., Κ = l). (Reproduced with permission from reference 31. Copyright 1983 Society of Plastics Engineers, Inc.) x

defined previously. A preference for association o f the M A to the a - h y d r o genated chlorocarbons is clearly established. F i g u r e 7 compares the relative interaction strengths o f an α-hydrogenated chlorocarbon (1,1,2,2-tetrachloroethane) ( 1 1 2 2 T C E ) , a β-hydrogenated chlorocarbon (1,1,1-trichloroethane) ( 1 1 1 T E C ) , a n d carbon tetrachloride ( C C l ) . These three solvents were chosen because they represent the possible interacting sites i n p o l y (vinyl chloride) a n d p o l y ( v i n y l i d e n e chloride), b o t h o f w h i c h f o r m compatible p o l y m e r blends w i t h polyester. T h e strength o f interacting abilities i n d e creasing order was established to b e α-hydrogenated chlorocarbon > βhydrogenated chlorocarbon > carbon tetrachloride. F u r t h e r m o r e , because the e q u i l i b r i u m constants f o r the 1 1 2 2 T C E heptane system a n d the 1 1 2 2 T C E - C C 1 system were f o u n d to b e nearly identical, the interaction 4

4



6.


W/[B]


207

Figure 7. Association behavior of methyl acetate in mixed solvents. Solvent A = 1122TCE. Solvent Β = carbon tetrabromide (A), 1,1,1 -trichloroethane (O). The dashed line corresponds to no preferen tial association of methyl acetate (i.e., Κ =l). (Reproduced with permission from reference 31. Copyright 1983 Soci ety of Plastics Engineers, Inc.)

strength o f b o t h heptane a n d carbon tetrachloride to the ester are essentially equal. Similar studies b y G a r t o n (31) revealed that the strength o f the interac tion o f ester w i t h carbon tetrabromide is m u c h greater than w i t h heptane. T h i s observation is attributed to the h i g h l y polarizable C - B r b o n d . It was also r e p o r t e d that i n a system o f α-hydrogenated chlorocarbon ( 1 1 2 2 T C E ) a n d a cyclic ether ( T H F ) , the e q u i h b r i u m only shghtly favors the 1 1 2 2 T C E ; thus, this establishes that α-hydrogenated chlorocarbons f o r m only shghtly stronger interactions than cyclic ethers. F r o m the m o d e l c o m p o u n d a n d solution studies, G a r t o n (31) established that these solvents interact w i t h a m o d e l ester according to the following strengths: α-hydrogenated chlorocarbons > T H F > β-hydrogenated chloro carbons — carbon tetrabromide > carbon tetrachloride — heptane. T h e preference to interact w i t h α-hydrogenated chlorocarbons was then d u p l i c a t e d b y G a r t o n (31) i n studies using P C L , rather than its m o d e l c o m p o u n d . H o w e v e r , w h e n the P C L was dissolved i n the 1 1 2 2 T C E - C C l system, the e q u i h b r i u m constant f o r M A was greater than that f o r P C L . T h i s result was attributed to stiffness, steric limitations, o r the conformation o f the P C L p o l y m e r chain, a l l o f w h i c h i n h i b i t the ester groups f r o m h y d r o g e n b o n d i n g w i t h the chlorocarbon. 4

C o l e m a n a n d M o s k a l a (32) p e r f o r m e d studies o n the p o l y m e r b l e n d system o f p o l y (hydroxy ether o f b i s p h e n o l A ) (phenoxy) a n d P C L , p l a c i n g emphasis o n the dependence o f intermolecular interactions o n b l e n d c o m p o sition. Results w e r e obtained first at 75 °C, w h i c h is above the m e l t i n g point


208

STRUCTURE-PROPERTY

RELATIONS IN

POLYMERS

of P C L . I n the carbonyl stretching region, the intensity of a shoulder at approximately 1720 c m " increased w i t h increasing concentration of phenoxy a n d was attributed to an intermolecular interaction between the carbonyl group o f P C L a n d the hydroxyl group of the phenoxy. C u r v e resolving studies were p e r f o r m e d o n the carbonyl region of the blends to reveal two c o m p o nents: a relatively n a r r o w b a n d centered at 1734 c m a n d a relatively b r o a d b a n d centered at 1720 c m . These two components corresponded to isolated P C L carbonyl groups a n d h y d r o g e n - b o n d e d carbonyl groups, respec tively. T h e b l e n d composition was d e t e r m i n e d f r o m the relative fraction o f h y d r o g e n - b o n d e d carbonyl groups. T h i s estimate was obtained b y taking the ratio of the area u n d e r the 1 7 2 0 - c m peak d i v i d e d b y the s u m of the areas u n d e r the 1720- a n d 1 7 3 4 - c m peaks. A plot of the relative fraction o f h y d r o g e n - b o n d e d carbonyl groups as a function of mole percent phenoxy reveals that the fraction of h y d r o g e n - b o n d e d carbonyls increases linearly w i t h increasing phenoxy concentration. 1

- 1

- 1

- 1


- 1

T h e hydroxyl stretching region of the p h e n o x y - P C L blends was also examined to determine the relative strength o f intermolecular interactions. A s seen i n F i g u r e 8, the p u r e phenoxy i n the spectrum consists o f two c o m p o nents: a relatively narrow peak centered at 3570 c m a n d a b r o a d peak at 3450 c m . These two peaks correspond to the free hydroxyl groups a n d the hydrogen-bonded hydroxyl groups, respectively. T h e peak due to the hydro gen-bonded hydroxyl shifts to higher frequencies u p o n b l e n d i n g of the phenoxy w i t h P C L . H o w e v e r , the peak assigned to the free hydroxyls remains unchanged u p o n addition o f the P C L . P u r c e l l d e t e r m i n e d that the difference between the frequencies o f the peak due to the free hydroxyls a n d that o f the peak due to the h y d r o g e n - b o n d e d hydroxyls is a measure of the average strength o f the intermolecular hydrogen b o n d i n g . Based o n this conclusion, the changes i n the hydroxyl stretching region o f the p h e n o x y - P C L b l e n d indicate that the hydrogen b o n d i n g between the P C L carbonyl group a n d the phenoxy hydroxyl group is weaker than the hydrogen-bonding interaction i n pure phenoxy. - 1

- 1

A study was also c o n d u c t e d b y C o l e m a n a n d M o s k a l a ( 3 2 ) o n p h e n o x y - p o l y ( e t h y l e n e oxide), ( P E O ) blends to determine the relative strength o f intermolecular interactions; the same technique as e m p l o y e d for the p h e n o x y - P C L system (32) was used. I n contrast to the p h e n o x y - P C L system, the h y d r o g e n - b o n d e d hydroxyl peak of p h e n o x y - P E O blends shifts to lower frequencies w i t h the addition o f P E O . T h u s it was c o n c l u d e d that the hydrogen-bonding interaction between the P E O carbonyl group a n d the phenoxy hydroxyl group is stronger than the intermolecular hydrogen b o n d i n g i n p u r e phenoxy. M o s k a l a a n d C o l e m a n (33) expanded their study o f phenoxy blends b y examining blends of phenoxy w i t h poly (vinyl alkyl ethers); specifically, the compatible b l e n d p h e n o x y - p o l y (vinyl m e t h y l ether) ( P V M E ) a n d the i n c o m patible blends p h e n o x y - p o l y (vinyl ethyl ether) ( P V E E ) a n d p h e n o x y -


6.



209


3505

3700

cm'

1

3100

Figure 8. FTIR spectra recorded at 75 °C of PCL-phenoxy blends containing 0 (Aj, 10 (B), 20 (C), 30 (D), 40 (E), and 50 (F) weight percent PCL. (Reproduced with permission from reference 32. Copyright 1983 Butterworth.)

p o l y (vinyl isobutyl ether) ( P V I E ) . A s was seen previously for the p h e n o x y P C L blends (32), the phenoxy h y d r o g e n - b o n d e d hydroxyl stretching peak i n p h e n o x y - P V M Ε blends shifts to higher frequencies u p o n addition o f P V M E . H o w e v e r , this shift to higher frequencies is accompanied b y a corresponding decrease i n relative broadness o f the peak. T h e w i d t h at half-height for this peak decreases f r o m 260 c m for p u r e phenoxy to 150 c m for a 20:80 w t % p h e n o x y - P V M E b l e n d . B y contrast, the spectra o f the incompatible p h e n o x y - P V E E a n d the p h e n o x y - P V I E blends show neither a shift i n frequency n o r a n a r r o w i n g o f the phenoxy h y d r o g e n - b o n d e d hydroxyl peak u p o n addition of the respective poly (vinyl ether). This n a r r o w i n g o f the self-associated hydroxyl peak i n the p h e n o x y - P V M E b l e n d was explained i n the following way. T h e b u l k y benzene ring o f the phenoxy creates steric - 1

- 1


210

STRUCTURE-PROPERTY


problems a n d prevents a consistency o f hydrogen b o n d distances a n d geome tries. H o w e v e r , the hydroxyl stretching m o d e narrows u p o n b l e n d i n g w i t h P V M E because the hydroxyl group can f o r m a more homogeneous distribu tion o f hydrogen b o n d lengths a n d geometries w i t h the relatively flexible P V M E molecule. M o s k a l a et a l . (34) demonstrated the use o f the hydroxyl stretching region to determine b l e n d compatibility f o r a group o f P V P h blends; specifi cally, P C L - P V P h , P V P r - P V P h , P E O - P V P h , a n d blends o f P V P h w i t h two different poly (vinyl alkyl ethers). T h e hydroxyl stretching region o f the P C L - P V P h blends c a n be resolved into three peaks. T h e non-hydrogenb o n d e d hydroxyls o f P V P h are responsible f o r a peak at 3525 c m ~ . A second peak at 3370 c m results f r o m the self-association o f P V P h hydroxyl groups a n d a peak at 3420 c m is assigned to P V P h hydroxyl groups hydrogenb o n d e d to P C L carbonyl groups. A s the P C L concentration i n the b l e n d is increased, there is a corresponding decrease i n intensity i n the first t w o peaks a n d a n increase i n the 3 4 2 0 - c m peak. M e a s u r i n g the intermolecular interaction strength as the difference i n frequency between the peak result i n g f r o m the interaction a n d that o f the free hydroxyls leads to the conclusion that the strength o f the self-associated hydroxyl groups i n P V P h (àv = 165 c m " ) is stronger than the hydrogen b o n d f o r m e d b e t w e e n the P V P h hydroxyl groups a n d the P C L carbonyl group ( Δ ν = 105 c m ) . S i m i l a r results w e r e obtained for P P L - P V P h blends, w i t h the peak due to the hydroxyl groups h y d r o g e n - b o n d e d to carbonyl groups shifted slightly to 3440 c m . U n l i k e the P C L - P V P h a n d P P L - P V P h blends, the self-associated hydroxyl peak occurs at a higher frequency (3360 c m ) than the hydrogenb o n d e d hydroxyl peak (3230 c m ) i n the P V P r - P V P h blends. It is therefore c o n c l u d e d that the intermolecular interaction between the P V P h hydroxyl group a n d the P V P r carbonyl group is stronger than the h y d r o x y l - h y d r o x y l interaction. F o r the P E O - P V P h blends, a peak at 3200 c m results f r o m the P V P h hydroxyl group h y d r o g e n - b o n d e d to the ether oxygen o f P E O . T h e frequency difference between this peak a n d that due to the free hydroxyl groups is 325 c m . T h i s frequency difference is greater than the difference f o u n d i n the previously studied (32) p h e n o x y - P E O blends (270 c m ) a n d reflects the fact that the P V P h hydroxyl groups have a greater affinity to hydrogen b o n d than do the phenoxy hydroxyls. F u r t h e r evidence o f the greater affinity o f the P V P h hydroxyl groups c a n be seen b y c o m p a r i n g the blends o f P V M E - P V P h a n d p h e n o x y - P V M E . T h e difference i n frequency between the peak assigned to the hydroxyl groups h y d r o g e n - b o n d e d to ether oxygens a n d that due to the free hydroxyl groups is 205 c m f o r the P V M E - P V P h b l e n d a n d 150 c m for the p h e n o x y - P V M E b l e n d . C o u s i n a n d P r u d ' h o m m e (22) c o n d u c t e d a miseibility study o f several polyesters w i t h p o l y (vinyl halides) a n d c o n c l u d e d that intermolecular interac tions occur o n a mole-to-mole basis for compatible P E - P V X blends. T h e degree o f frequency shift i n the carbonyl stretching b a n d is plotted as a 1


-

1

-

1

- 1

1

- 1

- 1

- 1

- 1

-

1

- 1

- 1

-

-

1


1


6.


211


function o f weight percent p o l y (vinyl halide) i n the b l e n d , as shown i n F i g u r e 9. W i t h i n experimental error, no shifting o f the carbonyl peak occurs for the P C L - p o l y ( v i n y l fluoride) ( P V F ) blends. T h u s it can be c o n c l u d e d that the P C L carbonyl peak does not interact w i t h the P V F methine hydrogen a n d the P C L - P V F b l e n d is, therefore, incompatible. I n contrast, a shifting o f the carbonyl peak i n the P C L - P V C a n d P C L - p o l y (vinyl b r o m i d e ) ( P V B ) blends indicates that these two mixtures are compatible. It can also be seen that the carbonyl b a n d frequency shift differs shghtly w h e n the P C L - P V C a n d P C L - P V B blends are c o m p a r e d . H o w e v e r , as seen i n F i g u r e 10, i f this frequency shift is p l o t t e d as a f u n c t i o n o f m o l a r composition o f poly (vinyl halide) rather than weight percent, the degree o f shifting experienced b y the two blends is nearly identical. T h i s observation leads to the conclusion that the interactions between the P C L a n d P V X occur o n a mole-to-mole basis. Similar results were obtained w h e n P V C a n d P V B were b l e n d e d w i t h other polyesters, namely, polyQiexamethylene sebacate) ( P H M S ) a n d poly(valerolactone) ( P V L ) .

βι-

Ό

Κ)

20

30

40

50

60

70

60

90

TOO

PVX Figure 9. Frequency shift of the carbonyl group of PCL as a function of the PVX weight percent in the mixture. Measurements made at 80 °C. Key: • , PVF; Φ, PVB; and A , PVC. (Reproduced with permission from reference 22. Copyright 1983 Butterworth.)



212

STRUCTURE-PROPERTY

40 60 PVX (mo!%)


POLYMERS

100

Figure 10. Frequency shift of the carbonyl group of PCL as a function of the PVX mole percent in the mixture. Measurements made at 80 °C. Key: • , PVC; • , PVB. (Reproduced with permission from reference 22. Copyright 1983 Butterworth.)

F o r all six compatible blends studies b y C o u s i n and P r u d ' h o m m e (22), the amount o f shift o f the carbonyl b a n d increased w i t h increasing P V X concentration. T h i s relationship occurs because at l o w P V X concentrations, only a small amount o f P V X molecules are i n t e r m i x e d w i t h i n the polyester matrix. Therefore, the n u m b e r o f interactions f o r m e d between the P E carbonyls a n d the P V X methine hydrogens is small. H o w e v e r , at h i g h P V X concentrations a small amount o f P E molecules are i n t e r m i x e d w i t h the P V X matrix, w h i c h allows a larger percentage o f the available carbonyl groups to f o r m hydrogen bonds w i t h the P V X a n d results i n a larger shift o f the carbonyl stretching peak. Blends o f p o l y (butylène adipate) ( P B A ) w i t h P V C a n d P V B were studied i n a similar manner. A l t h o u g h a decrease i n the carbonyl frequency o c c u r r e d w i t h increasing P V C content for P B A - P V C blends, the amount of shift was small, especially at l o w P V C concentrations. F o r the P B A - P V B blends, no frequency shift was detected for blends w i t h less than 6 0 - m o l % P V B . T h u s it was c o n c l u d e d that there was only partial miseibility i n the P B A - P V B blends. T o establish a relationship between the degree of miseibility a n d the molecular structure o f the various p o l y (vinyl halides), a comparison o f the amount o f carbonyl shifting at h i g h P V X concentrations w i t h i n the respective blends was made. T h e carbonyl shifts o f the various polyesters decreased i n the following order: P H M S < P C L < P V L < P B A . T h i s ranking corre-


6.



213


sponds to a decrease i n the order of carbonyl concentration o n the repeat unit o f the polyester a n d an increase i n the order o f rigidity o f the polyester chain. B o t h trends result i n a decrease i n the n u m b e r o f intermolecular interactions that may occur between the carbonyl groups a n d the P V X hydrogens. C o l e m a n et al. (35) further expanded the study o f the shift i n the carbonyl b a n d b y examining the dependence o f the shift o n temperature. Studies of P V C - e t h y l e n e v i n y l acetate ( E V A ) c o p o l y m e r blends a n d P V C chlorinated polyethylene ( C P E ) blends were undertaken to determine w h e t h e r a correlation exists between the strength o f the intermolecular interaction as d e t e r m i n e d b y the frequency shift o f the carbonyl b a n d a n d the onset o f phase separation at the lower critical solution temperature ( L C S T ) . T h e i n f r a r e d spectrum o f the carbonyl b a n d for C P E - E V A a n d P V C - E V A blends were r e c o r d e d at r o o m temperature. A s expected, the occurrence o f specific intermolecular interactions is revealed t h r o u g h a shift to lower frequencies a n d an increase i n the w i d t h at half-height. W a l s h et al. ( 3 6 ) previously r e p o r t e d that the L C S T for C P E - E V A blends is b e l o w 130 °C. T h i s value was c o n f i r m e d b y C o l e m a n et al. b y examination o f the carbonyl peak o f the i n f r a r e d spectra r e c o r d e d after heating the blends for approxi mately 3 h at 130 °C. Incompatibility at this temperature was demonstrated b y the fact that the spectra o f the p u r e E V A a n d blends containing 40- a n d 8 0 - w t % C P E are nearly identical w i t h i n experimental error. Attempts were then made to determine the L C S T f r o m changes i n the frequency o f the carbonyl b a n d as a f u n c t i o n o f temperature. Studies c o n d u c t e d to determine frequency changes i n the p u r e E V A carbonyl b a n d w i t h increasing temperature showed that the frequency increased shghtly i n a linear relationship w i t h increasing temperature. Spectra o f an 8 0 : 2 0 - w t % C P E - E V A b l e n d obtained i n a r o o m temperature to 160 °C range showed that the frequency o f the carbonyl peak increased w i t h temperature. F i g u r e 11 shows a plot o f the frequency versus temperature for b o t h the p u r e E V A a n d the 8 0 : 2 0 - w t % C P E - E V A b l e n d . T h e relative strength o f the interaction at any temperature is d e t e r m i n e d b y the difference between the carbonyl frequency i n the b l e n d a n d i n the p u r e E V A . T h i s difference becomes smaller as the temperature is increased, w h i c h indicates a decrease i n the strength o f the interaction between the two components o f the b l e n d . A t temperatures ranging f r o m 35 to 90 °C, the strength o f the intermolecular interaction as d e t e r m i n e d b y the difference i n frequencies is great enough to result i n a compatible b l e n d . H o w e v e r , above approximately 110 °C, the interaction has decreased to such a degree that phase separation occurs. T h u s the p r e d i c t e d L C S T of C P E - E V A blends occurs b e t w e e n 90 a n d 110 °C, w h i c h corresponds to a critical value of the strength o f the interaction. Similar behavior was observed for 80:2()-wt% P V C - E V A blends, a n d the L C S T o f this system was p r e d i c t e d to be between 110 a n d 130 °C. These results suggest that changes i n the intermolecular interactions between b l e n d


214

STRUCTURE-PROPERTY


140 Ο

i

100

LCST-

φ Ω. Ε Φ


60

1733

1735 1737 Wavenumber (cm- )

1739

1

Figure 11. Plot of the temperature versus the carbonyl peak position for an 80:20-wt% CPE-EVA blend. (Reproduced with permission from reference 35. Copyright 1983 Butterworth.) components analysis.

as a function o f temperature

can be determined b y F T I R

T h e effect o f temperature o n hydrogen b o n d i n g i n compatible blends was studied b y T i n g et al. ( 3 7 ) . A poly(styrene-co-vinylphenyl trifluoromethyl carbinol) ( P F A ) a n d P E O b l e n d a n d a poly(styrene-co-vinylphenol hexafluor o d i m e t h y l carbinol) ( P H F A ) a n d P E O b l e n d i n a 4:1 ratio were studied at temperatures o f 2 5 , 75, 125, a n d 175 ° C . A t 75 °C, t h e h y d r o g e n - b o n d e d hydroxyl peak o f the P F A - P E O b l e n d decreased i n intensity, while that o f the P H F A - P E O b l e n d d i d not change. T h i s observation was attributed to the fact that t h e strong acidity o f t h e P H F A hydroxyl group resulted i n t h e formation o f a stronger hydrogen b o n d . A t 175 °C, hydrogen-bond dissocia t i o n i n b o t h blends is evidenced b y the significant decrease i n the intensity o f the h y d r o g e n - b o n d e d hydroxyl group a n d the formation o f peaks attributed to free hydroxyl groups at 3550 c m " i n the P F A b l e n d a n d at 3600 a n d 3520 cm i n the P H F A b l e n d . T h e s e samples were t h e n slowly cooled to r o o m temperature, d u r i n g w h i c h t i m e t h e hydrogen bonds were reformed, as indicated b y the reappearance o f the hydrogen-bonded hydroxyl peaks. It was c o n c l u d e d that t h e phase separation o f P F A - P E O a n d P H F A - P E O blends that o c c u r r e d u p o n heating to 175 °C was reversible. A study was undertaken b y Skrovansk a n d C o l e m a n (38) to determine the ability o f a strongly self-associated p o l y m e r to interact w i t h another 1

- 1


6.


215



p o l y m e r at the molecular level. Before intermolecular interactions between the t w o component polymers c a n b e f o r m e d , the interactions present i n the pure polymers must be b r o k e n . Therefore, a large negative enthalpy o f m i x i n g w i l l only occur i f the strength o f the association between the t w o polymers exceeds the average o f the strength o f self-association o f each o f the component polymers. Therefore, a strongly self-associated p o l y m e r s h o u l d react w i t h another p o l y m e r that is weakly self-associated b u t contains a c h e m i c a l moiety that has the potential to f o r m a relatively strong intermolecu lar interaction. F o r this study, Skrovansk a n d C o l e m a n (38) chose polyamide as the strongly self-associated p o l y m e r because o f extensive intermolecular h y d r o gen b o n d i n g o f the amide group. T h e polyamide was b l e n d e d w i t h poly(vinyl-2-vinyl p y r i d i n e ) ( P 2 V P ) , w h i c h is a weakly self-associated p o l y m e r that has intermolecular forces o f a dispersive nature. I n addition, P 2 V P has a nitrogen atom that contains a lone pair o f electrons that c a n act as a n excellent site f o r hydrogen b o n d i n g to a labile p r o t o n . E v i d e n c e that intimate m i x i n g has o c c u r r e d i n the b l e n d o f polyamide a n d P 2 V P is seen i n the N - H stretching region o f the i n f r a r e d spectrum o f the b l e n d . T h e spectrum o f amorphous polyamide at r o o m temperature consists o f t w o peaks; the most p r o m i n e n t peak is centered at 3310 c m . T h e extreme broadness o f this b a n d is caused b y the w i d e distribution i n strengths o f the h y d r o g e n - b o n d e d N - H groups. T h e second peak is located at 3444 c m a n d has b e e n assigned to the free N - H groups. T h e hydrogenb o n d e d N - H peak appears to shift to lower frequencies w h e n the polyamide is b l e n d e d w i t h P 2 V P , b u t actually the overlapping o f two major components is b e i n g observed. T h e first component is attributed to the self-association o f polyamide (between the N - H group a n d the carbonyl group) a n d the second component results f r o m the association o f the polyamide N - H group to the P 2 V P nitrogen atom. T h e apparent relative shift i n frequency after b l e n d i n g leads to the conclusion that the association o f polyamide to P 2 V P is stronger than the self-association o f polyamide. - 1

-

1

F u r t h e r evidence o f a n interaction between the polyamide a n d P 2 V P c a n be seen i n the amide I a n d amide II bands o f the infrared spectra obtained at r o o m temperature after removal b y subtraction o f the spectral contributions f r o m P 2 V P . T h e amide I b a n d , centered at 1640 c m , results f r o m carbonyl stretching a n d is c o m p o s e d o f t w o major contributions: one at 1640 c m f r o m the carbonyl groups that are h y d r o g e n - b o n d e d to N - H groups a n d the other f r o m free carbonyl groups that appear at 1670 c m . T h e peak at 1670 cm becomes more p r o n o u n c e d after b l e n d i n g , w h i c h indicates an increase i n the fraction o f free carbonyl groups. A s expected, the contribution o f the free carbonyl b a n d was calculated b y a curve fitting technique to b e approxi mately twice that f o u n d i n the pure amorphous polyamide. This increase i n the n u m b e r o f free carbonyl groups was interpreted as a result o f a specific interaction between the two components o f the b l e n d . I n p u r e amorphous - 1

-

- 1

- 1


1

216

STRUCTURE-PROPERTY


POLYMERS

polyamide at r o o m temperature, nearly all the amide groups are hydrogenb o n d e d a n d there are an equal n u m b e r o f h y d r o g e n - b o n d e d N - H a n d C = 0 groups. W h e n polyamide is b l e n d e d w i t h P 2 V P , the n u m b e r o f free carbonyl groups increases because an interaction between a polyamide N - H group a n d a P 2 V P nitrogen atom must be p r e c e d e d b y a corresponding break o f a hydrogen b o n d between a polyamide N - H group a n d a polyamide carbonyl group. T h e amide II b a n d located at 1542 c m results f r o m N - H in-plane b e n d i n g . T h i s b a n d also indicates the presence o f an intermolecular interac tion i n the b l e n d because it shifts to a higher frequency a n d broadens. T h i s frequency shift a n d broadening is interpreted to be the result o f an increase i n the strength o f the interaction involving the N - H groups. Downloaded by UNIV LAVAL on July 11, 2016 | http://pubs.acs.org Publication Date: May 5, 1993 | doi: 10.1021/ba-1993-0236.ch006

1

C o m p a t i b i l i t y studies o f P S - P V M E were c o n d u c t e d b y L u et al. (39). A 50:50 molar composition o f h i g h molecular weight P S - P V M E produces a compatible b l e n d w h e n cast f r o m toluene. H o w e v e r , w h e n the same p o l y m e r system is cast f r o m c h l o r o f o r m or trichloroethylene ( T C E ) , an incompatible b l e n d results. Spectroscopic differences between the compatible a n d i n c o m patible blends can be f o u n d i n the 1100- a n d 7 0 0 - c m " regions. T h e P S spectrum was subtracted f r o m the 1 1 0 0 - c m region o f b o t h types o f blends. T h e doublet present i n the resultant difference spectrum is c o m p o s e d o f two peaks located at 1107 a n d 1085 c m . T h e relative intensities o f these two peaks is d e t e r m i n e d b y whether the b l e n d is compatible or incompatible. T h e 1 0 8 5 - e m b a n d has greater intensity i n the compatible blends whereas the 1 1 0 7 - c m " b a n d dominates i n the incompatible blends. Snyder a n d Z e r b i (40) c o n c l u d e d that the difference between the intensities o f the two peaks i n this doublet results f r o m changes i n the C - O - C asymmetric stretching i n the C O C H group o f P V M E . T h e spectrum o f P S shows the most significant changes w i t h b l e n d i n g i n the 7 0 0 - c m region. T h e C - H out-of-plane b e n d i n g vibration o f the p h e n y l r i n g is located at 697.7 c m for pure P S , at 699.5 c m for compatible P S - P V M E blends o f equal molar composition, and at a frequency between these two values for incompatible 50:50 P S - P V M E blends. Changes i n these two spectral regions indicate that the interaction that occurs i n these blends is between the p h e n y l r i n g o f P S a n d the C O C H 3 group o f P V M E . 1

- 1

- 1

_ 1

1

3

- 1

- 1

- 1

T h e dependence o f compatibility o n b l e n d composition was also d e m o n strated i n this experiment. A s previously stated, 50:50 P S - P V M E blends were compatible w h e n cast f r o m toluene but w e r e incompatible w h e n cast f r o m c h l o r o f o r m or T C E . H o w e v e r , blends o f 15:85 P S - P V M E were compatible w h e n cast f r o m toluene, c h l o r o f o r m , or T C E a n d showed no differences i n the 1100- or 7 0 0 - c m " regions. Sargent a n d K o e n i g (41) studied the compatibility o f p o l y ( v i n y l i d e n e fluoride) ( P V F ) a n d poly (vinyl acetate) ( P V A c ) blends as a function o f t h e r m a l treatment a n d b l e n d composition. F i l m s o f the p o l y m e r blends w i t h weight ratios ranging f r o m 10:90 to 90:10 P V F ~ P V A c were cast f r o m 1

2

2


6.

217



solution onto K B r plates a n d p l a c e d i n an oven u n d e r v a c u u m for 1 h . T w o sets o f samples were prepared. T h e samples d i f f e r e d i n the temperature at w h i c h the solvent was evaporated. T h e first set o f samples was p l a c e d i n a 75 °C oven a n d the second set was p l a c e d i n a 175 °C oven. F a c t o r analysis was used to determine whether blends o f the various compositions o f the two homopolymers were compatible. Results indicated that the P V F - P V A c blends heat treated at b o t h 75 a n d 175 °C were compatible. T h e spectroscopic changes that resulted f r o m the intermolecular interactions between the two homopolymers were isolated b y subtracting the spectra o f the p u r e P V F a n d p u r e P V A c f r o m the spectra o f the blends. T h e relative amounts o f each o f these components present i n the blends were t h e n d e t e r m i n e d b y least squares curve fitting a n d the percent contributions of each o f the interaction spectra were analyzed as a function o f the b l e n d composition a n d t h e r m a l treatment of the samples. A s shown i n F i g u r e 12, for b o t h the samples heat treated at 75 °C and 175 °C a general increase occurs i n the degree o f interaction w i t h a corresponding increase i n P V F concentration. T h e degree o f interaction 2


2

2

35

1

—

—

x ο ο

ο

ο

ο

ο

ο 40

20

60 •t.Χ

60

PVF„

Figure 12. Plot of the percent contribution of the interaction spectrum versus weight percent PVF for PVF -PVAc blends heat treated at 75 °C (O) and 175°C(X)(41). 2

2


100

218

STRUCTURE-PROPERTY


w i t h i n the blends was greater i n the samples heat treated at 175 °C than i n those heat treated at 75 ° C . T h i s difference was attributed to the fact that those samples heat treated at 175 °C were subjected to a t h e r m a l treatment that was at o r above the m e l t i n g point o f the blends, w h i c h results i n a n increase i n the m o b i l i t y o f the h o m o p o l y m e r molecules a n d allows m o r e intimate mixing between the components. A n i m p r o v e m e n t i n the degree o f m i x i n g i n a p o l y m e r b l e n d system thus promotes the formation o f intermolec ular interaction between specific c h e m i c a l groups o f the component poly mers, as reflected b y the higher interaction contribution i n the blends heat treated at 175 °C. T h e percent contribution o f the interaction spectrum was also f o l l o w e d as a function o f t i m e f o r the P V F ~ P V A c blends. E x a m i n a t i o n o f the data revealed that w i t h i n experimental error the interaction contribution d i d not decrease over a p e r i o d o f 24 days. T h i s observation indicates that n o detectable phase separation o c c u r r e d w i t h i n the blends d u r i n g this t i m e p e r i o d o f analysis.


2

Summary and Conclusions C o m p a t i b l e p o l y m e r blends c a n b e characterized b y the formation o f inter molecular interactions between specific c h e m i c a l groups o f the component polymers. These interactions c a n b e studied o n the molecular level using i n f r a r e d spectroscopy. Spectral data processing techniques, such as factor analysis, difference spectroscopy, a n d least squares curve fitting, have b e e n a p p l i e d to characterize these interactions. E x a m i n a t i o n o f frequency shifts, b a n d broadening, a n d changes i n peak intensity w i t h i n the b l e n d spectrum can also b e used to identify the interactions. Additionally, the dependence o f compatibility o n such factors as b l e n d composition a n d temperature c a n be d e t e r m i n e d f r o m infrared spectroscopy studies.

References 1. Flory, P. J. Principles of Polymer Chemistry; Cornell University Press: Ithaca, NY, 1953. 2. Tompa, H . Polymer Solutions; Butterworths: London, England, 1956. 3. Bohn, L. Z. Kolloid, Polym. 1966, 213, 55. 4. Rosen, S. L. Poly Eng. Sci. 1967, 7, 115. 5. Krause, S.; J. Macromol. Sci., Chem. 1972, C7, 251. 6. Bralow, J. W.; Paul, D. R. Polym. Eng. Sci. 1981, 21, 985. 7. Olabisi, O.; Robeson, L. M . ; Shaw, M . T. Polymer-Polymer Miscibility; Aca demic: Orlando, F L , 1979. 8. Polymer Blends; Paul, D. R.; Newman, S., Eds.; Academic: Orlando, F L , 1978; Vols. I and II. 9. Coleman, M . M.; Painter, P. C. J. Macromol. Sci., Chem. 1977, C16, 197. 10. Coleman, M . M . ; Painter, P. C. Appl. Spectrosc. Rev. 1984, 20, 255. 11. Gillette, P. C.; Lando, J. B.; Koenig, J. L . Fourier Transform Infrared Spectroscopy; Academic: Orlando, F L , 1985; Vol. 4, Chapter 1.



6.

SARGENT AND KOENIG


219

12. Flory, P. J. Principles of Polymer Chemistry; Cornell University Press: Ithaca, NY, 1962; Chapter XII. 13. Scott, R. L. J. Chem. Phys. 1949, 17, 279. 14. Howe, S. E.; Coleman, M . M . Macromolecules 1986, 19, 72. 15. Antoon, M . K.; D'Esposito, L.; Koenig, J. L. Appl. Spectrosc. 1979, 33, 351. 16. Malinowski, E. R. Anal. Chem. 1977, 49, 612. 17. Koenig, J. L. Appl. Spectrosc. 1975, 29, 293. 18. Blackburn, J. A. Anal. Chem. 1965, 37, 1000. 19. Koenig, J. L.; Tovar Rodriquez, M . J. M . Appl. Spectrosc. 1981, 35, 543. 20. Wellinghoff, S. T.; Koenig, J. L.; Baer, E. J. Polym. Sci., Polym. Phys. Ed. 1977, 15, 1913. 21. Leonard, C.; Halary, J. L.; Monnerie, L. Polymer 1985, 26, 1507. 22. Cousin, P.; Prud'homme, R. E. Multicomponent Polymer Materials; Paul, D. R.; Sperling, L. H., Eds.; Advances in Chemistry 211; American Chemical Society: Washington, DC, 1986; pp 87-110. 23. Coleman, M . M.; Zarian, J. J. Polym. Sci., Polym. Phys. Ed. 1979, 17, 837. 24. Kirkpatrick, H . M.S. Thesis, University of Cincinnati, Cincinnati, O H , 1975. 25. Varnell, D. F.; Moskala, E. J.; Painter, P. C.; Coleman, M . M . Polym. Eng. Sci. 1983, 23, 658. 26. Coleman, M . M.; Varnell, D. F. J. Polym. Sci., Polym. Phys. Ed. 1980, 18, 1403. 27. Garton, A. J. Polym. Sci., Polym. Lett. Ed. 1983, 21, 45. 28. Coleman, M . M . ; Varnell, D. F.; Runt, J. P. Contemporary Topics in Polymer Science; Bailey, W. J.; Tsuruta, T., Eds.; Plenum: New York, 1984; Vol. 4, p 807. 29. Moskala, E. J.; Varnell, D. F.; Coleman, M . M . Polymer 1985, 26, 228. 30. Benedetti, E.; Posar, F.; D'Alessio, Α.; Vergamini, P.; Aglietto, M . M . ; Ruggeri, G.; Ciardelli, F. Br. Polym. J. 1985, 17, 34. 31. Garton, A. Polym. Eng. Sci. 1983, 23, 663. 32. Coleman, M . M.; Moskala, E. J. Polymer 1983, 24, 251. 33. Moskala, E. J.; Coleman, M . M . Polymer. Commun. 1983, 24, 206. 34. Moskala, E. J.; Varnell, D. F.; Coleman, M . M . Polymer 1985, 26, 228. 35. Coleman, M . M.; Moskala, E. J.; Painter, P. C.; Walsh, D. J.; Rostami, S. Polymer 1983, 24, 1410. 36. Walsh, D. J.; Higgins, J. S.; Rostami, S. Macromolecules 1983, 16, 388. 37. Ting, S. P.; Bulkin, B. J.; Pearce, E. M.; Kwei, T. R. J. Polym. Sci., Polym. Chem. Ed. 1981, 29, 1451. 38. Skrovansk, D. J.; Coleman, M . M . Polym. Eng. Sci. 1987, 27, 857. 39. Lu, F. J.; Benedetti, A.; Hsu, S. L. Macromolecules 1983, 16, 1525. 40. Snyder, R. G.; Zerbi, G. Spectrochim. Acta Part A 1967, 23A, 391. 41. Sargent, M.; Koenig, J. L. Vibrational Spectrosc. 1991, 2, 21. RECEIVED for review May 14, 1991. ACCEPTED revised manuscript June 11, 1992.


Structure-Property Relations in Polymers - American Chemical Society

Recommend Documents