Thermal and spectroscopic properties of zinc perchlorate/poly(vinylpyrrolidone) blends and a comparison with related hydrogen bonding systems

(1)

Thermal and spectroscopic properties of zinc

perchlorate/poly(vinylpyrrolidone) blends and a comparison with related

hydrogen bonding systems

Shiao-Wei Kuo, Chih-Feng Huang, Chung-Hsi Wu, Feng-Chih Chang*

Institute of Applied Chemistry, National Chiao-Tung University, Hsin-Chu 30050, Taiwan, ROC

Received 13 April 2004; received in revised form 6 July 2004; accepted 13 July 2004 Available online 30 July 2004

Abstract

We have investigated the thermal and spectroscopic properties of blends of poly(vinylpyrrolidone) (PVP) with zinc perchlorate. Analyses

by differential scanning calorimetry indicates that blending with zinc perchlorate increases the values of Tgof PVP. We calculated the

interaction strength of the zinc salt/PVP blends based on an extended configuration entropy model. The presence of ion–dipole interactions between PVP and the zinc salt was confirmed based on Fourier transform infrared (FTIR) and solid-state NMR spectroscopies, which suggest

that the zinc cations coordinate with the carbonyl groups of PVP. The single value of T1rH measured by solid-state NMR spectroscopy

observed for all the zinc salt/PVP blends is smaller than that of pure PVP, which is a finding that indicates that the domain size of this blend system decreases upon increasing the zinc salt content. Based on FTIR and solid-state NMR spectroscopic analyses, we conclude that the ion–dipole interactions in the zinc salt/PVP blend are stronger than the hydrogen bonds in systems such as the poly(vinylphenol) (PVPh)/PVP blend and the PVPh-co-PVP copolymer.

Keywords: Hydrogen bonding; Zinc salt; Solid state NMR

1. Introduction

Polymer blending is a convenient and attractive route for obtaining new polymeric materials that have desirable combination of physical properties. Most polymers pairs, however, are immiscible because the high degree of polymerization of polymers results in a vanishingly small entropy term for the free energy of blending. To enhance the formation of one single-phase system in polymer blends, it is necessary to ensure that specific interactions, such as ionic, hydrogen bonding, and dipole–dipole interactions, exist between the two components. Recently, polymer blending with inorganic salts has become an interesting method for producing new materials with better thermal, mechanical, and conductivity properties, as well as for

materials that can separate olefin/paraffin gas mixtures [1–4].

Increasing both the intermolecular cohesion and glass transition temperatures of polymers is a subject of long-standing interest. Polymers possessing higher glass tran-sition temperatures are attractive in polymer science as a result of the strong economic incentives that arise from their potential applications. In a previous study[5], we found that the values of Tgof PVPh/PVP blends are higher than their

mother polymers. Furthermore, based on DSC and Fourier transform infrared (FTIR) and solid-state NMR spectro-scopic analyses, the values of Tgand the hydrogen bonding

strengths of PVPh-co-PVP copolymers are greater than those of their corresponding PVPh/PVP blends at the same mole fractions of PVPh[6,7]. In addition, the values of Tgof

PVPh-co-PVP-co-POSS nanocomposites are increased dra-matically in comparison with the corresponding PVPh-co-PVP copolymers because of the strong hydrogen bonding that exists in this system [8]. PVP is amorphous and possesses a high value of Tgbecause of its rigid pyrrolidone

www.elsevier.com/locate/polymer

* Corresponding author. Tel.: C886-357-270-77; fax: C886-357-195-07.

(2)

group, which is capable of forming a variety of complexes with many inorganic salts[9–11]. The variations of Tgwith

respect to the zinc salt concentration are of special interest to researchers studying polymer electrolytes. The incorpor-ation of zinc salts into a PVP matrix causes the polymer chains to rigidify as a result of the physical cross-linking that occurs between the PVP chains and the metal atoms; this process results in the value of Tg increasing. In this

study, therefore, we have investigated, by DSC and FTIR and solid-state NMR spectroscopic analyses, the mechanism by which the enhancement of the values of Tgof PVP occurs

upon blending it with zinc perchlorate and provide a comparison of this system with regard to related hydrogen bonding systems.

Solid-state NMR and FTIR spectroscopies are powerful tools for characterizing the structural details of polymer interactions because these specific interactions affect the local electron density and, consequently, corresponding frequency shifts can be observed[12–15]. Moreover, solid-state NMR spectroscopy can be used to measure the phase behavior and the morphology of polymer blends, which can be estimated from the proton spin–lattice relaxation time in the rotating frame ðT_1rHÞ: In this paper, we report the complex interactions that exist between PVP and zinc perchlorate, and provide direct evidence to support the nature of the proposed ionic interactions.

2. Experimental 2.1. Materials

Poly(vinylpyrrolidone) (PVP) having MnZ58,000 g/mol

was obtained from the Acros Chemical Company. Zinc perchlorate hexahydrate [Zn(ClO4)2$6H2O], obtained from

Aldrich, was dried in a vacuum oven at 70 8C for 24 h. N,N-Dimethylformamide (DMF) was used as received.

2.2. Preparation of polymer–salt mixtures

Zinc salt/PVP blends having varying zinc content were prepared by dissolving the desired amounts of the polymer and Zn(ClO4)2$6H2O in DMF. After stirring continuously

for 12 h, these solutions were maintained at 50 8C for 24 h on Teflon plates to facilitate desolvation, and then dried further under vacuum at 90 8C for 2 days.

2.3. Characterization

Thermal properties were measured using a TA Instru-ment DSC 2010. The measureInstru-ments were conducted under a nitrogen flow rate of ca. 25 ml/min. All samples were heated to 300 8C, subsequently annealed for 3 min, rapidly cooled to 20 8C, and then reheated. A heating rate of 20 8C/min was used in all cases. The glass transition temperature (Tg) was

taken as the midpoint of the transition. All IR spectra were

recorded in the range 4000–400 cmK1 at a spectral resolution of 1 cmK1after 32 scans using a Nicolet Avatar 320 FTIR spectrophotometer. IR spectra of mixture films were measured using the conventional KBr disk method: a DMF solution of the mixture was cast onto a KBr disk and the solvent was evaporated at 90 8C for 24 h. IR spectra were obtained at 120 8C by using a cell mounted inside the temperature-controlled compartment of the spectrometer. The film used in this study was sufficiently thin to obey the Beer–Lambert law. High-resolution solid-state 13C NMR spectroscopy experiments were carried out at room temperature using a Bruker DSX-400 spectrometer operat-ing at resonance frequencies of 399.53 and 100.47 MHz for

1

H and13C nuclei, respectively. The13C CP/MAS spectra were measured using a 3.9 ms 908 pulse, a 3 s pulse delay time, and an acquisition time of 30 ms, with a total of 2048 scans. All NMR spectra were recorded at 25 8C using broad-band proton decoupling and a normal cross-polarization pulse sequence. A magic-angle sample-spinning (MAS) rate of 5.4 kHz was used to avoid absorption overlapping. The proton spin–lattice relaxation time in the rotating frame ðTH

1rÞ was determined indirectly by observing carbon nuclei using a 908-t-spin-lock pulse sequence prior to cross polarization. The data acquisition was performed by 1H decoupling with delay times (t) ranging from 0.1 to 12 ms and a contact time of 1.0 ms.

3. Results and discussion 3.1. Analyzing thermal properties

Fig. 1 shows DSC analyses of zinc salt/PVP blends having various compositions. Clearly, the value of Tg

increases as the zinc salt concentration increases, with the maximum enhancement of Tg being ca. 47 8C for PVP

containing 40 wt% of the zinc salt, and then the values of Tg

decrease at higher zinc salt concentration. Recently, Kim et al. proposed[16]an extended configuration entropy model to predict the behavior of Tg of complexing polymer–salt

systems: lnTg12 Tg1 Z b 1 K g12ln z K 1 e f1 r1 ln f1C f2 r2 ln f2 K f2 r2vm 4 3AdissI3=2tðI 1=2 Þ (1) where Tg1and Tg12are the glass transition temperatures of

the pure polymer and of the polymer–salt blend, respect-ively, and fiand riare the volume fraction and the degree of

polymerization, respectively, for component i. Here, r2Z1 and bZzR/M1uDCpp, where z (Z12) is the lattice

coordination number, R is the gas constant, M1u is the

molecular weight of the repeating unit of the polymer, and

S.-W. Kuo et al. / Polymer 45 (2004) 6613–6621 6614

(3)

DCpp is the isobaric specific heat of polymer. The

proportionality constant r12 represents the strength of the

interaction between the polymer and salt. The equation ðf2=r2vmÞð4=3ÞAdissI3=2tðI1=2ÞZADH is extended to account for the cation–anion interaction.Table 1lists the molecular weight, Tg, specific heat, and density of the components

used in this study. It is well known that the presence of the zinc salt leads to retardation of chain mobility, which results in higher values of Tg. Polymer–cation interactions occur

commonly in metal salt/polymer blend system, but an excessive metal salt content tends to decrease the value of the polymer’s Tgbecause the interchain distance increases

and micro-phase separation occurs above a certain opti-mized concentration[17].Fig. 2shows the variation in Tgas

a function of zinc salt content; the solid lines, calculated using Eq. (1), imply that good correlations exist between the experimental data and model’s predictions. The interaction parameters of the zinc salt/PVP blend (r12Z1.50, ADHZ 0.06) are lower than those of the zinc salt/poly(4-vinylpyridine) blend system (r12Z3.60, ADHZ0.06) [18]. This result is consistent with the findings of a previous study in which the ionic interactions in the poly(vinylpyridine) blend system were determined to be greater than those in the PVP blend system[19].

3.2. Infrared spectroscopic analyses

The most frequently employed method for quantifying the relative fraction of free and bonded carbonyl sites within a PVP polymer chain is to measure the extent of such interactions by monitoring the carbonyl stretching bands in the IR spectra as a function of the blend composition.Fig. 3 shows the carbonyl stretching region, in the range 1520– 1760 cmK1, of the IR spectra of zinc salt/PVP blends recorded at 120 8C. The stretching band of the free carbonyl group of uncomplexed PVP appears at 1680 cmK1. This band broadens gradually as the zinc salt content increases and a new band, at ca. 1615 cmK1, appears and grows. This new band corre-sponds to the coordination interaction between the zinc ion and the carbonyl oxygen atom. The fraction of the complexed carbonyl groups can be measured at different zinc salt ratios by decomposing the carbonyl stretching region into two Gaussian peaks, as is shown inFig. 4.Table 2summarizes results from curve fitting of these peaks; we observe that the fraction of complexed carbonyl groups increases upon increasing the zinc salt content. These results suggest that Zn(ClO4)2dissociates

into the PVP matrix through the coordination of zinc ions with carbonyl oxygen atoms and that one zinc cation binds to each carbonyl oxygen atom of PVP.

Fig. 1. DSC scans of zinc salt/PVP blends having a variety of compositions.

Table 1

Molecular weights, values of Tg, specific heats, and densities of the components used in this study

Mw(g/mol) Tg(8C) DCpp(J/(kg K)a Density (g/cm3) PVP 58,000 173 2080 1.563 Zn(ClO4)2$6H2O 372 – – 2.052b Blend system z b r12 A DH PVP/Zn(ClO4)2 12 0.6809 1.5 0.06 a

Values obtained from the group contribution method[22].

b

(4)

The change in the relative intensity and the location of the ClOK

4 anion within the complex depend on the concentration of the Zn(ClO4)2 salt; consequently, these

changes can be attributed to interactions involving the ClOK 4:

Fig. 5showns IR spectra of pure PVP and various zinc salt/ PVP complexes, presenting the ClOK

4 stretching bands in the

range from 650 to 600 cmK1, recorded at 120 8C. Within this region, the absorptions at 627 and 635 cmK1, which represent nðClOK

4Þ vibrations, correspond to free and contact ions, respectively[20]. When the zinc salt concentration is increased, the contact ion band shifts to higher frequency and the band becomes asymmetric. The asymmetric shape of the band may be attributed to the existence of both free ions and ion pairs.Fig. 5also shows the deconstruction of the bands of ClOK

4 anion of the pure zinc salt into two Gaussian peaks. Table 3 summarizes results from curve fitting and indicates that the fraction of ClOK

4 anions existing in ion pairs increases as the zinc salt content increases; this finding can be interpreted as the ClOK

4 anion interacting gradually with zinc cation to facilitate the cation–anion interaction with increasing zinc salt concentration.

3.3. Analyses by solid state NMR spectroscopy

Solid-state NMR spectroscopy provides further insight into the interaction behavior, phase behavior, and mor-phology of the polymer blends formed from PVP and the zinc salt. Upon coordination, the coordinating units experience small perturbations in the magnetic shielding of their nuclei, which results in downfield chemical shifts of their corresponding carbon atoms relative to those of the free polymers. Figure 6 shows the 13C CP/MAS NMR spectra of pure PVP and its blends with zinc perchlorate; Scheme 1indicates the atom numbering used to assign these peaks. The spectrum of pure PVP exhibits peaks for six major resonances. Table 4 summarizes the values of the chemical shifts observed in the13C CP/MAS NMR spectra of the zinc salt/PVP blends. Compared with the 13C CP/ MAS NMR spectra of the pure polymers, the spectra of the zinc salt/PVP blends display significant changes, especially for the resonances of the carbon atoms involved in intermolecular ion–dipole interactions. For example, the

Fig. 2. Variations in the values of Tgof PVP blended with the zinc perchlorate; the solid lines are values calculated from the configuration entropy model.

Fig. 3. FTIR spectra, presenting the region 1760–1520 cmK1_{, of weight}

fraction of the zinc salt/PVP blends recorded at 120 8C.

(5)

carbonyl resonance of PVP is shifted downfield by ca. 2 ppm at a zinc salt content of 40 wt%. It is well known that intermolecular interactions in polymer blends can affect the chemical environments of the interacting molecules, which can result in relatively downfield chemical shift. The signals of the C2, C4, and C5 atoms, as well as the carbonyl group’s resonance (C-6), shift downfield upon increasing the zinc

salt concentration, as indicated in Table 4. These carbon atoms display the character of being positioned adjacent to electron donating units: the C4 and C5 atoms are adjacent to N atoms and the C2 atom is near the carbonyl group. The strong electron withdrawing character of the N atom and carbonyl is able to function as a quasi-cation within the PVP chain to attract the anion that result in down-filed shift. In

(6)

addition, the resonance of the C1 atom is practically independent of the concentration of the zinc salt, which is consistent with our argument because this carbon atom is located far from the coordinating groups.

3.4. Proton spin lattice relaxation time in the rotating frame analyses

We examined the spin lattice relaxation times in the rotating frame ðT_1rHÞ to measure the homogeneity of the zinc salt/PVP blends on the molecular scale. According to the spin locking mode employed in this study, the magnetiza-tion of resonance is expected to decay according to the exponential function model

MtZ MoexpbKt=T1rðHÞc (2)

where T1rH is the spin lattice relaxation time in the rotating frame, t is the delay time used in the experiment, and Mtis the corresponding resonance. The value of T1rH can be obtained from the slope of the plot of lnðMt=MoÞ versus t.

Figure 7shows the T1rH relaxation behaviors of these blends (PVP, 176 ppm) and reveals that both pure PVP and the zinc salt/PVP blends exhibit only a single relaxation throughout the whole range of blends, which indicates the good miscibility and dynamic homogeneity of the PVP phase. Table 5 summarizes the values of T_1rH derived from the

Tabl e 2 Cu rve fitting of the area fract ions of the carb onyl stret ching bands in the FTIR spectra of zinc salt/PVP blends reco rded at 120 8C Zn salt/PVP wt ratio Free C a O Ionic inte raction C a O( fi * ) n (cm K 1)W 1/2 (cm K 1) Af (% ) n (cm K 1) W1/2 (cm K 1) Ai (%) 0/100 1682 29 100 – – – 0 5/95 1681 34 95.5 1629 22 4.5 3.5 10/90 1678 40 91.6 1621 22 8.4 6.6 20/80 1677 38 81.5 1620 32 18.5 14.8 30/70 1674 42 70.2 1619 36 29.8 24.6 40/60 1674 36 57.9 1618 41 42.1 35.9 50/50 1671 36 53.2 1617 38 46.8 40.4 60/40 1665 44 44.2 1614 36 55.8 49.3 70/30 1664 36 21.5 1614 40 78.5 73.8 n, wavenumbe r; W1/2 , half width and * f i , fract ion of ionic intera ction.

Fig. 5. FTIR spectra, presenting the region 650–600 cmK1, of the weight fraction zinc salt/PVP blends.

(7)

binary exponential analysis. The blend exhibits a single exponential and the value of T1rH decreases upon increasing the zinc content. This observation indicates that the domain size of the zinc salt/PVP blends decreases, based on one

dimension spin-diffusion equation [21], as a result of the strong ion–dipole interactions between the zinc cations and the carbonyl groups. In addition, T_1rH is a measure of the chain mobility, the deeper slopes suggest that the chain become stiffer (less mobile) due to the increase in physical cross-link density caused by ionic complexation.

3.5. Comparisons with hydrogen bonding systems

Fig. 8shows the carbonyl stretching region, ranging from 1550 to 1730 cmK1, of the FTIR spectra of various pyrrolidone-containing systems recorded at room tempera-ture: 0.1 M ethylpyrrolidone (EPr, a model compound for PVP) in cyclohexane[6], PAS74-co-PVP26 copolymer[6], pure PVP, PVPh74/PVP26 blend [5], PVPh74-co-PVP26 copolymer [6], and the zinc salt/PVP (40:60) blend. The wavenumber and half-width significantly depend upon dipole interactions and hydrogen bonding and ionic interactions in and between the polymer chains. The carbonyl stretching band of the EPr solution is positioned at a higher wavenumber than that of the pure PVP because of decreased probability of carbonyl–carbonyl interactions. Similarly, the half-width of the band for pure PVP at 1680 cmK1is decreased and shifted to higher wavenumber

Table 3

Curve fitting of the area fractions of the free and ion-paired perchlorate ions as determined from the FTIR spectra of zinc salt/PVP blends recorded at 120 8C wt%, Zn salt/PVP Free ClOK

4 Ion pair ClOK4

n (cmK1) W1/2(cm K1 ) Af(%) n (cm K1 ) W1/2(cm K1 ) Ai(%) 5/95 624 9 100 – – – 10/90 624 9 100 – – – 20/80 625 10 84.1 635 8 15.9 30/70 625 12 83.8 636 7 16.2 40/60 626 12 79.4 636 6 20.6 100/0 627 9 77.7 637 5 22.3

n, wavenumber; W1/2, half width.

Fig. 6.13_{C CPMAS spectra of weight fraction of the zinc salt/PVP blends}

recorded at room temperature. Scheme 1. Chemical structure of PVP and its atom numbering scheme. Table 4

Chemical shifts (ppm) observed in the13C CP/MAS/DD spectra of PVP and its zinc salt blends

Zn Salt (wt%) C-1 C-2 C-4 and C-5 C-6 0 19.02 32.17 43.90 176.58 5 19.02 32.37 44.10 176.98 10 19.02 32.38 44.11 177.20 20 19.02 32.57 44.30 177.59 30 19.02 32.77 44.31 177.99 40 19.02 32.98 44.32 178.60

(8)

(1682 cmK1) when the acetoxystyrene monomer is incor-porated into the PVP chain (PAS74-co-PVP26). After deacetylation of PAS-co-PVP into PVPh-co-PVP, however, the carbonyl group absorption shifts from 1682 to 1651 cmK1; this shift is attributed to the formation of hydrogen bonds between the vinylphenol and vinylpyrroli-done segments of the copolymer. In contrast, the same carbonyl stretching frequency of the PVPh/PVP blend shifts to only 1660 cmK1. This observed difference can be interpreted as suggesting that the number and/or strength of the hydrogen bonds within the PVPh-co-PVP copolymer are greater than those in the corresponding PVPh/PVP blend. In addition, the carbonyl band of the zinc salt/PVPZ 40/60 blend has the relatively lowest wavenumber at 1615 cmK1. Similar results are also observed in the solid-state NMR spectroscopic analysis shown inFig. 9 [5,6]: the carbonyl carbon atoms’ resonance in the zinc salt/PVP

blends experiences the greatest downfield shift, when compared with the other hydrogen bonding systems. Clearly, the FTIR and solid-state NMR spectroscopic analyses indicate the relative strength of the specific interactions: zinc salt/PVP blendOPVPh-co-PVP copolymerOPVPh/PVP blendOpure PVPOPAS-co-PVP copolymer.

4. Conclusions

The glass transition temperatures of PVP are increased by 47 8C relative to that of pure PVP when it is blended with zinc perchlorate at 40 wt%. We calculate the interaction strength of the zinc salt/PVP blend to be 1.50 based on an extended configuration entropy model equation. The IR and

Fig. 7. Semi-logarithmic plots of the magnetization intensities of the signal at 176 ppm versus the delay time for the zinc salt/PVP blends at a contact time of 1 ms.

Table 5 Values of TH

1rof pure PVP and its various zinc salt blends determined from proton spin lattice relaxation experiments performed at room temperature

Zn/PVP C-1 C-2 C-4 and C-5 C-6

0/100 13.98 15.21 14.37 11.96

20/80 7.71 7.13 7.59 6.79

40/60 4.28 4.00 4.19 4.35

Fig. 8. FTIR spectra, recorded at room temperature and displaying the region 1730–1550 cmK1_{, of 0.1 M ethylpyrrolidone (EPr, a model}

compound for PVP) in cyclohexane, PAS74-co-PVP26 (mol%) copolymer, pure PVP, PVPh74/PVP26 (mol%) blend, PVPh74-co-PVP26 (mol%) copolymer, and the zinc salt/PVP (40:60) blend.

(9)

solid-state NMR spectroscopic data shows that the PVP has the ability to interact ionically with the zinc salt; these interactions are stronger than those present in related hydrogen bonding systems.

Acknowledgements

The authors thank the National Science Council, Taiwan, Republic of China, for supporting this research financially under Contract No. NSC-92-2216-E-009-018.

References

[1] Rivas BL, Pereira ED, Moreno-Villoslada I. Prog Polym Sci 2003;28: 173.

[2] Hong SU, Jin JH, Won J, Kang YS. Adv Mater 2000;12:968. [3] Ma X, Sauer JA, Hara M. Macromolecules 1995;28:3953.

[4] Balfiore LA, McCurdie MP, Ueda E. Macromolecules 1993;26:6908. [5] Kuo SW, Chang FC. Macromolecules 2001;34:5224.

[6] Kuo SW, Xu H, Huang CF, Chang FC. J Polym Sci, Polym Phys Ed 2002;40:2313.

[7] Kuo SW, Chang FC. Polymer 2003;44:3021.

[8] Xu H, Kuo SW, Lee JS, Chang FC. Polymer 2002;43:5117. [9] Wu HD, Wu ID, Chang FC. Polymer 2001;42:555.

[10] Liu M, Yan X, Liu H, Yu W. React Funct Polym 2000;44:55. [11] Hee J, Hong SU, Won J, Kang YS. Macromolecules 2000;33:4932. [12] Coleman MM, Graf JF, Painter PC. Specific interactions and the

miscibility of polymer blends. Lancaster, PA: Technomic Publishing; 1991.

[13] Hill DJT, Whittaker AK, Wong KW. Macromolecules 1999;32:5285. [14] Kim JH, Min BR, Kim CK, Won J, Kang YS. J Phys Chem B 2002;

106:2786.

[15] Kuo SW, Chang FC. Macromolecules 2001;34:4089.

[16] Kim JH, Min BR, Won J, Kang YS. J Phys Chem B 2003;107:5901. [17] Kim JH, Min BR, Kim CK, Won J, Kang YS. Macromolecules 2002;

35:5250.

[18] Kuo SW, Wu SC, Chang FC. Macromolecules 2004;37:192. [19] Li XD, Goh SH. Polymer 2002;43:6853.

[20] Chen SW, Wu HD, Chang FC. Polymer 2002;43:5011.

[21] Demco DE, Johansson A, Tegenfeldt J. Solid State Nucl Magn Reson 1995;4:13.

[22] van Krevelen DW. Properties of polymer. Amsterdam: Elsevier; 1990. Fig. 9. Dependence of the chemical shift of the carbonyl carbon atoms on

the compositions of the PAS-co-PVP copolymer, the PVPh/PVP blend, the PVPh-co-PVP copolymer, and the zinc salt/PVP blend.