Linear Viscoelasticity and Cation Conduction in Polyurethane

Feb 20, 2018 - phase glass transition temperature, but the degree of microphase separation and ionic conductivity were found to be significantly affec...
4 downloads 6 Views 1MB Size
Download PDF
Article Cite This: Macromolecules XXXX, XXX, XXX−XXX

Linear Viscoelasticity and Cation Conduction in Polyurethane Sulfonate Ionomers with Ions in the Soft Segment−Multiphase Systems Shih-Wa Wang† and Ralph H. Colby*,‡ †

Department of Chemical Engineering and ‡Department of Materials Science and Engineering, Pennsylvania State University, University Park, Pennsylvania 16802, United States S Supporting Information *

ABSTRACT: PEO600-based polyurethane ionomers with various hard segment contents were synthesized and characterized by both linear viscoelastic (LVE) properties and dielectric relaxation spectroscopy. The ions were placed in the soft segment to achieve better ionic conductivity while the hard phase can provide mechanical strength. Microphase separation was observed in all samples with more than 23 wt % hard segment. The samples that show evidence of microphase separation share similar soft phase glass transition temperature, but the degree of microphase separation and ionic conductivity were found to be significantly affected by specimen preparation method (hot pressed or solution cast). Both ionic conductivity and polymer chain mechanical relaxation show VFT or WLF temperature dependence. At 150 °C, the microphaseseparated samples were found preserving both the ionic conductivity and mechanical modulus. While most literature focuses on gel polymer electrolytes or block copolymers to obtain both high modulus and high conductivity in single-ion conductors, our polyurethane ionomers demonstrate an alternative path to simultaneously high modulus and ionic conductivity.

■

INTRODUCTION Polyethylene glycol (PEG or PEO)-based polymer electrolytes, polyelectrolytes, and single-ion conductors have drawn the attention of many researchers, since PEO was found to have excellent ability to solvate cations.1−4 The flexible ether linkage and low Tg of amorphous PEO facilitate ion transport and boost ionic conductivity but at the same time sacrifice the mechanical strength,5,6 which is very important for ionconducting polymer applications including lithium ion battery membranes and ionic actuators. In order to meet both requirements of ionic conductivity and mechanical strength, cross-linked PEO-based gel is a popular option but requires solvents to boost conductivity.3,6,7 Besides polymer gels, materials with both nanophase separated hard phase (with good mechanical strength) and soft phase (fast relaxation and good conductivity) are also potential candidates. Instead of block copolymers,8−12 we turn to PEO-based polyurethanes; such segmented polymers spontaneously microphase separate into desired hard and soft microphases. Meanwhile, we select single-ion conductors, rather than polymer−salt systems, for their advantage of high transference number.13 There are several ways to prepare polyurethane anionomers:14,15 (1) replacing the urethane proton by ioncontaining groups;16,17 (2) introducing the ionic group into the hard segment by ionic chain extenders such as dimethylolpropionic acid (DMPA);18−21 (3) end-capping the diisocyanate−polyol prepolymer with monofunctional ion-containing alcohol;22,23 and (4) attaching ionic groups in the soft segment by using ion-containing diols.24−26 Replacing the urethane proton interferes with microphase separation due to loss of ability to form hydrogen bonds through −NH groups. Our © XXXX American Chemical Society

previous study suggests that placing ionic groups in the hard segment increases ionic conductivity but reduces mechanical strength because ions in the hard segment compete for hydrogen bonds.27 Polyurethane sulfonate ionomers based on p-phenylene diisocyanate (pPDI) with no chain extender, which contain ions in the soft segment, do not microphase separate and thus have low modulus above Tg.28 In this paper, the same PEO-based sulfonate-centered diol with M ≈ 1300 g/mol was used as the soft segment, and the urethane linkage is still pPDI. Butanediol was added as a chain extender to investigate the hard segment content needed for microphase separation and its effects on morphology, ionic conductivity, viscoelastic, and dielectric properties.

■

MATERIALS AND SYNTHESIS METHODS

Materials. p-Phenylene diisocyanate (pPDI), dimethyl fumarate, sodium bisulfite, 1,4-butanediol, terephthaloyl chloride, fumaryl chloride, potassium carbonate, and dibutyltin oxide were purchased from Sigma-Aldrich. Poly(ethylene glycol) 600 (PEG600, Mw = 600 with Mw/Mn < 1.2 from the manufacturer and confirmed by aqueous SEC and 1H NMR in DMSO-d6) was purchased from TCI America, Inc. Anhydrous N,N-dimethylformamide (DMF), dimethyl sulfoxide (DMSO), tetrahydrofuran (THF), toluene, methanol, and diethyl ether were purchased from EMD. All reagents were used as received without further purification, except pPDI, which was purified by sublimation at 80 °C overnight as described in the literature,29 and PEG600 was dried under vacuum at 80 °C for at least 24 h to remove water (with less than 100 ppm by Karl Fischer titration). Received: November 27, 2017 Revised: February 20, 2018

A

DOI: 10.1021/acs.macromol.7b02510 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules Scheme 1. Chemical Structure of SC-Diol Based Polyurethane Ionomersa

a

The EO/Na+ ratio is fixed at 24. x = 1, 0.5, 0.33, and 0.25. process took ∼10 days, and the final products were vacuum-dried at 80 °C.

Polyurethane (PU) Ionomer Synthesis. Synthesis of PU Ionomers with Sulfonate-Centered Diol (SC-Diol). The synthesis method of SC-diols was described previously.28 The prepared SC-diol was dried at 80 °C under vacuum overnight before use. 1,4-Butanediol was dried with molecular sieves for a week (moisture content was below 100 ppm by Karl Fischer titration). Dried SC-diol, 1,4butanediol, and pPDI (with −NCO:−OH = 1:1) were dissolved in anhydrous DMF and then reacted at 60 °C for 5−6 h until the −NCO absorption peak at ∼2270 cm−1 disappeared. The crude products were obtained by precipitation with excess diethyl ether and vacuum-dried at 80 °C. The chemical structure is shown in Scheme 1. Synthesis of Polyester Ionomer Based on SC-Diol. Vacuum-dried PEG600 dissolved in THF and potassium carbonate was charged into a three-neck flask and further dried by stirring with molecular sieves for 2 days and cooled in an ice bath before reaction. The fumaryl chloride−THF solution was dropped into the dried solution above for 1 h, and the solution was stirred for another hour. The terephthaloyl chloride−THF solution was then added dropwise for 1 h. The solution (20 wt % in THF) was allowed to warm up to room temperature and react overnight. THF was then removed by rotary evaporation. The product was then dissolved in methanol and mixed with sodium bisulfite (10% excess) aqueous solution, and the solution was allowed to react at 80 °C overnight. The chemical structure is shown in Scheme 2.

■

Differential Scanning Calorimetry (DSC). The calorimetric glass transition temperature Tg was measured using a Seiko Instruments SSC/5200. All samples were dried under vacuum at 120 °C for 2 days before measurement. All samples were first heated to 120 °C, annealed for 3 min, and cooled down to −80 °C at 10 K/min. Tg was then determined from the change in heat capacity during the second heating at a heating rate of 10 K/min under an ultrahigh purity nitrogen purge. Thermogravimetric Analysis (TGA). The thermal degradation was probed by a TA Instruments Q50. All samples were dried under vacuum at 120 °C for 2 days before measurement. All samples were heated to 800 °C at a heating rate of 10 K/min with 20 mL/min nitrogen purge, and an upper bound on the degradation temperature Td was determined by the onset of mass loss (intersection of stable and degradation curves). Dielectric Relaxation Spectroscopy (DRS). Liquid samples were dried at 120 °C under vacuum for 1 day to remove moisture and then sandwiched between two freshly polished brass electrodes with 100 μm Teflon spacers. The prepared cells were annealed in a vacuum oven for an additional 24 h at 120 °C before measurement. For solid samples, films were prepared by either solution casting or hot pressing. Solution cast samples were prepared from 10 wt % DMSO polymer solution (∼100 μm thick), and hot pressed samples were prepared at 160 °C under a pressure of 5000 lb (∼200−300 μm, no spacer). All prepared sample films were sandwiched between two freshly polished brass electrodes without spacers, dried, and annealed in a vacuum oven at 120 °C for 2 days. Dielectric (impedance) spectra were measured using a Novocontrol GmbH Concept 40 broadband dielectric spectrometer in the frequency range of 1 × 10−2 to 1 × 107 Hz with 0.1 V amplitude. Samples were annealed at 150 °C for 30 min in the instrument and allowed to equilibrate at each progressingly lower temperature for 5 min before each isothermal measurement. The measurements were done from high to low temperature and then from low to high temperature to confirm no changes due to measurement history. Linear Viscoelasticity (LVE). PU23 and PU32 were vacuummolded at 160 °C for 20 min to prepare approximately 1 mm thick, 8

Scheme 2. Chemical Structure of SC-Diol-Based Polyester Ionomera

a

EXPERIMENTAL METHODS

The EO/Na+ ratio is 24.

Purification. The crude polymer products were either dissolved (PE and PU10, water-soluble) or soaked (PU23, PU32, and PU40, insoluble but swell in water) in deionized water and injected into a Slide-A-Lyzer G2 dialysis cassette with 3500 molar mass cutoff, which was then immersed in fresh deionized water (changed every 2−8 h) until the dialysate reached a conductivity lower than 2 μS/cm. The

Table 1. Chemical Composition of SC-Diol-Based Polyurethane and Polyester Ionomers and Their Physical Properties xa

sample SC-diol600 PE PU10f PU23 PU32 PU40h

hard segment contentb (wt %)

f

1 0.5 0.33 0.25

0 10 23 32 40

physical state at room temp viscous liquid viscous liquid viscous liquid solid, G′ = 1 MPa solid, G′ = 10 MPa solid

ion content p0c (nm−3)

Tgd (°C)

Tde (°C)

0.48 0.44 0.43 0.37 0.32 0.29

−30 −33 −5 −1, 2 −2, 2 2

237 238 247 243 259 257

g

a

x is the fraction of diols that are SC-diol600 shown in Scheme 1. bHard segment content = (diisocyanate + butanediol)/(diiscocyanate + butanediol + SC-diol). cAssuming all samples have the same density of 1.1 g/cm3. dSoft phase Tg measured by DSC. For PU23 and PU32, the first temperature listed shows the Tg of hot pressed films and the second is the Tg of the DMSO solution cast films. eOnset of thermal degradation in TGA measurement, thought to be an upper bound as DSC and LVE indicate degradation starts near 200 °C. fSC-diol600 and PU10 are the same samples as SC-diol600 and PU600s in ref 28. gSC-diol600 also shows a small melting endotherm on heating at 15 °C. hPU40 cannot be easily formed into a nice film by either DMSO solution casting or hot pressing, which means it cannot be fully characterized. B

DOI: 10.1021/acs.macromol.7b02510 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules mm diameter pellets. All samples were annealed at 120 °C under vacuum for 2 days before measurement. Linear viscoelastic response was measured under oscillatory shear using a Rheometrics RDS-II with Bendix spring transducer. PU23 and PU32 samples were annealed at 180 °C between two 8 mm parallel plates for 30 min with compressional normal force to ensure good contact and tested between 180 and 30 °C. PU10 was tested with 25 mm parallel plates from 80 to 25 °C. All samples were allowed to equilibrate at each temperature for 25 min before each isothermal frequency sweep. Small-Angle X-ray Scattering (SAXS). Samples were prepared by the same methods as described in the DRS section. Sample thickness (0.1−0.6 mm) and Kapton (two sheets with a 0.15 mm total thickness) and empty background (with Kapton) were also recorded for data reduction. The data were collected on a Molecular Metrology SAXS instrument using Cu Kα radiation (λ = 1.54 Å) and a twodimensional multiwire detector. The sample-to-detector distance was 150 cm. At least 1 000 000 counts were collected for good quality of data (with transmission above 70%), which takes 4−6 h.

■

Figure 1. SAXS spectra of PU ionomer samples. Intensity was corrected by background (empty cell with only Kapton) and normalized by specimen thickness. Open symbols represent samples prepared by DMSO solution casting, and filled symbols represent samples prepared by hot pressing (except PU10).

RESULTS AND DISCUSSION

Chemical Structure and Physical Properties. Table 1 summarizes the chemical composition of the ionomers studied in this paper and some of their basic physical properties. PE is the polyester form of corresponding PU ionomers (without −NH− group). The two-digit number XX in PUXX represents hard segment (HS) content in wt % of PU ionomer samples calculated from the feed. Both PE and SC-diol600 were used as reference samples with no hard segment and only have a single phase30 with a few ion aggregates.31 Both of these two samples have a soft microphase Tg around −30 °C and are viscous liquids at room temperature, indicating the extra ester linkage in PE does not have strong interaction to raise Tg. In contrast, Tg = −5 °C of PU10 is 25 °C higher owing to the presence of −NH− groups acting as strong hydrogen bond donors.28 Note that PU10 is also a viscous liquid at room temperature and shows no evidence of microphase separation in SAXS (Figure 1). However, the Fox equation does not work in PU10 by using SC-diol600 Tg = −30 °C as the soft segment Tg, as that suggests a hard phase Tg over 2000 °C. The deviation from the Fox equation has been interpreted to imply microphase separation32,33 and/or different heat capacity contributions of hard and soft segments even in a single phase.34 Samples with higher HS content start to microphase separate and are soft and rubbery pale yellow powders at room temperature, with a soft phase Tg near 0 °C regardless of HS content. This has been observed in the nonionic polyurethane literature35 and implies the amount of HS mixed in the soft phase is a constant once it reaches the critical concentration to microphase separate and/ or more ions are trapped with increased hard phase content. It is interesting that there is no significant difference (±5 °C) in soft phase Tg with samples from different preparation methods. No hard phase Tg or Tm was observed in DSC before thermal degradation above 200 °C. Although the TGA measurement indicates the onset of thermal degradation is above 230 °C and the degradation temperature increases with HS content, sample degradation was observed below 230 °C during heating (charred sample), suggesting the degradation occurs slowly at lower temperature. The solubility is also affected by HS content. For HS content less than 10 wt %, ionomers are soluble in polar solvents including water, DMSO, and DMF. PU23 and PU32 are soluble in DMSO but have poor solubility in water and DMF. PU40 can only swell in DMSO and is insoluble in both water and DMF.

Small-Angle X-ray Scattering and Morphology. In order to further investigate microphase separation, SAXS spectra were collected and are shown in Figure 1. PU10, which is a viscous liquid at room temperature, shows minimal scattering, indicating single-phase morphology. In contrast, PU23 and PU32, prepared by either DMSO solution casting or hot pressing, show clear microphase separation peaks. The peak position represents the spacing between two hard domains.36,37 The spacing of ∼14 nm can be calculated by d = 2π/qmax and is reported in Table 2. Hot pressed PU32 shows similar spacing as Table 2. Spacing between Hard Domains from SAXS PU23

PU32

sample

hot pressed

DMSO cast

hot pressed

DMSO cast

spacing (nm)

13.2

15.6

13.7

14.7

hot pressed PU23 but with significantly larger intensity, indicating more microphase separation and/or better contrast between hard and soft phases. Both DMSO cast PU23 and PU32 films show significantly larger spacing than hot pressed samples, while PU23 shows lower intensity but PU32 shows higher intensity compared to the hot pressed samples. Solution casting from DMSO solution helps polymer chains relax and reach a state closer to equilibrium, with microphase separation better developed, hence the slightly larger spacing. It is interesting that the two processing methods give opposite results in PU23 and PU32. This implies the morphology (thus ionic conductivity which will be discussed later) can be significantly affected and can perhaps be tuned by the processing method, but the mechanisms will need further investigation. Here we provide one possible explanation: the presence of DMSO allows urethane linkages to rearrange, interact more with the soft segment, and possibly release the soft segment trapped in the hard domain in hot pressed samples, resulting in larger spacing and lower intensity (less amount of hard domain) for PU23. On the other hand, DMSO also allows the urethane linkage in PU32 to rearrange and interact with ion-containing soft segment, but the larger amount of hard segment and urethane linkage also get a chance to trap the ion-containing soft segment since it is believed that −NH groups can strongly interact with −SO3− C

DOI: 10.1021/acs.macromol.7b02510 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules anions in the soft segment,38 imparting better contrast and higher intensity. Ions trapped in the hard domain will be revisited in a later discussion.

PU10 and ionomers discussed in our previous study28). This is very likely because interfacial polarization (Maxwell−Wagner− Sillars, MWS, relaxation) occurs on a similar time scale, and the tan δ peak is the combination of both MWS relaxation and EP. This also implies that the predictions of τσ and τEP (as well as p and μ discussed below) become less reliable with high degree of microphase separation. At any given instant, the number density of simultaneously conducting ions, p, and their mobility, μ, can be obtained, with L the sample film thickness: p=

μ=

Ionic Conductivity. Figure 2 shows the temperature dependence of ionic conductivity. The SC-diol600 has the highest ionic conductivity, providing an upper limit of possible conductivity if microphase separation were complete. PE and PU10 also have higher conductivity compared to other samples due to their lower Tg. Note that conductivity of PE and PU10 merge at high temperature, suggesting similar ion conduction in PE and PU10. PU23 and PU32 have the same soft phase Tg but different ionic conductivity. At the same time, sample preparation (thus morphology) also affects conductivity; however, the effect is very different in PU23 and PU32. For PU23, the DMSO cast sample shows higher conductivity while the hot pressed sample has higher conductivity for PU32. This phenomenon shows good correlation with SAXS results: samples with stronger SAXS peaks (hot pressed PU23 and DMSO cast PU32) have lower conductivity. One possible explanation is that some ions are trapped in the hard phase, contributing to stronger SAXS peaks and leaving fewer ions for conduction in the soft phase (discussed next). Conducting Ion Concentration and Mobility. To further investigate ion transport, an electrode polarization model (EP model, Macdonald/Coelho model)39−41 was applied to separate conducting ion content and their mobility at any given instant in time. For a single-ion conductor with blocking electrodes under an a.c. electric field, the conductivity and the polarization can be described as σDC = eμp (1) ωτEP ε″ = ε′ 1 + ω 2τστEP

εEPε0 σDC

(5)

eL2τσ 4τEP 2kT

⎛ −E ⎞ p = p∞ exp⎜ a ⎟ ⎝ kT ⎠

(6)

(7)

where Ea is an activation energy and p∞ is the conducting ion content at infinite high temperature. Note that p∞ here represents the total number density of Na+ counterions that are available to contribute to conductivity under an applied electric field and is not necessarily equal to the theoretical total ion content p0 for microphase separated ionomers.

Figure 3. Normalized concentration of simultaneously conducting ions as a function of temperature. Solid lines are best fits to the Arrhenius equation (eq 7) with fitting parameters in Table 3.

Table 3 gives the fitting parameters of the Arrhenius equation for conducting ion content (eq 7). At any given instant, less than 10% of the Na+ are participating in conduction and the majority of Na+ are in pair or aggregate states. PE, PU10, DMSO cast PU23, and hot pressed PU32 can be fitted with p∞ = p0 (Table 1), suggesting all Na+ ions are available for conduction. But p∞ = p0 does not yield reasonable fits for hot pressed PU23 and DMSO cast PU32. p∞ of hot pressed PU23 and DMSO cast PU32 needs to be less the 10% of their p0, indicating that the majority of Na+ ions are trapped in the hard domains and cannot contribute to conductivity. This observation agrees with the analysis of SAXS spectra and also explains the lower conductivity of hot pressed PU23 and DMSO cast PU32 in Figure 2. PE has slightly lower activation energy compared to our previous studies30,41 on PEO-based polyester single-ion conductors, possibly due to different anion

(2)

where τσ and τEP are time scales of ion conduction and full electrode polarization: εε τσ ≡ s 0 σDC (3)

τEP ≡

e 2L2τσ

Figure 3 shows conducting ion concentration divided by theoretical ion content, p0, as a function of temperature. It is clear that the conducting ion concentration roughly follows the Arrhenius equation:

Figure 2. Ionic conductivity as a function of reciprocal temperature.

tan δ =

4σDCτEP 2kT

(4)

where εs is the static dielectric constant of polymer, εEP is the dielectric constant when EP is completed, and ε0 is the vacuum permittivity. Please note that the peak of tan δ becomes broader, with smaller magnitude, in microphase-separated samples compared to the homogeneous samples (such as D

DOI: 10.1021/acs.macromol.7b02510 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules

Table 3. Fitting Parameters for Temperature Dependence of Conducting Ion Concentration (Eq 7) and Their Mobility (Eq 8) conducting ion concentration p (nm−3) sample PE PU10 PU23 PU32 a

conducting ion mobility μ (cm2/(V s))

preparation method

Tg (K)

p∞

p∞/p0

Ea (kJ/mol)

log μ0

T0 (K)

D

hot pressed DMSO cast hot pressed DMSO cast

240 268 272 275 271 275

0.44 0.43 0.023 0.37 0.32 0.0063

1a 1a 0.06 1a 1a 0.02

15.0 15.2 13.6 16.3 19.5 13.8

−1.90 −2.29 −1.46 −2.08 −1.86 −0.80

190 222 219 222 215 217

4.4 3.3 4.3 3.9 3.9 3.9

Can be fitted by defining p∞ = p0.

However, for ion-conducting polymers, ion conduction usually dominates the loss spectra and often masks polymer relaxations. A common method to reveal polymer relaxations underneath ion conduction is to use eq 9, derived from the Kramers− Kronig relation, to obtain “conduction-free” derivative spectra.42,44

species (sulfonated phthalate vs sulfonated aliphatic diester). PU10, DMSO cast PU23, and hot pressed PU32 have similar p∞ but lower p at all studied temperatures and larger activation energy with increasing HS content, implying that the urethane linkage impedes ionic conduction, either by directly binding Na+ or by effectively tying up ether oxygens and preventing them from solvating Na+. The activation energy decreases with fewer available conducting ions due to higher EO/Na+ ratio, as reported for PEO-based ionomers when the ion content was varied (see refs 28 and 30). The mobility of these “free” Na+ follows the Vogel− Fulcher−Tammann (VFT) equation and is shown in Figure 4 with fitting parameters summarized in Table 3: ⎛ −DT0 ⎞ μ = μ0 exp⎜ ⎟ ⎝ T − T0 ⎠

π ∂ε′(ω) (9) 2 ∂ ln ω Figure 5 shows the derivative spectra of ionomers with different amounts of HS. From high to low frequency, peaks εder(ω) = −

(8)

where μ0 is the unconstrained (high T) cation mobility, T0 is the Vogel temperature, and D is a parameter that is related to fragility. The Vogel temperatures are found to be 40−60 K below the DSC Tg, which is very common and suggests that the soft phase is always continuous. It is interesting that PE, PU10, and DMSO cast PU23 have the same mobility, while other microphase-separated samples have higher mobility. A possible explanation is that the presence of the hard phase provides wellordered −NH− groups at the interface that can interact with sulfonate groups and allow cations to move along the interface, resulting in higher mobility. On the other hand, this can also simply be the error of EP fitting because of the possible presence of MWS relaxations in the same frequency range. Polymer Dielectric Relaxation. DRS is also a useful tool for studying polymer relaxation and microphase separation.42,43 Polymer relaxation can be probed from dielectric loss spectra.

Figure 5. “Conduction-free” derivative spectra of ionomers with various hard segment content and sample preparation methods at T = Tg + 50 K (±5 K).

corresponding to polymer relaxation are first observed then the electrode polarization starts and shows a large increase in the spectra. εder can be fitted with a power-law based EP and a Havriliak−Negami (HN) function: εder = Aω−s −

π ⎡ ∂ε′HN (ω) ⎤ ⎢ ⎥ 2 ⎣ ∂ ln(ω) ⎦α

(10)

⎧ ⎫ Δε ⎬ ε′HN (ω) = Re⎨ a b ⎩ [1 + (iωτHN) ] ⎭

(11)

εs ≡ lim [ε′α (ω)] + ε∞ = Δεα + ε∞

(12)

ω→0

⎪

⎪

⎪

⎪

The first term in eq 10 represents the EP part with constant A and s (= 1.6−2.0), and the second term represents the α relaxation, where Δε is the relaxation strength, a and b are shape parameters, τHN is a characteristic relaxation time, and εs is determined by eq 3. It is worth noting that although the six samples have different morphology, they can be fitted reasonably with one power-law EP plus one HN function. Meanwhile, as shown in Figure 6a, the α relaxation peak

Figure 4. Mobility of the simultaneously conducting ions as a function of Tg/T. Solid lines are fits to the VFT equation (eq 8) with fitting parameters in Table 3. E

DOI: 10.1021/acs.macromol.7b02510 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules

separation so the Δεα is lower in PU23 and PU32. With the most microphase separation and the most ions trapped in the hard phase, DMSO cast PU32 has the lowest Δεα.

frequency can be nicely reduced into a single curve by normalizing by the DSC Tg. The temperature dependence of the α relaxation frequency shown in Figure 6a follows the VFT equation: ⎛ −DT0 ⎞ ωα(T ) = ωα 0 exp⎜ ⎟ ⎝ T − T0 ⎠

(13)

where ωα0 is the unconstrained (high T) frequency, T0 is the Vogel temperature, and D is a parameter that is related to fragility. The fitting parameters are summarized in Table 4. These VFT fits use the same T0 from fitting conducting ion mobility (Table 3), which is 40−60 K below the DSC Tg.

Figure 7. Correlation between the reciprocal of the time scale of ion conduction τσ and the peak frequency of the α relaxation. The solid line represents eq 14 with B = 3.6 and the dashed line with B = 1.

In microphase-separated PUs, Maxwell−Wagner−Sillars (MWS) interfacial polarization, originating from different dielectric constants between two phases, is commonly observed above their soft phase Tg.45−48 However, we do not observe a clear MWS peak because of the strong α relaxation and EP. The fact that the α relaxation data can be reduced nicely by DSC Tg and shares the same Vogel temperature T0 as the conducting ion mobility (derived from EP) suggests that any MWS relaxation very likely has a magnitude much smaller compared to EP and has relatively little impact on both the α relaxation and our EP analysis. In our previous study,28 single-phase polyurethane ionomers were found to follow the Barton−Nakajima−Namikawa (BNN) correlation,49−52 which suggests the dielectric relaxation and conduction originate from one diffusion process, and the correlation between DC conductivity and dielectric relaxation can be described as σDC 1 = = Bωα ε0εs τσ (14)

Figure 6. (a) Temperature dependence of the α relaxation peak frequency. Solid lines represent the fitting results to the VFT equation (eq 13) with fitting parameters shown in Table 4. (b) Temperature dependence of Δεα. The symbols are identified in the Legend in part (a).

Table 4. Fitting Parameters for Temperature Dependence of α Relaxation fit to the VFT Equation (Eq 13)

with B = 3.6 for PU10.28 Figure 7 shows the correlation between the time scale of ion conduction, τσ, and ion rearrangement τα = 1/ωα. PE, DMSO cast PU23, and hot pressed PU32 have similar B to PU10 as expected. Hot pressed PU23 and DMSO cast PU32 have smaller B simply because a significant amount of the ions are trapped in the hard phase, resulting in higher EO/Na+ ratio in the soft phase. Linear Viscoelastic Properties. Microphase separation of PU can also be probed by linear viscoelastic response.53,54 Figure 8 shows the master curves of PU10, PU23, and PU32. PU10 is a viscous liquid at room temperature while PU23 and PU32 are solids that were vacuum molded at 160 °C. Time− temperature superposition (TTS) applies reasonably well for PU10, suggesting single-phase morphology (with no morphology changes), which agrees with our SAXS results. Both PU23 and PU32 show storage moduli dominating (tan δ = G″/G′ < 1), and TTS only applies at low temperature and starts to fail above 140 °C. The failure of TTS implies breaking of hydrogen bonding and possibly some morphology change. PU23 has a small relaxation in G″ in the temperature range 100−140 °C, but this relaxation was not observed at higher temperature (160

α relaxation sample PE PU10 PU23 PU23 PU32 PU32

preparation method

log ωα0

T0 (K)

D

hot pressed DMSO cast hot pressed DMSO cast

9.9 8.7 9.3 9.2 9.2 9.7

190 222 219 222 215 217

6.3 4.3 5.2 5.1 5.7 6.0

Figure 6b shows the strength of α relaxation, Δεα, as a function of temperature. PE contains no urethane group and has Δεα around 50. The urethane linkages, as well as the resulting hydrogen bonding, have strong dipole and significantly increase Δεα in PU10. A stronger temperature dependence is also observed, probably from the fact the hydrogen bonding is relatively weak and easy to break with temperature. Further increase of urethane linkage (via increasing hard segment content) results in microphase F

DOI: 10.1021/acs.macromol.7b02510 Macromolecules XXXX, XXX, XXX−XXX

Macromolecules

■

and 180 °C). It was found that this relaxation does not show up every time and disappears with longer annealing time. This might be due to unequilibrated hydrogen bonding or incomplete microphase separation. The temperature dependence of the frequency shift factors aT, shown in Figure 9 and Table 5, follows the Williams− Landel−Ferry (WLF) equation: log a T = −

c1(T − Tr) c 2 + (T − Tr)

Article

CONCLUSION Polyurethane ionomers based on sulfonate-centered PEO diols (SC-diols) with various hard segment contents using pPDI and

(15)

where c1 and c2 are constants. The WLF equation is mathematically equivalent to the VFT equation with D ≅ 2.303c1c2/T0 and the Vogel temperature T0 = Tr − c2. The temperature dependence can be fitted with the same Vogel temperatures discussed in previous sections. As expected, the mechanical relaxation shares the same T0 with the dielectric α relaxation and conducting ion mobility.

Figure 10. (a) Relationship between ionic conductivity and storage modulus for PEO-based single-ion conductors in the literature, PU23 and PU32. Dark red diamond: PU23 at 30 °C (open) and 150 °C (filled); teal down triangle: PU32 at 30 °C (open) and 150 °C (filled); pink square: P(STFSILi)-b-PEO-b-P(STFSILi) block copolymer ionomer at 60 °C;11 green square: poly[Sty-b-(Sty-Tf2N-Li-coDEGMEMA)-b-Sty] block copolymer ionomer at 90 °C and 10% RH;55 blue up triangle: polyurethane ionomers22,23,25 at room temperature and 50 °C. (b) Ionic conductivity (black circle symbols) and storage modulus (red triangle symbols) as a function of hard segment content at 30 °C (solid symbols) and 150 °C (open symbols). Note that PU10 (10 wt % hard segment) does not have a separated hard phase and is a liquid at 150 °C with G′ lower than 1 Pa.

Figure 8. Master curves of PU samples at Tr = DSC Tg + 30 K (298 K for PU10, 302 K for PU23, and 301 K for PU32).

butanediol were synthesized. Such polyurethane ionomers with 23 or 32 wt % hard segment microphase separate, while the one with only 10 wt % hard segment (having no butanediols) does not. Polymer dynamics were studied both dielectrically and mechanically. The correlation between ion conduction and polymer relaxation is similar to previous observations in the single phase PU ionomer systems,28 and no clear MWS interfacial polarization was observed. The main goal of this study was to synthesize a multiphase material that can achieve reasonable mechanical strength and ionic conductivity at the same time. Unfortunately, neither PU23 nor PU32 shows good ionic conductivity at room temperature. However, at 150 °C, PU32 shows both high storage modulus and high ionic conductivity. Figure 10a compares the storage modulus and ionic conductivity (hot pressed samples) of PU23 and PU32 at 150 °C with literature data. PU32 outperforms other PU-based ionomers and is comparable to the poly[Sty-b-(Sty-Tf2N-Li-co-DEGMEMA)-bSty] block copolymer.55 This proves the concept of microphase-separated polyurethane ionomers can provide the

Figure 9. Temperature-dependent frequency scale shift factors aT of PU samples. Reference temperature Tr is selected at DSC Tg + 30 K. The solid lines represent fits to the WLF equation (eq 15) with the fitting parameters summarized in Table 5.

Table 5. Temperature Dependence of the Frequency Scale Shift Factor aT Fit to the WLF Equation (Eq 15), the Corresponding VFT Equation Parameters, and the 150 °C Shear Modulus of Polyurethane Ionomers sample

c1

c2 (K)

T0 (K)

D

G′ at 150 °C (Pa)

PU10 PU23 PU32

5.9 6.8 9.2

76 83 86

222 219 215

4.7 6.0 8.4

2 × 105 3 × 106

G

DOI: 10.1021/acs.macromol.7b02510 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules

(7) Chandrasekhar, V. Polymer Solid Electrolytes: Synthesis and Structure. Adv. Polym. Sci. 1998, 135, 139−205. (8) Singh, M.; Odusanya, O.; Wilmes, G. M.; Eitouni, H. B.; Gomez, E. D.; Patel, A. J.; Chen, V. L.; Park, M. J.; Fragouli, P.; Iatrou, H.; Hadjichristidis, N.; Cookson, D.; Balsara, N. P. Effect of Molecular Weight on the Mechanical and Electrical Properties of Block Copolymer Electrolytes. Macromolecules 2007, 40 (13), 4578−4585. (9) Gomez, E. D.; Panday, A.; Feng, E. H.; Chen, V.; Stone, G. M.; Minor, A. M.; Kisielowski, C.; Downing, K. H.; Borodin, O.; Smith, G. D.; Balsara, N. P. Effect of Ion Distribution on Conductivity of Block Copolymer Electrolytes. Nano Lett. 2009, 9 (3), 1212−1216. (10) Soo, P. P.; Huang, B.; Jang, Y.-I.; Chiang, Y.-M.; Sadoway, D. R.; Mayes, A. M. Rubbery Block Copolymer Electrolytes for Solid-State Rechargeable Lithium Batteries. J. Electrochem. Soc. 1999, 146 (1), 32− 37. (11) Bouchet, R.; Maria, S.; Meziane, R.; Aboulaich, A.; Lienafa, L.; Bonnet, J.-P.; Phan, T. N. T.; Bertin, D.; Gigmes, D.; Devaux, D.; Denoyel, R.; Armand, M. Single-Ion BAB Triblock Copolymers as Highly Efficient Electrolytes for Lithium-Metal Batteries. Nat. Mater. 2013, 12, 452−457. (12) Mineart, K. P.; Al-Mohsin, H. A.; Lee, B.; Spontak, R. J. WaterInduced Nanochannel Networks in Self-Assembled Block Ionomers. Appl. Phys. Lett. 2016, 108, 101907. (13) Hallinan, D. T.; Balsara, N. P. Polymer Electrolytes. Annu. Rev. Mater. Res. 2013, 43, 503−525. (14) Jaudouin, O.; Robin, J.-J.; Lopez-Cuesta, J.-M.; Perrin, D.; Imbert, C. Ionomer-based Polyurethanes: A Comparative Study of Properties and Applications. Polym. Int. 2012, 61 (4), 495−510. (15) Nelson, A. M.; Long, T. E. Synthesis, Properties, and Applications of Ion-containing Polyurethanes Segmented Copolymers. Macromol. Chem. Phys. 2014, 215 (22), 2161−2174. (16) Visser, S. A.; Cooper, S. L. Comparison of the Physical Properties of Carboxylated and Sulfonated Model Polyurethane Ionomers. Macromolecules 1991, 24, 2576−2583. (17) Ding, Y. S.; Register, R. A.; Yang, C.-Z.; Cooper, S. L. Synthesis and Characterization of Sulphonated Polyurethane Ionomers based on Toluene Diisocyanate. Polymer 1989, 30, 1204−1212. (18) Yang, C.-Z.; Grasel, T. G.; Bell, J. L.; Register, R. A.; Cooper, S. L. Carboxylate-containg Chain-Extended Polyurethanes. J. Polym. Sci., Part B: Polym. Phys. 1991, 29, 581−588. (19) Krol, P.; Krol, B.; Holler, P.; Telitsyna, N. Polyurethane Anionomers Synthesised with Aromatic, Aliphatic or Cycloaliphatic Diisocyanates, Polyoxyethylene Glycol and 2,2-bis(hydroxymethyl)propionic acid. Colloid Polym. Sci. 2006, 284 (10), 1107−1120. (20) Kakati, D. K.; George, M. H. Polyurethane Ionomers Containing Phosphate Groups. Polymer 1993, 34 (20), 4319−4324. (21) Williams, S. R.; Wang, W.; Winey, K. I.; Long, T. E. Synthesis and Morphology of Segmented Poly(tetramethylene oxide)-based Polyurethane containing Phosphonium Salts. Macromolecules 2008, 41, 9072−9079. (22) Filip, D.; Asandulesa, M.; Macocinschi, D.; Aflori, M.; Vlad, S. Molecular Dynamics, Conductivity and Morphology of Sodium Deoxycholate-based Poly(ester ether)urethane Ionomer Biomaterials. J. Mater. Sci. 2016, 51, 8516−8528. (23) Filip, D.; Macocinschi, D.; Vlad, S.; Lisa, G.; Cristea, M.; Zaltariov, M. F. Structure-Property Relationship of Sodium Deoxycholate based Poly(ester ether)urethane Ionomers for Biomedical Applications. J. Appl. Polym. Sci. 2016, 133, 42921. (24) Polizos, G.; Kyritsis, A.; Pissis, P.; Shilov, V. V.; Shevchenko, V. V. Structure and Molecular Mobility Studies in Novel Polyurethane Ionomers based on Poly(ethylene oxide). Solid State Ionics 2000, 136137, 1139−1146. (25) Xu, H.-S.; Yang, C.-Z. The Ionic Conductive Property of Sulfonated Polyethylene Oxide Polyurethane Ionomers. J. Polym. Sci., Part B: Polym. Phys. 1995, 33, 745−751. (26) Fragiadakis, D.; Dou, S.; Colby, R. H.; Runt, J. Molecular Mobility, Ion Mobility, and Mobile Ion Concentration in Poly(ethylene oxide)-based Polyurethane Ionomers. Macromolecules 2008, 41, 5723−5728.

mechanical strength through their hard phase and still have reasonable ion conduction in the soft phase. Figure 10b demonstrates the trade-off between ionic conductivity and storage modulus. Ionic conductivity is higher at lower hard segment content (i.e., PU10) and higher temperature, but the modulus is too low without microphase separation. With increasing hard segment content, the modulus can be significantly improved by microphase separation but ionic conductivity is lower than the single-phase PU10. Here, the fact that our materials stay in the solid state even at 150 °C gives considerable room for further improvement and optimization. More optimization can help improve both mechanical strength and ionic conductivity. The soft phase Tg is only slightly lower than room temperature and might be lowered by using longer PEO chains and/or a blend of PEO/ SC-diol as the soft segment to further boost the ionic conductivity. At the same time, using an amine-based chain extender may help improve microphase separation, which would improve the mechanical strength and possibly lower the soft phase Tg simultaneously.

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.macromol.7b02510. 1 H NMR of samples in DMSO-d6 (PDF)

■

AUTHOR INFORMATION

Corresponding Author

*E-mail [email protected] (R.H.C). ORCID

Ralph H. Colby: 0000-0002-5492-6189 Present Address

S.-W.W.: Axalta Coating Systems, 200 Powder Mill Rd., Wilmington, DE 19308. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was supported by the U.S. Army Research Office under Grant W911NF-07-0452 Ionic Liquids in Electroactive Devices (ILEAD) MURI. The authors thanks Timothy Long and James Runt for many useful discussions.

■

REFERENCES

(1) Fenton, D. E.; Parker, J. M.; Wright, P. V. Complexes of Alkali Metal Ions with Poly(ethylene oxide). Polymer 1973, 14, 589. (2) Meyer, W. H. Polymer Electrolytes for Lithium-Ion Batteries. Adv. Mater. 1998, 10 (6), 439−448. (3) Song, J. Y.; Wang, Y. Y.; Wan, C. C. Review of Gel-type Polymer Electrolytes for Lithium-Ion Batteries. J. Power Sources 1999, 77, 183− 197. (4) Mogurampelly, S.; Borodin, O.; Ganesan, V. Computer Simulations of Ion Transport in Polymer Electrolyte Membranes. Annu. Rev. Chem. Biomol. Eng. 2016, 7, 349−371. (5) Snyder, J. F.; Carter, R. H.; Wetzel, E. D. Electrochemical and Mechanical Behavior in Mechanically Robust Solid Polymer Electrolytes for Use in Multifunctional Structural Batteries. Chem. Mater. 2007, 19, 3793−3801. (6) Stephan, A. M. Review on Gel Polymer Electrolytes for Lithium Batteries. Eur. Polym. J. 2006, 42, 21−42. H

DOI: 10.1021/acs.macromol.7b02510 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules (27) Wang, S.-W.; Liu, W.; Colby, R. H. Counterion Dynamics in Polyurethane-Carboxylate Ionomers with Ionic Liquid Counterions. Chem. Mater. 2011, 23, 1862−1873. (28) Wang, S.-W.; Colby, R. H. Linear Viscoelasticity and Cation Conduction in Polyurethane Sulfonate Ionomers with Ions in the Soft Segment−Single Phase System. Macromolecules 2018, 51, in press, DOI: 10.1021/acs.macromol.7b02509. (29) Sheth, J. P.; Klinedinst, D. B.; Wilkes, G. L.; Yilgor, I.; Yilgor, E. Role of Chain Symmetry and Hydrogen Bonding in Segmented Copolymers with Monodisperse Hard Segments. Polymer 2005, 46, 7317−7322. (30) Fragiadakis, D.; Dou, S.; Colby, R. H.; Runt, J. Molecular Mobility and Li+ Conduction in Polyester Copolymer Ionomers based on Poly(ethylene oxide). J. Chem. Phys. 2009, 130, 064907. (31) Wang, W.; Liu, W.; Tudryn, G. J.; Colby, R. H.; Winey, K. I. Multi-Length Scale Morphology of Poly(ethylene oxide)-based Sulfonate Ionomers with Alkali Cations at Room Temperature. Macromolecules 2010, 43, 4223−4229. (32) McLennaghan, A. W.; Pethrick, R. A. Linear Segmented Polyurethane Electrolytes - I. Morphology and Physical Properties. Eur. Polym. J. 1988, 24 (11), 1063−1071. (33) Lee, D.-C.; Register, R. A.; Yang, C.-Z.; Cooper, S. L. Methylenebis(p-phenyl isocyanate)-based Polyurethane Ionomers. 2. Structure-Property Relationships. Macromolecules 1988, 21, 1005− 1008. (34) Koberstein, J. T.; Leung, L. M. Compression-Molded Polyurethane Block Copolymers. 2. Evaluation of Microphase Compositions. Macromolecules 1992, 25, 6205−6213. (35) Koberstein, J. T.; Galambos, A. F.; Leung, L. M. CompressionMolded Polyurethane Block Copolymers 0.1. Microdomain Morphology and Thermomechanical Properties. Macromolecules 1992, 25 (23), 6195−6204. (36) Choi, T.; Weksler, J.; Padsalgikar, A.; Runt, J. Influence of Soft Segment Composition on Phase-Separated Microstructure of Polydimethylsiloxane-based Segmented Polyurethane Copolymers. Polymer 2009, 50, 2320−2327. (37) Bonart, R.; Muller, E. H. Phase Separation in Urethane Elastomers as Judged by Low-Angle X-Ray Scattering. I. Fundamentals. J. Macromol. Sci., Part B: Phys. 1974, 10 (1), 177−189. (38) Lam, P. K. H.; George, M. H.; Barrie, J. A. Sulfonated Polyurethane Ionomers with New Ionic Diols. Polymer 1989, 30 (12), 2320−2323. (39) Macdonald, J. R. Theory of AC Space-Charge Polarization Effects in Photoconductors, Semiconductors, and Electrolytes. Phys. Rev. 1953, 92 (1), 4−17. (40) Coelho, R. On the Static Permittivity of Dipolar and Conductive Media - An Educational Approach. J. Non-Cryst. Solids 1991, 131−133, 1136−1139. (41) Klein, R. J.; Zhang, S.; Dou, S.; Jones, B. H.; Colby, R. H.; Runt, J. Modeling Electrode Polarization in Dielectric Spectroscopy: Ion Mobility and Mobile Ion Concentration of Single-Ion Polymer Electrolytes. J. Chem. Phys. 2006, 124, 144903. (42) Kremer, F.; Schonhals, A. Broadband Dielectric Spectroscopy; Springer: 2003. (43) Pissis, P.; Polizos, G. Molecular Dynamics and Ionic Conductivity Studies in Polyurethane Thermoplastic Elastomers. In Handbook of Condensation Thermoplastic Elastomers; Fakirov, S., Ed.; Wiley-VCH: 2005; pp 381−429. (44) Wubbenhorst, M.; van Turnhout, J. Analysis of Complex Dielectric Spectra. I. One-Dimensional Derivative Techniques and Three-Dimensional Modelling. J. Non-Cryst. Solids 2002, 305 (1−3), 40−49. (45) North, A. M.; Reid, J. C.; Shortall, J. B. Some Physical Properties Associated with the Urethane Group - II. Dielectric Relaxation in Thermoplastic Polyurethane Elastomers. Eur. Polym. J. 1969, 5, 565− 573. (46) Pissis, P.; Kanapitsas, A.; Savelyev, Y. V.; Akhranovich, E. R.; Privalko, E. G.; Privalko, V. P. Influence of Chain Extenders and Chain

End Groups on Properties of Segmented Polyurethanes. II. Dielectric Study. Polymer 1998, 39 (15), 3431−3435. (47) Georgoussis, G.; Kyritsis, A.; Pissis, P.; Savelyev, Y. V.; Akhranovich, E. R.; Privalko, E. G.; Privalko, V. P. Dielectric Studies of Molecular Mobility and Microphase Separation Polyurethanes. Eur. Polym. J. 1999, 35, 2007−2017. (48) Kanda, D. H. F.; Nagashima, H. N.; Malmonge, J. A.; Sakamoto, W. K.; Chierice, G. O. Effect of Polyol Excess on the Electrical Property of Vegetable-Polyurethane Film. J. Mater. Sci. 2008, 43, 5436−5440. (49) Namikawa, H. Multichannel Conduction in Alkali Silicate Glasses. J. Non-Cryst. Solids 1974, 14 (1), 88−100. (50) Namikawa, H. Characterization of the Diffusion Process in Oxide Glasses based on the Correlation between Electric Conduction and Dielectric Relaxation. J. Non-Cryst. Solids 1975, 18 (2), 173−195. (51) Dyre, J. C. Universal Low-Temperature AC Conductivity of Macroscopically Disordered Nonmetals. Phys. Rev. B: Condens. Matter Mater. Phys. 1993, 48, 12511−12526. (52) Macdonald, J. R. Addendum to “Fundamental Questions Relating to Ion Conduction in Disordered Solids″. J. Appl. Phys. 2010, 107, 101101. (53) Velankar, S.; Cooper, S. L. Microphase Separation and Rheological Properties of Polyurethane Melts. 3. Effect of Block Incompatibility on the Viscoelastic Properties. Macromolecules 2000, 33, 395−403. (54) Han, C. D. Linear Dynamic Viscoelasticity of TPUs. In Rheology and Processing of Polymeric Materials; Oxford University Press: New York, 2007; Vol. 1, pp 484−486. (55) Jangu, C.; Savage, A. M.; Zhang, Z.; Schultz, A. R.; Madsen, L. A.; Beyer, F. L.; Long, T. E. Sulfonimide-containing Triblock Copolymers for Improved Conductivity and Mechanical Performance. Macromolecules 2015, 48, 4520−4528.

I

DOI: 10.1021/acs.macromol.7b02510 Macromolecules XXXX, XXX, XXX−XXX