Investigating the Influence of Morphology in the ... - ACS Publications

Article pubs.acs.org/JPCB

Investigating the Influence of Morphology in the Dynamical Behavior of Semicrystalline Triton X‑100: Insights in the Detection/ Nondetection of the α′-Process Esther G. Merino,† Florence Danéde,‡ Patrick Derrollez,‡ Carlos J. Dias,§ M. Teresa Viciosa,∥ Natália T. Correia,†,‡ and Madalena Dionísio*,† †

REQUIMTE/CQFB, Departamento de Química, Faculdade de Ciências e Tecnologia, Universidade Nova de Lisboa, 2829-516 Caparica, Portugal ‡ Unité Matériaux Et Transformations (UMET), UMR CNRS 8207, UFR de Physique, BAT P5, Université Lille Nord de France, F-59655 Villeneuve d’Ascq Cedex, France § CENIMAT/I3N, Departamento de Ciência dos Materiais, Faculdade de Ciências e Tecnologia, FCT, Universidade Nova de Lisboa, 2829-516 Caparica, Portugal ∥ CQFMCentro de Química-Física Molecular and INInstitute of Nanoscience and Nanotechnology, Instituto Superior Técnico, Universidade Técnica de Lisboa, Avenida Rovisco Pais, 1049-001 Lisboa, Portugal ABSTRACT: The paper investigates the influence of the crystalline structure in the dynamical behavior of semicrystalline Triton X-100 allowing enlightening the reason for the detection/nondetection of the α′-process. The work was preceded by the study of the full amorphous material for which dielectric relaxation spectroscopy (DRS) identified multiple relaxations: the α-process associated with the dynamical glass transition and two secondary relaxations (β- and γ- processes). To evaluate how crystallinity affects the detected relaxation processes, different crystallizations were induced under high and low undercooling conditions. While the secondary relaxations are unaffected by crystallization, the mobility of the cooperative bulk α-process is sensitive to the distinct morphologies. The distinct semicrystalline states were structurally characterized by X-ray diffraction and polarized optical microscopy (POM). Differential scanning calorimetry (DSC) was used as a complementary tool. Depending on the extension of undercooling, large and well-defined shperulites or grainy-like structure emerge, respectively, for low and high undercooling degrees, as monitored by POM. In the two crystalline structures, X-ray diffraction patterns detected the amorphous halo meaning that both are semicrystalline. However, no differences between the amorphous regions are indentified by this technique; the distinction was done by means of dielectric measurements probing different mobilities in each of those regions. When the large spherulites evolve, the bulk-like α-process never goes to extinction and slightly shifts to low frequencies increasing the associated glass transition by 2−3 K, as confirmed by DSC; the slight change is an indication that the dimensions of the persisting amorphous regions become comparable to the length scale inherent to the cooperative motion that determines the glass transition in the full amorphous material. For the grainy-like structure, the αprocess becomes extinct and an α′-process evolves as revealed by isochronal plots of dielectric measurements, with the features of a glass transition as confirmed by temperature modulated differential scanning calorimetry; both techniques indicate a 10−12 K displacement of the associated hindered glass transition toward higher temperatures relative to the amorphous glass transition. It is concluded that the detection of the α′-process in Triton X-100 is greatly determined by the high degree of constraining of the amorphous regions imposed by the grainy crystalline structure disabling the occurrence of a bulk-like α-process. Triton X-100 can be taken as a model for understanding low molecular weight materials crystallization, allowing correlating the observed dynamical behavior with the achieved crystalline morphology.

1. INTRODUCTION Recently, the temperature driven phase transformations of the water-soluble liquid surfactant Triton X-100, with the molecular formula C14H22O(C2H4O)n (n = 9−10), were investigated by some of us,1 mainly by using dielectric relaxation spectroscopy (DRS); differential scanning calorimetry (DSC) and polarized optical microscopy were used as complementary tools. © 2013 American Chemical Society

Due to its high dielectric response and ability to crystallize from both molten and glassy states (melt-crystallization at ∼255 K at 1 K·min−1; cold-crystallization at ∼232 K), Triton Received: April 29, 2013 Revised: July 18, 2013 Published: August 12, 2013 9793

dx.doi.org/10.1021/jp4042414 | J. Phys. Chem. B 2013, 117, 9793−9805

The Journal of Physical Chemistry B

Article

emerge when crystallization occurs closer to Tm (≈273 K in Triton X-100) at 24 K ≤ Tm, Tcr ≤ 44 K.1 The evolving of a new relaxation process with a more hindered mobility that α-relaxation associated with the dynamic glass transition, designated α′-process, is found in crystallizable systems either polymeric,4,18−22 or low molecular weight materials.14,15,23 For Triton X-100 isothermally crystallized near Tg, a low frequency process was detected1 exhibiting features common to this α′-process; this will be investigated here in more detail with the help of TMDSC. The present paper is a further contribution to the understanding of low molecular weight materials crystallization, trying to correlate the different behaviors observed for the decay of the α-relaxation and the emergence of a α′-process with a particular morphology of the crystalline fraction.

X-100 revealed to be a good material to study simultaneously the crystallization process and the mobility of the coexistent amorphous fraction. Moreover, it was found that crystallization can be avoided and a fully amorphous material is obtained, if it is cooled from the isotropic liquid at a rate ≥10 K·min−1. It exhibits a glass transition temperature around 212 K taken at the midpoint of the transition region measured on heating by differential scanning calorimetry (DSC), allowing to classify it as a glass former; however, the mobility of completely amorphous Triton X-100 was not wholly explored at that time, being now characterized in more detail. Additionally, the mobility of the material undergoing nonisothermal meltcrystallization and after isothermal cold-crystallization is investigated. Complementary techniques are used, dielectric relaxation spectroscopy (DRS) and X-ray diffraction (XRD), probing different media; DRS probes the bulk/constrained amorphous fraction while XRD directly probes the emergent crystalline fraction; additionally, both standard and temperature modulated DSC (TMDSC) are applied for the thermal characterization of Triton X-100 under different conditions. The evolving of crystallization can change the molecular dynamics of the coexisting noncrystalline material as long as the crystalline lamellae cause geometrical confinement on the remaining amorphous phase. For instance, in polymers when the crystalline phase confines the amorphous regions to dimensions smaller than a few nanometers, strong variations from the bulk dynamical behavior can be observed leading to either slower or enhanced dynamics as observed in semicrystalline poly-L-lactic acid (PLLA).2 The monitoring of crystallization can be carried out by dielectric relaxation spectroscopy, for example, by the reduction of the dielectric strength of the αrelaxation, as has been done in either polymers such as poly(ethylene terephthalate) (PET),3,4 PLLA,5−8 and polycarbonate/poly(ε-caprolactone) blends9 or low molecular weight materials such as isooctyloxycyanobiphenyl,10 ethyleneglycol dimethacrylate,11 and also in Triton X-100.1 In these low molecular weight materials, it was found that the α-process depletes upon isothermal crystallization with no significant changes either in position or in shape as observed also in some pharmaceutical drugs.12,13 However, this behavior is not common to all materials: in sorbitol14 or in a terephthalic acid dipropyl ester,15 the peak position shifts toward lower frequencies, while in 2-propanol a shift to either higher or lower frequencies is observed depending on the crystallinity degree (for crystallinities lower than 80% the shift is to higher frequencies and the opposite occurs for higher crystallinities).16,17 As above-mentioned, for Triton X-100 the α-peak extinguishes while keeping the same location upon isothermal crystallization, either cold or melt, carried out in the vicinity of the calorimetric glass transition temperature, Tg (6 to 8 K above Tg),1 with a concomitant evolving of a new process. The morphology of the crystalline fraction obtained under these conditions revealed to be a grain-like microstructure with no optically resolvable spherulites as found by polarized optical microscopy (POM). This was interpreted as the concurrent effect of a high dynamic fragility involving high cooperativity between relaxing units and the occurrence of a nucleation/ crystal grow catastrophic process. However, it was found that the morphology depends on the undercooling degree (temperature difference between the melting (Tm) and the occurrence of crystallization (Tcr)): large and well-defined spherulites

2. EXPERIMENTAL SECTION 2.1. Materials. Triton X-100, polyethylene glycol tertoctylphenyl ether, C14H22O(C2H4O)n with an average number, n ∼ 9−10 of oxyethylene units per molecule (MW ∼625) (see Scheme 1) was reagent grade purchased from Fluka (catalogue no. 93420; CAS no.: 9002-93-1). Hereafter, it will be designated as Triton. Scheme 1. Chemical Structure of Triton X-100; n ∼ 9−10

Karl Fisher analysis showed water content of 0.28% (w/w). It was used without further purification. For the dielectric and calorimetric analysis, Triton was studied after being submitted to different thermal treatments, which allowed obtaining samples in different physical states. These are summarized in Tables 1 and 2. 2.2. Experimental Techniques. 2.2.1. Dielectric Relaxation Spectroscopy (DRS). Dielectric measurements were carried out using the ALPHA-N impedance analyzer from Novocontrol Technologies GmbH. A drop of Triton as Table 1. Thermal Treatment and Subsequent Physical State of the Samples Prior to the DRS Experiments; Designation Used in the Paper thermal treatment I. Cooling from the melt at 11 K·min−1 from 373 to 153 K II. (i) Cooling from the melt at 11 K·min−1 from 373 to 153 K; (ii) heating to 219 at 9 K·min−1; (iii) isothermal crystallization 2 h at 219a K; (iv) cooling to 153 at 11 K· min−1 III. Cooling from the melt at 1 K·min−1b from 373 to 153 Kc

physical state obtained

designation

fully solid amorphous material at 153 K semicrystalline

amorphous

semicrystalline

SC nonisothermal melt

SC isothermal cold

a The isothermal crystallization was carried out near Tg at 219, 220, and 221 K leading to very similar spectra after crystallization; sample crystallized at 219 K was taken for further analysis. bThe cooling rate is an average value estimated from the time taken to isothermal spectral acquisition. cThe nonisothermal crystallization occurred near Tm.

9794



Article

Table 2. Thermal Treatment and Subsequent Physical State of the Samples Prior to the DSC Experiments; Designation Used in the Paper thermal treatment −1

I. Cooling at 20 K·min

physical state obtained

from 303 to 173 K

fully solid amorphous material at 173 K semicrystalline

II. Cooling from the melt at 1 K·min−1 from 303 to 173 K

a

III. (i) Cooling from the melt at 20 K·min−1 from 303 to 223 K; (ii) isothermal annealing/crystallization 2 h at 223 Kb; (iii) cooling to 173 at 20 K·min−1 IV. (i) Cooling from the melt at 20 K·min−1 from 303 to 173 K; (ii) heating to 223 at 20 K·min−1; (iii) isothermal crystallization 2 h at 223 Kb; (iv) cooling to 173 at 20 K·min−1 a

semicrystalline semicrystalline

designation amorphous SC nonisothermal melt SC isothermal melt SC isothermal cold

The nonisothermal melt-crystallization occurred near Tm. bThe isothermal crystallization was carried out near Tg.

From the estimated values of τHN, αHN, and βHN parameters, a model-independent relaxation time, τ = 1/(2πf max), was determined (see references 24−26 for details). Close to the glass transition temperature, the temperature dependence can be described by the well-known Vogel/ Fulcher/Tammann/Hesse equation,27−29 which reads,

received was placed between two gold plated electrodes (diameter of inferior electrode = 10 mm) of a parallel plate capacitor, BDS 1200 with two silica spacers, 50 μm thickness. The sample cell was mounted on a cryostat, BDS 1100, and exposed to a heated gas stream being evaporated from a liquid nitrogen Dewar. The temperature control is assured by the Quatro Cryosystem and performed within ±0.5 K. Novocontrol Technologies GmbH supplied all these modules. Before isothermal dielectric measurements the sample was previously heated to 373 K, under the nitrogen stream, to eliminate residual water. The full amorphous material (Amorphous) and two semicrystalline samples (SC isothermal cold and SC nonisothermal melt) (see Table 1) were studied by DRS. Since in previous studies no differences were observed between samples isothermally crystallized coming from either melt or cold,1 only DRS measurements of sample cold-crystallized will be taken for analysis. For the Amorphous and SC isothermal cold samples, the spectral acquisition was carried out in increasing different temperature steps: in the range 153 K ≤ T ≤ 193 and 273 K ≤ T ≤ 373 K each spectrum was recorded every 5 K; in the remaining temperature region the spectra were recorded every 2 K. For sample SC nonisothermal melt, the spectral acquisition was done upon cooling in the following temperature steps: in the range 373 K ≥ T ≥ 273 K, each spectrum was recorded in steps of 10 K, and in the range 273 K > T ≥ 193 K, the spectra were recorded every 2 K; from 193 K > T ≥ 153 K, the spectral collection was done in steps of 5 K; nonisothermal melt crystallization occurred during measurements in the range 273 K > T ≥ 193 K. Dielectric Data Analysis. To analyze the isothermal dielectric data, the model function introduced by Havriliak− Negami (HN) was fitted to both imaginary and real components of complex permittivity using the WinFit software (version 3.4, Novocontrol Technologies, 2011). Because multiple peaks are observed in the available frequency window, a sum of HN-functions was employed: ε*(f ) = ε∞ +

∑ j

⎛ B ⎞ τ(T ) = τ∞exp⎜ ⎟ ⎝ T − T0 ⎠

τ∞ and B are constants and T0 is the so-called Vogel temperature. 2.2.2. Differential Scanning Calorimetry (DSC). DSC experiments were carried out in a DSC Q10 from TA Instruments interfaced with a liquid nitrogen cooling accessory (LNCA). The DSC runs cover a temperature range from 173 to 303 K with heating rate of 5 K·min−1. A small amount of sample (∼5 mg) was placed in an aluminum hermetic pan. Measurements were performed under dry high-purity helium gas (at flow rate of 25 mL·min−1) to improve the thermal conductivity. The baseline was calibrated scanning the temperature range of the experiments with two empty pans. Calibration was carried out using indium for temperature transitions and the heats of fusion. Analogous to what was done by DRS, samples prior submitted to different thermal treatments, reproducing close the DRS conditions, were calorimetrically investigated. A cooling rate of 20 K·min−1 was used (instead of 11 K·min−1 as in DRS). Table 2 summarizes the thermal treatments/ physical states/designations of the samples tested by DSC. The thermograms that will be later analyzed were obtained upon a final heating scan from 173 K, after each thermal treatment from I to IV (Table 2). Temperature modulated DSC (TMDSC) was also performed using a Q200 from TA Instruments (MDSC technology) equipped with a RCS cooling system. The TMDSC experiments were carried out in the “heat-only” modulation mode (oscillation period 60 s, amplitude 0.08 K, heating rate 0.5 K·min−1), from 183 to 303 K, for the SC isothermal cold sample (after thermal treatment IV in Table 2). 2.2.3. X-ray Diffraction. X-ray diffraction (XRD) analysis was carried out with a diffractometer using Cu Kα radiation (selected wavelength λCu Kα1 = 1.54056 Å) in the 2θ range 0.2− 113.8° (2θ step 0.015°). The X-ray diffraction patterns were recorded in real time with an INEL curved position-sensitive detector (CPS 120) giving simultaneous detection over 120°. The Triton samples were enclosed in a Lindemann glass capillary (diameter 0.7 mm) mounted on the axis of a G3000 goniometer (see ref 30 for more details). A Cryostream Plus

Δεj βHN

[1 + (iωτHNj)αHNj ]

j

(2)

(1)

where j is the index over which the relaxation processes are summed, Δε is the dielectric strength, τHN is the characteristic HN relaxation time, and αHN and βHN are fractional parameters (0 < αHN < 1 and 0 < αHNβHN < 1) describing, respectively, the symmetric and asymmetric broadening of the complex dielectric function.24 9795



Article

controller from Oxford Cryosystems was used to regulate the temperature. 2.2.4. Polarized Optical Microscopy. Polarized optical microscopy (POM) was performed on an Olympus Bx51 optical microscope equipped with a Linkam LTS360 liquid nitrogen-cooled cryostage. Photomicrographs were taking using an Olympus C5060 wide zoom camera. Images were obtained at a magnification of 500×; for more details see ref 1.

3. RESULTS AND DISCUSSION 3.1. Morphological and Structural Characterization. The temperature history specified in Table 2 to be followed in the calorimetric analysis was applied to observe the emergence of crystallinity by X-ray diffraction and POM. The time evolution of crystallization induced nonisothermally by cooling at 1 K·min−1 from the molten state (SC nonisothermal melt), was monitored by POM. The respective microphotographs taken at different decreasing temperatures are shown in Figure 1a. It is observed that large and welldefined spherulites emerge at 266 K, near below Tm, and for temperatures lower than 232 K, no further changes occur. In Figure 1b, a grain-like microstructure with no optically resolvable spherulites is observed for a sample rapidly cooled from melt and isothermally crystallized near Tg at 223 K (SC isothermal melt). The microphotographs taken now by POM disclose identical morphologies as those previously reported,1 revealing high reproducibility in the crystalline state achieved by Triton when submitted to the same thermal treatment. Figure 2a presents the X-ray diffraction patterns collected at 223 K for three Triton samples: isothermally cold- and meltcrystallized near Tg (223 K) and nonisothermal melt-crystallized near below Tm (cooled at 1 K·min−1 from the liquid state); the designation included in Table 2 is applied. In the inset, the diffraction pattern for liquid Triton (taken at room temperature) is included as a reference. All these diffractions patterns were obtained for an acquisition time of 7 h. As a first observation, the XRD patterns of the SC isothermal cold and melt samples fall in a single plot. This agrees with what was observed previously by DRS1 where no differences were detected. Additionally, the two patterns act as a probe to evaluate the reproducibility of the X-ray trials. Second, two strong reflections centered at 2θ = 19.2° and 23.3° are observed in the diffraction patterns of these SC isothermal cold and melt samples providing a clear evidence of crystallinity. On the other hand, the inset in Figure 2a presents the full pattern over a broader 2θ range, which superimposes the halo for the liquid/noncrystalline sample; this result sustains that the SC isothermal cold and melt samples are semicrystalline. The diffraction pattern for the sample nonisothermally crystallized from the melt (SC nonisothermal melt) exhibits two Bragg peaks that emerged in the same 2θ region as for the samples isothermally crystallized near Tg; however, their full width at half-maximum is narrower, which is a feature compatible with higher crystal dimensions, in agreement with what is observed by POM (Figure 1a). Therefore, the small differences between the X-ray patterns of isothermally cold/ melt crystallized and nonisothermal crystallized are due to different crystal dimensions instead to different crystal forms. This was not obvious from POM, since really distinct morphologies were observed, leading to speculate if it were caused by polymorphism. X-ray diffraction analysis pulls apart

Figure 1. Microphotographs taken by POM: (a) at the specified temperatures on cooling from the liquid state at 1 K·min−1 (sample SC non isothermal melt) and (b) at 223 K for the SC isothermal melt sample.

this hypothesis, at least for samples crystallized under these specific conditions. In Figure 2b, the X-ray pattern of a sample quenched from room temperature to 173 K is presented. The similarity with the X-ray spectrum for the liquid shown in Figure 2a confirms that a full amorphous material is obtained exhibiting the halo characteristic of the amorphous form with no observable Bragg peaks. Additionally, the time evolution of this initially full amorphous sample kept at 223 K is presented (corresponding to the SC isothermal cold); in this case, each spectrum was obtained with an acquisition time of 30 min. The first spectrum (tc = 30 min) already shows the crystalline pattern, and no significant evolutions are observed for diffractograms taken after ∼2 h. Additionally, the XRD patterns over all 2θ range (5−115°) were deconvoluted by the Voigt function of the 9796



Article

degree from which it is concluded that the maximum crystallization (37.6%) is attained after the first 100 min of isothermal annealing. This will be commented further on. 3.2. Dynamical Characterization. Amorphous Triton. In a previous work,1 it was shown that crystallization can be circumvented for Triton when cooled down from the molten state at a rate higher than 10 K·min−1. Therefore, to evaluate the mobility in the glass and supercooled liquid state, the sample, after heated up to 373.15 K for removal of residual water, was cooled down at 11 K·min−1, and the dielectric loss spectra were taken in increasing temperatures steps from 153 K up to 273 K (see the Experimental Section). Representative ε″ spectra thus collected are shown in Figure 4. The spectra of Amorphous Triton evidence a multimodal character with several relaxation processes. At the lowest frequencies and highest temperatures, a strong relaxation process is observed, which will be attributed later on to the

Figure 2. (a) X-ray diffraction patterns for semicrystalline Triton samples (same designation as used in DRS and DSC) recorded at 223 K; the inset shows the XRD spectra over a larger 2θ range where the diffraction pattern for liquid Triton taken at room temperature is also included (black line). (b) XRD diffraction pattern of Triton obtained after an initial quenching to 173 K (black line) corresponding to a full amorphous form together with the time evolution (see legend inside) of the Bragg peaks during cold-crystallization at 223 K.

PeakFIT software. The degree of crystallinity was estimated by the ratio of the area of the Bragg peaks to the total area. Figure 3 shows the time evolution of the thus estimated crystalline

Figure 4. (a) Isothermal dielectric loss spectra (logarithmic scale) for Amorphous Triton collected between 153 and 217 K, after a cooling ramp from 373 to 153 K carried out at 11 K·min−1 (this rate assures that melt-crystallization is avoided). The individual relaxation processes considered in the HN fit of the raw data are depicted in the figure for the spectrum obtained at 201 K: two secondary processes, β, γ, and the α-process. (b) Isothermal dielectric loss spectra illustrating the decrease in the dielectric response from 221 to 223 K due to crystallization; the inset presents the normalized dielectric loss curves for the α-relaxation illustrating its invariant shape. The high frequency side is affected by the secondary processes. The solid line represents the fit of the one-sided Fourier transform of the KWW function with βKWW = 0.39.

Figure 3. Time evolution of the degree of crystallinity estimated from the deconvolution of the X-ray diffraction patterns for the SC isothermal cold which difractograms are depicted in Figure 2b. The solid line is only a guide for the eyes. The dispersion of the points at long times provides an estimation of the standard uncertainties. 9797



Article

conversion to τ (see the Experimental Section). The relaxation processes were designated as usual in an increasing order of frequency at which they are detected for a fixed T, as α, β, and γ. The temperature dependence of the relaxation times of the γprocess is Arrhenian; that is, it obeys ln τ = ln τ∞ + Ea/RT where τ∞ is the pre-exponential factor and Ea is the activation energy; R is the ideal gas constant, and T the temperature. This behavior is characteristic of a local relaxation process confirming that it is thermally activated. Although the βprocess is less defined, a linear plot is also obtained; the usual curvature is obtained for the α-process denoting the cooperative nature of the underlying relaxation mechanism. The activation parameters are summarized in Table 3. Interesting, both the pre-exponential factor (τ∞) and Ea value obtained for the γ-process of Triton are comparable with the respective parameters obtained for the γ-process observed in the low members of the n-ethylene glycol dimethacrylate (nEGDMAs) series (1 ≤ n ≤ 4), respectively (1.2 ± 0.9) × 10−16 s and 39 ± 3 kJ·mol−1.23 This relaxation in those materials was assigned to dipolar fluctuations within the ethylene glycol moiety,31 a molecular group that also exists in the molecular chemical structure of Triton (see Scheme 1), and therefore, the previous attribution seems to aplly to the process now detected as well. Concerning the mechanism of the β-process, it must be associated with a higher length scale relative to the γ-process. It can be conceived that it can be associated with hindered rotations of the octylphenyl ether group (see Scheme 1). The bulkiness of this unit is coherent with the higher value of activation energy. Additionally, its molecular origin can be rationalized as involving some cooperativity as denounced by a prefactor of ∼10−23 s, far from the one expected for noncooperative local processes (10−12 to 10−14 s).32 The solid line in Figure 5 describing the α-process is the fit obtained from the VFTH law to the τ(T) data (eq 2). From the extrapolation of the VFTH equation to τ = 100 s,33,34 a glass transition temperature of 208.3 K is obtained in good accordance with the onset of the calorimetric value;1 the respective VFTH parameters are given in Table 3/Amorphous. In our previous paper,1 the fragility index, which is a quantitative measure of the degree of deviation from Arrhenius-type temperature dependence near Tg, allowing to classify the glass formers in terms of fragility35,36 was estimated for amorphous Triton based on fewer τ(T) data points. It is

dynamic glass transition; therefore, it is designated hereafter as α-process. This process shifts to higher frequencies with the temperature increase due to the mobility enhancement enabled by the temperature increasing. At the lowest temperatures, a less strong but well-defined process is found that also goes toward high frequencies with T increasing, although in a lesser extent. In between both relaxations, another process is felt but never emerges as a definite one. At higher temperatures, crystallization occurs between 221 and 223 K, as confirmed by the abrupt decrease in the spectrum taken at the latter temperature (seen in the isotherms shown in Figure 4b). To evaluate the temperature dependence of the relaxation times of each detected process in the amorphous state before crystallization, the HN equation (eq 1) was fitted to each spectrum allowing drawing a relaxation map (Figure 5), after

Figure 5. Relaxation map of all detected processes for the Amorphous sample (filled circles); the solid lines are the Arrhenius and VFTH fits for, respectively, the secondary and dynamic glass transition processes. The map also includes the relaxation times obtained for semicrystalline Triton: SC isothermal cold open trianglesτ(T) values for the secondary βsc and γsc processes detected after isothermal coldcrystallization at 219 K (extinction of the α-process); SC nonisothermal melt gray starsτ(T) values obtained during melt-crystallization for which αsc-process is observed; the arrow indicates the temperature (232 K) from which no more changes are observed by POM in the crystalline morphology; the dashed curved line is the VFTH fit for the MWS process for SC nonisothermal melt.

Table 3. Summary of the Activation Parameters for All Detected Processes in Triton at the Different Conditions Investigated in the Present Work semicrystalline Triton relaxation process α

β γ MWS

activation parameters

amorphous

τ∞/s B/K T0/K Tg (τ = 100 s)/K m Ea/kJ·mol τ∞/s Ea/kJ·mol−1 τ∞/s τ∞/s B/K T0/K

(1.0 ± 0.2) × 10−14 1080 ± 11 178.9 ± 0.4 208.3 ± 0.9 113 ± 10 80 ± 3 (5 ± 4) × 10−23 36.3 ± 0.1 (7.5 ± 0.4) × 10−16

75 ± 5 (1.6 ± 1.5) × 10−21 35.6 ± 0.2 (9 ± 1) × 10−16

not observed

not analyzed

9798

SC isothermal cold not observed

SC nonisothermal melt (1.2 ± 0.4) 722 ± 14 189 ± 1 211 ± 1 130 ± 14 76 ± 4 (1.3 ± 1.2) 37.4 ± 0.2 (2.2 ± 0.3) (1.1 ± 0.6) 699 ± 72 184 ± 2

× 10−12

× 10−21 × 10−16 × 10−8



Article

possible to recalculate it now for amorphous Triton as m = 113. Alternatively, m can be estimated by the correlation proposed by Böhmer et al.,33 m = 250(±30)-320(βKWW), where βKWW is the so-called “stretching parameter” of Kohlrausch−Williams− Watts (KWW) function37,38 ϕ(t) = exp[−(t/τKWW)βKWW] that describes the relaxation response in the time domain; 0 < βKWW < 1; βKWW = 1 for a Debye response. Using the βKWW value found from the fit of the one-sided Fourier transform of the KWW function to the normalization plot, 0.39 (see inset of Figure 4b), the proposed correlation gives a value of m = 125 ± 30, which includes the value estimated from the VFTH parameters. The fragility index of amorphous Triton is a rather high value for a relatively low molecular weight material,33,39 and it was discussed previously as one of the factors determining its ability to crystallize close to the glass transition.1 Semicrystalline Triton. In a previous work,1 the isothermal crystallization of Triton was dielectrically monitored at 219, 220, and 221 K, until the complete extinction of the α-process. Now, the dielectric spectra collected at the end of each isothermal cold-crystallization (see the Experimental Section), are presented and analyzed to evaluate the mobility of the remaining amorphous fraction of these samples, here designated as SC isothermal cold. A similar behavior was found for the three samples as evidenced by the superposition in the plots of both ε′(T) and ε″(T) traces at 104 Hz, shown respectively in Figure 6a and b; the corresponding plot of the Amorphous sample obtained during cooling (solid line) is included in Figure 6a for comparison, together with the subsequent heating run (Figure 6a and b) where nonisothermal cold crystallization occurred. Figure 6c presents the loss curves at some representative temperatures for the SC isothermal cold sample crystallized at 219 K. By comparing these spectra for the semicrystalline material with those taken for Amorphous Triton (Figure 4), the main differences are the above-mentioned suppression of the cooperative α-process and a better definition of the βrelaxation. The extinction of the α-process for the SC isothermal cold sample indicates a relatively high crystallinity of the sample. If the crystallinity degree is estimated from the reduction in the dielectric strength taking in account both main and secondary processes,1 a value of 0.87 is found, highly surpassing the one determined from the areas of the Bragg peaks obtained from XRD spectral deconvolution, 0.37 (see section 1.1). This could mean that the amorphous fraction that it is being considered in the calculation based on the dielectric data is underestimated, which can be taken as an indication of the existence of the α′process. Nevertheless, it is questionable if the same crystallinity degree is attained in the both experiments having different sample holders (glass capillary tube in XRD vs flat electrodes in DRS). Even so, it is interesting to note the parallelism between the time evolution of the crystallinity degree determined by XRD for the SC isothermal cold sample (remember Figure 3) and the time evolution of the crystallinity degree estimated by DRS;1 for the latter, no further crystallization occurs after ∼100 min, as observed by XRD. From the HN fit to the raw data, the relaxation times of the two secondary processes were obtained for the semicrystalline material and included in Figure 5 (open triangles). The Arrhenius plot for the relaxation times of the secondary γ-process detected in the SC isothermal cold sample (γsc) almost superimposes the one of the Amorphous sample, giving an activation energy of 36.3 ± 0.1 kJ·mol−1, close to the value

Figure 6. Temperature dependence of the (a) ε′ and (b) ε″ traces at 104 Hz (isochronal plots) for the three SC isothermal cold samples of Triton after being isothermally cold-crystallized at three different temperatures near Tg (see legend inside); as a solid line, the ε′(T) trace obtained upon cooling at 11 K·min−1 for the Amorphous sample is included for comparison in part a, and the subsequent heating run where the sample undergoes nonisothermal cold crystallization is plotted as well in both (a) and (b) (full black symbols); arrows in part b indicate the onset of αam- and α′-process and MWS process (see text). (c) Dielectric loss versus frequency for the SC isothermal cold after the sample being submitted to 2 h of isothermal coldcrystallization at 219 K; the crystalline degree was estimated as 0.87 (see text). The spectra shown here were collected between 153 and 203 K (presented in steps of 10 K); the figure also includes spectra collected at 209, 213, and 215 K; the full symbols indicate the frequency region where the β-process is observed. The bulk-like αprocess is not observed; the raise in ε″ is due to the incoming of the α′-process.

estimated for the full amorphous material (35.6 ± 0.2 kJ· mol−1). 9799



Article

loss peaks maxima. To have an evidence of this behavior, ε″(ω) will be compared; however, the dielectric loss is significantly affected by conductivity mainly in the semicrystalline material. To surpass this difficulty, it can be determined from ε′(ω), to which the dc conductivity does not contribute (as long as electrode polarization becomes negligible, which is the case). Therefore, ε″(ω) was calculated from the frequency dependence of the derivative of the real part of the complex dielectric constant, according to40−43

In spite of a slight decrease in its relaxation strenght, the secondary γsc-process keeps the same shape of the full amorphous process as seen by the coincidence of the normalized plot (not shown). The almost insensitivity of the γ-trace to crystallization, even observed under a high crystallinity, as also found for EGDMA,23 can be taken as a confirmation of a very localized process, intramolecular in nature, as attributed above. In regard to the βsc-relaxation, as already mentioned, it was possible to resolve this process in a wider temperature range for the semicrystalline material. The obtained activation energy, Ea(βsc) = 75 ± 5 kJ·mol−1, is similar within the experimental error to the value estimated for amorphous Triton; Ea(βam) = 80 ± 3 kJ·mol−1. The relaxation map in Figure 5 also includes the relaxation times obtained for the processes detected from dielectric measurements carried upon cooling from the liquid during which Triton undergoes nonisothermal melt-crystallization, that is, the relaxation times for the SC nonisothermal melt sample (isotherms shown in Figure 7).

″ ≈− εderiv

π ∂ε′(ω) 2 ∂ln ω

(3)

Equation 3 is an alternative to the numerical Kramers−Kronig relations to obtain ε″(ω) that allows, besides the elimination of the conductivity contribution, to improve the resolution of overlapping relaxation processes. The plot of the loss peaks thus obtained (labeled εderiv″(ω)) normalized by the maximum εderiv″ value, for both Amorphous and SC nonisothermal melt Triton are compared in Figure 8a. It is evident the displacement in the maximum of the peak for the semicrystalline material to lower frequencies confirming the difference in relaxation times shown in the activation plot; furthermore, the spectral shape of both peaks seems to be the

Figure 7. Dielectric spectra collected isothermally upon cooling from the melt at some representative temperatures (from 158 to 198 K, the spectra are shown in steps of 10 K; in the remaining temperature range, they are shown in steps of 4 K until 245 K (the spectrum taken at 247 K is also included); in full circles is depicted the spectrum taken at 221 K) for semicrystalline Triton: SC nonisothermal melt; several relaxation processes are observed including the αSC-process and a MWS process due to the coexistence of amorphous and crystalline fractions. In all the spectra here presented, Triton is already semicrystalline since the onset of crystallization according POM occurred at 266 K and no further crystallization occurs below 232 K.

All log(τ) vs 1000/T data values for the process labeled αsc (gray stars in Figure 5) concern to a semicrystalline material since the onset of crystallization occurred well above the evolving of the α-process; with the temperature decreasing, no further crystallization takes place. The POM observations support this, since from below 232 K the crystalline morphology does not undergo further changes. Nevertheless, the αsc-process in this semicrystalline material, SC nonisothermal melt, is highly depleted relative to the full amorphous material. From the decay in the dielectric strength of both αsc- and secondary peaks relative to full amorphous Triton, a crystallization degree of 0.69 is estimated. The τ values of this αsc-process are slightly shifted toward higher values relative to the Amorphous material indicating a slower mobility in the SC nonisothermal melt sample. The shift between relaxation times can also be seen by a deviation in the

Figure 8. Normalized loss peaks at 221 K obtained from the derivative of ε′( f) (eq 3) of Amorphous Triton (full circles) compared with semicrystalline Triton (stars): (a) loss peak collected after nonisothermal melt-crystallization (SC nonisothermal melt), evidencing the shift toward lower frequencies of the main relaxation process (αSC) relative to the amorphous material (αam); the lines are the individual HN functions and conductivity used in the fit of the normalized derivative plot for each derivative plot, the solid line over the experimental data points is the overall fit; (b) loss peak collected after 720 s under isothermal cold-crystallization1 at 221 K (SC isothermal cold sample) confirming the invariance in the position of the main process while another process emerges in the low frequency flank (it will be assigned later on text); lines are only a guide for the eyes. 9800



Article

melt crystallization. To ensure that the comparison between the different processes is done in the same way, the analysis taken the derivative of the real part of the complex dielectric constant was also carried out for the α-process monitored while Triton undergo isothermal crystallization at 221 K. This is shown in Figure 8b considering a spectrum for the sample exhibiting a similar crystallization degree as the one taken from the nonisothermal crystallization (presented in Figure 8a), that is, similar area under the ε″-peak. The comparison confirms the previous assumption that the relaxation time of the α-process can be kept constant.1 However, for an adequate description of the dominant relaxation taken at the earliest times (12 min), two processes must be considered. This establishes an important difference relative to the SC nonisothermal melt, where only a single function, αSC, was used to properly simulate the main process. The additional process to αSC needed to describe the spectrum for the SC cold-isothermal sample, was used to take in account the low frequency contribution, the origin of which was not unequivocally assigned in the previous paper;1 this will be next discussed. It should be stressed that the α-like process detected in the SC isothermal cold sample goes to extinction with the progress of crystallization. The same does not occur with the α-like process detected in the SC nonisothermal melt sample, remaining active upon further cooling. For this sample, MWS polarization is clearly detected in the covered temperature and frequency range exhibiting the features of a relaxation process. Having in mind the POM observations described earlier in the text, when crystallization occurs near the glass transition, a catastrophic nucleation/crystallization emerges, suppressing bulk amorphous regions, at least at the length scale of the microscopic observation, which, according DRS measurements, disables the bulk-like α-process. However, X-ray diffractograms clearly show that the amorphous halo is always present, confirming that the sample is semicrystalline (albeit some care is required in comparing crystallinity degrees, as aforementioned). This means that the remaining amorphous regions are highly constrained and a slower and more hindered αSC-process takes place, identical to the one occurring in rigid amorphous regions observed for polymers and also in low molecular weight materials (see the Introduction), although analyzed based in a different molecular origin as discussed in the end of this paper, designated as α′-process; the same terminology is here adopted. In Figure 6b, the arrows indicate in the ε″(T) trace the onset of both αam and α′-process being evident the displacement toward higher temperatures of the latter (∼10 K). Additionally, a MWS process is detected also for this sample, shifted ∼20K to higher temperatures relatively to the homologous process observed in the SC nonisothermal melt, as shown in Figure 6b through the ε″(T) trace at 104 Hz; it also becomes evident in the ε′(T) trace for lower frequencies as 10−1 Hz (not shown). Nevertheless, it never emerges as a defined peak being the reason why it respective relaxation times are lacking in the relaxation map presented in Figure 5. For samples crystallized near below Tm, POM evidenced that well-defined spherulites become to be observed at around 260 K, and bulk-like amorphous regions coexist with the crystalline phase being microscopically observable down to 232 K. Once more, the X-ray diffraction pattern exhibits the amorphous halo for the SC nonisothermal melt sample taken at 223 K confirming that the sample is semicrystalline. The dielectric analysis indicates that an α-process is detected in the remaining

same. This can be confirmed by analyzing the shape parameters used in the fit of the normalized derivative plots: the lines in Figure 8a are the individual HN functions describing the α, β, and γ processes used to fit the plot for the amorphous and semicrystalline curves; the solid line over the experimental data points is the overall fit. It should be pointed out that in the fit of each derivative plot, exactly the same shape parameters and relaxation times were used to simulate the β and γ processes: αHN,β =0.45; βHN,β = 1.00; τHN,β= 1.88 × 10−4 s; αHN,γ =0.63; βHN,γ = 0.86; τHN,γ = 7.70 × 10−7s. This confirms what was observed previously on analyzing the secondary relaxations detected in the SC isothermal cold sample. For the fit of the αlike process the difference in the used parameters αHN,Am = 0.66; βHN,Am = 0.64 and αHN,SC =0.64; βHN,SC = 0.64 is meaningless confirming the same shape of the relaxation process; only the respective relaxation times change: τHN,Am = 5.05 × 10−3 s; τHN,SC = 1.36 × 10−2 s. Therefore, the main relaxation detected for the SC nonisothermal melt is adequately described by a single process that keeps the same shape of the amorphous one. The shift toward lower frequencies can be taken as an indication that although the dimensions of the remaining amorphous regions are large enough to allow the manifestation of the α-process, it become to be of the order of the inherent length scale of the cooperative dynamics that determines the glass transition in full amorphous Triton. The low frequency tail in the spectrum of this semicrystalline material is due to the incoming of a MWS process due to the internal phase boundaries that develop between bulk-like amorphous regions that still persist and the crystalline phase,44 emerging as an intense and broad relaxation process at higher temperatures (remember Figure 7) (activation parameters in Table 3). As mentioned earlier, this type of representation enhances the underlying individual relaxation processes allowing here distinguishing better the secondary processes, in particular, the βSC-process, in the spectrum taken for the semicrystalline material relatively to Amorphous Triton; this is also a consequence of the shift toward lower frequencies of the αlike process (αSC). The invariant shape and frequency location of the secondary relaxations in both SC isothermal cold and SC nonisothermal melt samples, confirms that the length scale of the underlying molecular mechanism is below the length scale of the cooperative motion inherent to the αAm-process. From the dielectric strength of each individual process used to simulate the non-normalized ε″ derivative plot, the crystallinity degree was recalculated as 0.685. This value is in excellent agreement with the one obtained from the fit of the isothermal ε″( f) plots (0.69). The hindrance of molecular mobility associated with the dynamic glass transition in the semicrystalline material relatively to the bulk one is reinforced by the higher glass transition temperature obtained from dielectric data. Tg is estimated from the VFTH equation that describes the temperature dependence of the αsc nonisothermal melt process at τ = 100 (211.0 K) being 2.7 K above the glass transition temperature obtained from dielectric data for Amorphous Triton. It is interesting to reanalyze now the results obtained for Triton while undergoing isothermal cold-crystallization.1 At that time, it was found that the α-process under isothermal crystallization carried out in the proximity of the glass transition, depletes keeping its location invariant, which is not the case here for the αsc-process observed after nonisothermal 9801



Article

lization is slightly less mobile than the 100% amorphous sample. The ratio between the heat capacity jump of the semicrystalline material and the one of the full amorphous material45 allows to roughly estimating a crystallization degree of 0.72, which well agrees with the value found by DRS (0.69). At higher temperatures a small crystallization peak is observed (ΔH ∼ 0.8 J·g−1) followed by a broad and complexly structured endothermic peak that spans from 245 to 283 K with a melting enthalpy ΔH ∼ 56.0 J·g−1. Figure 9a also includes the thermogram for a sample isothermally cold-crystallized at Tcr = 223 K (SC isothermal cold). The thermogram for a sample melt-crystallized at the same temperature is also included superimposing the one of the cold-crystallized sample in agreement with the X-ray diffraction results. For these samples no heat capacity jump in the region of the amorphous material is detected. However, at higher temperatures, a small endothermic event is observed slightly above Tcr, which is immediately followed by crystallization and melting. The features of this endothermic discontinuity in the heat flux seem compatible with an extended glass transition, nevertheless significantly shifted to higher temperatures relatively to the amorphous glass transition. This could be an evidence of a hindered glass transition as the α′-process, as found in crystallizable systems (see the Introduction). To clarify the origin of this thermal event, two additional experiments were carried out. First, the influence of annealing time was analyzed. Figure 9b presents the thermograms collected after isothermal cold-crystallization at 223 K, which was carried out during different times. The inset in Figure 9b that illustrates the time dependence of the heat flow shows that crystallization was accomplished within the first 40 min in fair agreement with the XRD results where it was shown that the Bragg peaks do not evolve after 100 min; therefore, during the extra time spent at 223 K, Triton does not undergo any structural changes in its crystalline phase, and so, the modifications occur within the amorphous fraction manifesting as physical aging effects. As a consequence, a deviation to higher temperatures of the transition accompanied by an enhancement of the overshoot is observed, reinforcing the attribution to a glass transition. Second, TMDSC was carried out in order to deconvolute the total heat flow signal, as measured in standard DSC, in the heat capacity and kinetic components (also usually referred as reversible and nonreversible components). For the latter, kinetic events such as crystallization and enthalpy recovering contribute, while the glass transition manifests as a heat capacity change;46 melting can be visible in both components.47 Figure 10 shows the TMDSC signals: heat capacity and kinetic components and the total heat flow signal as well. The analysis allowed separating the glass transition from the kinetic signal, being identifiable in the heat capacity component. Moreover, in the kinetic signal, the enthalpy recovery and the cold-crystallization peak are clearly seen. In the inset the standard heat flow obtained at a heating rate of 5 K·min−1 is compared with the TMDSC total heat flow taken at 0.5 K· min−1, revealing the influence of the heating rate in the kinetic events: enthalpic recovery and crystallization. This gives a further evidence of the previous assignment of the endothermic event to a glass transition for the sample cold-crystallized near above Tg. This process corresponds to the low frequency contribution detected by DRS, designated as α′-process

amorphous regions slightly slower relative to the bulk-like but much more mobile relative to α′-process. This represents a plus of dielectric spectroscopy that allows differentiating through the mobility of a molecular probe, the persisting amorphous region, which is indistinguishable either by POM or X-ray. Differential Scanning Calorimetry (DSC) Analysis. To clarify the two behaviors observed for Triton upon crystallization, both standard and temperature modulated calorimetric analysis were carried out. Figure 9a presents the thermograms obtained by standard calorimetry for Triton submitted to different thermal histories,

Figure 9. (a) Thermograms obtained upon heating at 5 K·min−1 for Triton submitted previously to different thermal histories (see the Experimental Section): Amorphous and SC isothermal cold/melt at 223 K for 2 h and SC nonisothermal melt. The main events illustrated are due to glass transition and melting; the crystallization signal was partially suppressed in the figure for the amorphous sample. (b) Thermograms collected at 5 K·min−1 after isothermal coldcrystallization at 223 K for different annealing times; heat flow vs time during crystallization is presented in the inset.

which reproduce approximately the DRS conditions. The thermogram for the full amorphous sample (Amorphous in Table 2) is included as a reference exhibiting a glass transition taken at the midpoint at 212.0 K (ΔCp = 1.085 J·g−1·K−1) in agreement with previous calorimetric analysis.1 For the SC nonisothermal melt sample, a glass transition with a ΔCp = 0.307 J·g−1·K−1 was detected. However, its position is slightly shifted toward higher temperatures, 2.2 K above the glass transition of the full amorphous material; this temperature increase is in good agreement with the shift estimated by DRS (2.7 K). This result indicates once more that the remaining mobile amorphous fraction after nonisothermal melt-crystal9802



Article

dynamic information of the glassy, supercooled, and molten state. The results show two secondary processes (γAm and βAm) in the glass associated with the local mobility and a process associated with the dynamic glass transition (αAm). Due to the strong tendency of Triton to crystallize, three different crystallizations were induced and evaluated: cold and melt isothermal crystallization at temperatures near the glass transition temperature (high undercooling degree) and nonisothermal melt crystallization near below melting (low undercooling degree). The X-ray analysis gives evidence that the crystalline size for a sample crystallized upon low undercooling is larger than crystals formed upon high undercooling degree. This crystalline size difference was confirmed by POM, which revealed large and well-defined spherulites against grainy-like structure, respectively for low and high undercooling. This feature determines different constraining conditions for the αSCprocess, which was studied in detail by dielectric relaxation spectroscopy allowing establishing a relationship between structure and dynamics. From the derivative analysis of the real permittivity spectra taken for semicrystalline samples, two types of behavior were distinguished: (i) for crystallization occurring with high undercooling the αSC-process decays until complete extinction by keeping its shape and location invariant, going together with the evolving of a slower α′-like relaxation exhibiting the features of a glass transition as detected by TMDSC that revealed Tg,α′ 12 K higher than Tg,am; (ii) for crystallization occurring with low undercooling, the αSC-process slightly shifts toward lower frequencies maintaining its spectral shape, but never becomes extinct. For the latter, both DRS and DSC analysis give a slight shift to higher temperatures of the estimated glass transition temperature (∼2−3 K) relatively to full amorphous Triton. Therefore, the occurrence of a bulk-like α-process or its extinction with the evolving of a α′-process, is highly determined by the size of the amorphous regions entrapped in the interlamellar regions of the emerging spherulites. It was demonstrated for Triton X-100 that an interfacial amorphous fraction highly spatially confined in a grainy morphology disables the bulk-like α-process and gives rise to a slow process that is noticeable in both dielectric and calorimetric measurements, emerging, in the latter, as a hindered glass transition. On the other hand, when large spherulites develop, the bulk-like α-process remains active meaning that the dimensions of the remaining amorphous regions, oppositely to what happens in a grain like morphology, become comparable to the inherent length scale of the cooperative dynamics that determines the glass transition in the full amorphous material. The present paper allowed evaluating crystalline morphology as determined by the undercooling degree and to correlate it with the molecular mobility of the remaining amorphous region. This intends to be a contribution to shedding some light regarding the detection/undetection of a α′-process upon crystallization in low molecular weight materials.

Figure 10. TMDSC signals obtained on heating at 0.5 K·min−1 (heatonly modulation mode) a sample previously cold-crystallized during 2 h at 223 K: kinetic comp. (blue line), total HF (black line), and heat capacity comp. (red line) identification refer to kinetic component, total heat flow and heat capacity component of the total heat flow. Inset: the heat flow obtained for a sample cold-crystallized 2 h at 223 K obtained on heating rate at 5 K·min−1 is compared with the total heat flow signal from the TMDSC experiment shown in the main figure.

(remember Figure 8b). While in DRS the full peak never appears in the frequency window, DSC allowed to identified the heat capacity jump associated with the glass transition that is associated with α′-process, which onset is ∼12 K higher than the Tg,on of the full amorphous Triton; DRS results show in the ε″(T) trace (remember Figure 6b) a difference of around 10 K in close agreement with the shift observed by DSC. For the sample melt-crystallized at high temperatures, there is no evidence, either by DRS or DSC (red curve in Figure 9a), of the detection of an α′-process; instead, a MWS process is observed. The no detection by DSC of this MWS process seen by DRS reinforces its attribution to a interfacial polarization since DSC is blind to this kind of processes. Being established the detection of a α′-process, it seems that its emergence occurring concomitantly with the extinction of the bulk-like α-process, is highly determined by the size of the amorphous regions entrapped in the interlamellar regions of the emerging spherulites. For the grainy morphology, the nondetection of the bulk-like α-process means that the dimensions of the amorphous regions that still persist within the spherulites are smaller than the length-scale of the cooperative motion that is in its origin. Concerning, the α′process, it is detected in high crystalline low molecular weight materials as terephthalic acid dipropyl ester with only 2% of amorphous material between crystals,15 with a Tg 15 K higher to that of the full amorphous sample, close to the shift found here for Triton. This process was assigned to the mobility of molecules existing in interfacial amorphous regions adjacent to crystalline surfaces submitted to spatial confinement,15 identical attribution was done for high crystalline ethylene glycol dimethacrylate.11 Therefore, Triton provides a further example of a relatively low molecular weight material in which an interfacial amorphous fraction, characterized by a glass transition, remains in the semicrystalline material, when crystallization is induced under a high undercooling degree.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected].

4. CONCLUSIONS The dielectric behavior of fully amorphous (Am) TritonX-100 covering a broad frequency range was investigated providing

Notes

The authors declare no competing financial interest. 9803



■

Article

(17) Jiménez-Ruiz, M.; Ezquerra, T. A.; Sics, I.; Fernández-Díaz, M. T. Crystallization of 2-propanol Studied by Neutron Diffraction and Dielectric Spectroscopy in Real Time. Appl. Phys. A: Mater. Sci. Process. 2002, 74, S543−S545. (18) Shafee, E. El. The Influence of Semicrystalline Morphology on the Dielectric Relaxation Properties of Poly(3-hydroxybutyrate). Eur. Polym. J. 2001, 37, 1677−1684. (19) Fukao, K.; Miyamoto, Y. J. Relaxation Behavior of α-Process of Poly(ethylene terephthalate) During Crystallization Process. J. NonCryst. Solids 1997, 212, 208−214. Fukao, K.; Miyamoto, Y. J. Dielectric Relaxation and X-ray Scattering of Poly(ethylene terephthalate) During Crystallization Process. J. Non-Cryst. Solids 1998, 235−237, 534−538. (20) Mierzwa, M.; Floudas, G. IEEE Trans. Dielectr. Electr. Insul. 2001, 8, 359−364. Reproduced in Floudas, G. Effect of Pressure on the Dielectric Spectra of Polymeric Systems. In Broadband Dielectric Spectroscopy; Kremer, F., Schönhals, A., Eds.; Springer Verlag: Berlin, 2003; Chapter 8. (21) Cerveny, S.; Zinck, P.; Terrier, M.; Arrese-Igor, S.; Alegría, A.; Colmenero, J. Dynamics of Amorphous and Semicrystalline 1,4-transPoly(isoprene) by Dielectric Spectroscopy. Macromolecules 2008, 41, 8669−8676. (22) Alves, N. M.; Mano, J. F.; Balaguer, E.; Meseguer Dueñas, J. M.; Gómez Ribelles, J. L. Glass Transition and Structural Relaxation in Semicrystalline Poly(ethylene terephthalate): A DSC Study. Polymer 2002, 43, 4111−4122. (23) Viciosa, M. T.; Correia, N. T.; Salmerón Sanchez, M.; Gómez Ribelles, J. L.; Dionísio, M. Molecular Dynamics of Ethylene Glycol Dimethacrylate Glass Former: Influence of Different Crystallization Pathways. J. Phys. Chem. B 2009, 113, 14196−14208. (24) Schönhals, A.; Kremer, F. Analysis of Dielectric Spectra. In Broadband Dielectric Spectroscopy; Kremer, F., Schönhals, A., Eds.; Springer Verlag: Berlin, 2003; Chapter 3. (25) Boersema, A.; Van Turnhout, J.; Wübbenhorst, M. Dielectric Characterization of a Thermotropic Liquid Crystalline Copolyesteramide: 1. Relaxation Peak Assignment. Macromolecules 1998, 31, 7453−7460. (26) Schröter, K.; Unger, R.; Reissig, S.; Garwe, F.; Kahle, S.; Beiner, M.; Donth, E. Dielectric Spectroscopy in the αβ Splitting Region of Glass Transition in Poly(ethyl methacrylate) and Poly(n-butyl methacrylate): Different Evaluation Methods and Experimental Conditions. Macromolecules 1998, 31, 8966−8972. (27) Vogel, H. The Temperature Dependence Law of the Viscosity of Fluids. Phys. Zeit. 1921, 22, 645−646. (28) Fulcher, G. S. Analysis of Recent Measurements of the Viscosity of Glasses. J. Am. Ceram. Soc. 1925, 8, 339−355. (29) Tammann, G.; Hesse, W. The Dependency of Viscosity on Temperature in Hypothermic Liquids. Z. Anorg. Allg. Chem. 1926, 156, 245−257. (30) Derollez, P.; Dudognon, E.; Affouard, F.; Danède, F.; Correia, N. T.; Descamps, M. Ab Initio Structure Determination of Phase II of Racemic Ibuprofen by X-ray Powder Diffraction. Acta Crystallogr. B 2010, 66, 76−80. (31) Viciosa, M. T.; Brás, A. R.; Gómez Ribelles, J. L.; Dionisio, M. Polymerization Effects on Molecular Dynamics of n-Ethylene Glycol Dimethacrylates Followed by Dielectric Relaxation Spectroscopy. Eur. Polym. J. 2008, 44, 155−170. (32) Starkweather, H. W., Jr. Simple and Complex Relaxations. Macromolecules 1981, 14, 1277−1281. Starkweather, H. W., Jr. Aspects of Simple, Non-Cooperative Relaxations. Polymer 1991, 32, 2443− 2448. (33) Bö hmer, R.; Ngai, K. L.; Angell, C. A.; Plazek, D. J. Nonexponential Relaxations in Strong and Fragile Glass Formers. J. Chem. Phys. 1993, 99, 4201−4209. (34) Moynihan, C. T.; Macebo, P. B.; Montrose, C. J.; Gupta, P. K.; DeBolt, M. A.; Dill, J. F.; Dom, B. E.; Drake, P. W.; Esteal, A. J.; Elterman, P. B.; Moeller, R. P.; Sasabe, H.; Wilder, J. A. Structural Relaxation in Vitreous Materials. Ann. N.Y. Acad. Sci. 1976, 279, 15− 35.

ACKNOWLEDGMENTS Financial support for this work was provided through contract PEst-C/EQB/LA0006/2011 and the project PTDC/CTM/ 098979/2008 implemented within the framework of the ́ Programme “Promover a Produçaõ Cientifica, o Desenvolví e mento Tecnológico e a Inovaçaõ 002: Investigaçaõ Cientifica Tecnológica (3599-PPCDTI)” financed by Fundaçaõ para a Ciência e Tecnologia (FCT), IP. M. T. Viciosa acknowledges FCT for a postdoctoral grant from SFRH/BPD/39691/2007.

■

REFERENCES

(1) Merino, E. G.; Rodrigues, C.; Teresa Viciosa, M.; Melo, C.; Sotomayor, J.; Dionísio, M.; Correia, N. T. Phase Transformations Undergone by Triton X-100 Probed by Differential Scanning Calorimetry and Dielectric Relaxation Spectroscopy. J. Phys. Chem. B 2011, 115, 12336−12347. (2) Brás, A. R.; Malik, P.; Dionísio, M.; Mano, J. F. Influence of Crystallinity in Molecular Motions of Poly(L-lactic acid) Investigated by Dielectric Relaxation Spectroscopy. Macromolecules 2008, 41, 6419−6430. (3) Williams, G. Molecular Aspects of Multiple Dielectric Relaxation Processes in Solid Polymers. Adv. Polym. Sci. 1979, 33, 59. (4) Ezquerra, T. A.; Baltá-Calleja, F. J.; Zachmann, H. G. Real-time Dielectric Relaxation of Poly(ethylene terephthalate) During Crystallization from the Glassy State. Polymer 1994, 35, 2600. (5) Mijović, J.; Sy, J. W. Molecular Dynamics during Crystallization of Poly(L-lactic acid) As Studied by Broad-Band Dielectric Relaxation Spectroscopy. Macromolecules 2002, 35, 6370−6376. (6) Fitz, B. D.; Andjelić, S. Real-time Monitoring of Segmental Dynamics During Crystallization of poly(l(−)-lactide) by Simultaneous DRS/SALS Technique. Polymer 2003, 44, 3031. (7) Dionisio, M.; Viciosa, M. T.; Wang, Y.; Mano, J. F. Glass Transition Dynamics of Poly(L-lactic acid) during Isothermal Crystallisation Monitored by Real-time Dielectric Relaxation Spectroscopy Measurements. Macromol. Rapid Commun. 2005, 26, 1423− 1427. (8) Brás, A. R.; Viciosa, M. T.; Wang, Y.; Dionisio, M.; Mano, J. F. Crystallization of Poly(L-lactic acid) Probed with Dielectric Relaxation Spectroscopy. Macromolecules 2006, 39, 6513−6520. (9) Laredo, E.; Graimau, M.; Barriola, P.; Bello, A.; Müller, A. J. Effect of Isothermal Crystallization on the Amorphous Phase Mobility of Polycarbonate/Poly(ε-caprolactone) Blends. Polymer 2005, 46, 6532−6542. (10) Massalska-Arodz, M.; Williams, G.; Thomas, D. K.; Jones, W. J.; Dabrowski, R. Molecular Dynamics and Crystallization Behavior of Chiral Isooctyloxycyanobiphenyl as Studied by Dielectric Relaxation Spectroscopy. J. Phys. Chem. B 1999, 103, 4197−4205. (11) Viciosa, M. T.; Correia, N. T.; Salmerón Sánchez, M.; Carvalho, A. L.; Romão, M. J.; Gómez Ribelles, J. L.; Dionisio, M. Real-time Monitoring of Molecular Dynamics of Ethylene Glycol Dimethacrylate Glass Former. J. Phys. Chem. B 2009, 113, 14209−14217. (12) Alie, J.; Menegotto, J.; Cardon, P.; Duplaa, H.; Caron, A.; Lacabanne, C.; Bauer, M. Dielectric Study of the Molecular Mobility and the Isothermal Crystallization Kinetics of an Amorphous Pharmaceutical Drug Substance. J. Pharm. Sci. 2004, 93, 218−233. (13) Rengarajan, G. T.; Beiner, M. Relaxation Behavior and Crystallization Kinetics of Amorphous Acetaminophen. Lett. Drug Des. Discovery 2006, 3, 723−730. (14) Minoguchi, A.; Nozaki, R. Broadband Complex Permittivity Measurements of Sorbitol During Isothermal Crystallization. J. NonCryst. Solids 2002, 307−310, 246−251. (15) Dobbertin, J.; Hannemann, J.; Schick, C.; Potter, M.; Dehne, H. Broadband Complex Permittivity Measurements of Sorbitol During Isothermal Crystallization. J. Chem. Phys. 1998, 108, 9062. (16) Sanz, A.; Jiménez-Ruiz, M.; Nogales, A.; Martín y Marero, D.; Ezquerra, T. A. Hydrogen-Bond Network Breakage as a First Step to Isopropanol Crystallization. Phys. Rev. Lett. 2004, 93, 15503−15506. 9804



Article

(35) The Tg scaled viscosity plot that is in the origin of the fragility concept was originally given by Laughlin, W. T.; Uhlmann, D. R. J. Phys. Chem. 1972, 76, 2317. and Oldekop, W. Glasstechnische Berichte 1957, 30, 8. being revised in Ngai, K. L. Dynamic and Thermodynamic Properties of Glass-forming Substances. J. NonCryst. Solids 2000, 275, 7−51. (36) Angell, C. A. In Relaxations in Complex Systems; Ngai, K. L.; Wright, G. B., Eds.; National Technical Information Service, U.S. Department of Commerce, Springfield, VA, 1985; p 1; Angell, C. A. Relaxation in Liquids, Polymers, and Plastic CrystalsStrong/Fragile Patterns and Problems. J. Non-Cryst. Solids 1991, 131−133, 13−31; J. Res. Natl. Inst. Stand. Technol. 1997, 102, 171. (37) Kohlrausch, R. Nachtrag uber die Elastische Nachwirkung beim Cocon und Glasfaden, etc. Pogg. Ann. Phys (III) 1847, 12, 393−399. Kohlrausch, R. Theorie des Electrischen Ruckstandes in der Leidener Flasche. Pogg. Ann. Phys (IV) 1854, 1, 79−86. (38) Williams, G.; Watts, D. C. Nonsymmetrical Dielectric Relaxation Behaviour Arising from a Simple Empirical Decay Function. Trans. Faraday. Soc 1970, 66, 80−85. (39) Huang, D.; McKenna, G. B. New Insights into the Fragility Dilemma in Liquids. J. Chem. Phys. 2001, 114, 5621−5630. (40) Bottcher, C. J. F.; Bordewijk, P. Theory of Electric Polarization, Vol. 2: Dielectrics in Time-Dependent Fields; Elsevier: Amsterdam, 1978, completely revised. (41) Steeman, P. A. M.; Turnhout, van J. Fine Structure in the Parameters of Dielectric and Viscoelastic Relaxations. Macromolecules 1994, 27, 5421−5427. (42) Dias, C. Determination of a Distribution of Relaxation Frequencies Based on Experimental Relaxational Data. J. Phys. Rev. B 1996, 53, 14212−14222. (43) Wübbenhorst, M.; Turnhout, van. Analysis of Complex Dielectric Spectra. I. One-Dimensional Derivative Techniques and Three-dimensional Modelling. J. Non-Cryst. Solids 2002, 305, 40−49. (44) Steeman, P. A. M.; van Turnhout, J. Dielectric Properties in Inhomogeneous Media. Broadband Dielectric Spectroscopy; Kremer, F., Schoenhals, A., Eds.; Springer Verlag: Berlin, 2003, Chapter 13. (45) Delpouve, N.; Saiter, A.; Mano, J. F.; Dargent, E. Cooperative Rearranging Region Size in Semi-Crystalline Poly(L-lactic acid). Polymer 2008, 49 (13−14), 3130−3135. (46) Bustin, O.; Descamps, M. Slow Structural Relaxations of GlassForming Maltitol by Modulated DSC Calorimetry. J. Chem. Phys. 1999, 110, 10982−10992. (47) Sauer, B. B.; Kampert, W. G.; Neal Blanchard, E.; Threefoot, S. A.; Hsiao, D. S. Temperature Modulated DSC Studies of Melting and Recrystallization in Polymers Exhibiting Multiple Endotherms. Polymer 2000, 41, 1099−1108.

9805


Investigating the Influence of Morphology in the ... - ACS Publications

Recommend Documents