Reaction-induced phase separation in polyethersulfone-modified epoxy-amine systems studied by temperature modulated differential scanning calorimetry

In epoxy-amine systems with a thermoplastic additive, the initially homogeneous reaction mixture can change into a multi-phase morphology as a result of the increase in molecular weight or network formation of the curing matrix. Temperature modulated DSC (TMDSC) allows the real-time monitoring of this reaction-induced phase separation. A linear polymerizing epoxy-amine (DGEBA–aniline) and a network-forming epoxy-amine (DGEBA–methylene dianiline), both with an amorphous engineering thermoplastic additive (polyethersulfone, PES), are used to illustrate the effects of phase separation on the signals of the TMDSC experiment. The non-reversing heat flow gives information about the reaction kinetics. The heat capacity signal also contains information about the reaction mechanism in combination with effects induced by the changing morphology and rheology such as phase separation and vitrification. In quasi-isothermal (partial cure) TMDSC experiments, the compositional changes resulting from the proceeding phase separation are shown by distinct stepwise heat capacity decreases. The heat flow phase signal is a sensitive indication of relaxation phenomena accompanying the effects of phase separation and vitrification. Non-isothermal (post-cure) TMDSC experiments provide additional real-time information on further reaction and phase separation, and on the effect of temperature on phase separation, giving support to an LCST phase diagram. They also allow measurement of the thermal properties of the in situ formed multi-phase materials.     

Reversible and irreversible heat capacity of poly(trimethylene terephthalate) analyzed by temperature‐modulated differential scanning calorimetry

The heat capacity of poly(trimethylene terephthalate) (PTT) has been analyzed using temperature‐modulated differential scanning calorimetry (TMDSC) and compared with results obtained earlier from adiabatic calorimetry and standard differential scanning calorimetry (DSC). Using quasi‐isothermal TMDSC, the apparent reversing and nonreversing heat capacities were determined from 220 to 540 K, including glass and melting transitions. Truly reversible and time‐dependent irreversible heat effects were separated. The extrapolated vibrational heat capacity of the solid and the total heat capacity of the liquid served as baselines for the analysis. As one approaches the melting region from lower temperature, semicrystalline PTT shows a reversing heat capacity, which is larger than that of the liquid, an observation that is common also for other polymers. This higher heat capacity is interpreted as a reversible surface or bulk melting and crystallization, which does not need to undergo molecular nucleation. Additional time‐dependent, reversing contributions, dominating at temperatures even closer to the melting peak, are linked to reorganization and recrystallization (annealing), while the major melting is fully irreversible (nonreversing contribution). © 2000 John Wiley & Sons, Inc. J Polym Sci B: Polym Phys 38: 622–631, 2000     

Reversible melting of polyethylene extended‐chain crystals detected by temperature‐modulated calorimetry

The melting and crystallization of extended‐chain crystals of polyethylene are analyzed with standard differential scanning calorimetry and temperature‐modulated differential scanning calorimetry. For short‐chain, flexible paraffins and polyethylene fractions up to 10 nm length, fully reversible melting was possible for extended‐chain crystals, as is expected for small molecules in the presence of crystal nuclei. Up to 100 nm length, full eutectic separation occurs with decreasingly reversible melting. The higher‐molar‐mass polymers form solid solution crystals and retain a rapidly decreasing reversible component during their melting that decreases to zero about 1.5 K before the end of melting. An attempt is made to link this reversible melting to the known, detailed morphology and phase diagram of the analyzed sample that was pressure‐crystallized to reach chain extension and practically complete crystallization. © 2002 Wiley Periodicals, Inc. J Polym Sci Part B: Polym Phys 40: 2219–2227, 2002     

Isothermal crystallization of concentrated amorphous starch systems measured by modulated differential scanning calorimetry

The slow isothermal crystallization of concentrated amorphous starch systems is measured by Modulated Differential Scanning Calorimetry (MDSC). It can be followed continuously by the evolution (stepwise decrease) of the MDSC heat capacity signal (Cp), as confirmed with data from X-ray diffractometry, Dynamic Mechanical Analysis, Raman spectroscopy, and conventional Differential Scanning Calorimetry. Isothermal MDSC measurements enable a systematic study of the slow crystallization process of a concentrated starch system, such as a pregelatinized waxy corn starch with 24 wt % water and 76 wt % starch. After isothermal crystallization, a broad melting endotherm with a bimodal distribution is observed, starting about 10°C beyond the crystallization temperature. The bulk glass transition temperature (T_g) decreases about 15°C during crystallization. The isothermal crystallization rate goes through a maximum as a function of crystallization time. The maximum rate is characterized by the time at the local extreme in the derivative of Cp (t_max), or by the time to reach half the decrease in Cp (t_1/2). Both t_max and t_1/2 show a bell-shaped curve as a function of crystallization temperature. The temperature of maximum crystallization rate, for the system studied, lies as high as 75°C. This is approximately 65°C above the initial value of T_g. Normalized Cp curves indicate the temperature dependence of the starch crystallization mechanism. © 1999 John Wiley & Sons, Inc. J Polym Sci B: Polym Phys 37: 2881–2892, 1999     

Temperature modulated calorimetry and dielectric spectroscopy in the glass transition region of polymers

The results from temperature modulated DSC in the glass transition region of amorphous and semicrystalline polymers are described with the linear response approach. The real and the imaginary part of the complex heat capacity are discussed. The findings are compared with those of dielectric spectroscopy. The frequency dependent glass transition temperature can be fitted with a VFT-equation. The transition frequencies are decreased by 0.5 to 1 orders of magnitude compared to dielectric measurements. Cooling rates from standard DSC are transformed into frequencies. The glass transition temperatures are also approximated by the VFT-fit from the temperature modulated measurements. The differences in the shape of the curves from amorphous and semicrystalline samples are discussed.Dedicated to Professor Bernhard Wunderlich on the occasion of his 65th birthday     

Specific reversible melting of polymers

Temperature‐modulated differential scanning calorimetry can detect a certain amount of reversible latent heat in flexible macromolecules. In short, one can identify a reversible melting in such polymers earlier thought to exhibit only fully irreversible crystallization and melting. Details of the reversible melting of isotactic polypropylene and ethylene‐1‐octene copolymers of low and medium densities have newly been measured and linked to the crystallization, annealing, or melting temperature. It is possible to assign the experimental reversibility of melting to specific crystal fractions that ultimately melt irreversibly at higher temperatures; that is, it is suggested that reversible melting mainly occurs only between the temperatures of their formation and their zero‐entropy‐production melting temperature, at which they change to a melt of the same degree of metastability. This is supported by the almost complete absence of reversibility below the temperature of crystal formation and the observation of a distinct relationship between the amount of irreversibly by annealing reorganized material and reversibility in the case of isotactic polypropylene. A given crystal fraction, characterized by its formation temperature and zero‐entropy‐production melting temperature, has a specific reversibility of the melt‐to‐crystal transition, which is represented by the ratio of the reversible latent heat to the total enthalpy change when the crystal fraction of interest ultimately melts. This specific reversibility is, for ethylene‐1‐octene copolymers, at least 25% at temperatures in the primary crystallization range, and this indicates that the reversible contribution to the total of the melting processes is much larger than expected from simple calculations by the excess apparent reversible heat capacity being referred to the heat of fusion of the polymer, as is commonly done. © 2003 Wiley Periodicals, Inc. J Polym Sci Part B: Polym Phys 41: 2039–2051, 2003     

Reversibility of the low‐temperature transitions of polytetrafluoroethylene as revealed by temperature‐modulated differential scanning calorimetry

Temperature‐modulated differential scanning calorimetry reveals distinct differences in the kinetics of the low‐temperature phase transitions of polytetrafluoroethylene. The triclinic to trigonal transition at 292 K is partially reversible as long it is not complete. As soon as the total sample is converted, supercooling is required to nucleate the reversal of the helical untwisting involved in the transition. The trigonal phase can be annealed in the early stages after transformation with a relaxtion time of about 5 minutes. The dependence of the reversing heat capacity on the modulation amplitude, after a metastable equilibrium has been reached, is explained by a non‐linear, time‐independent increase of the heat‐flow rate, perhaps caused by an increased true heat capacity. The order‐disorder‐transition at 303 K from the trigonal to a hexagonal condis phase is completely reversible and time‐independent. It extends to temperatures as low as the transition at 292 K or even lower. Qualitatively, the thermal history and crystallization conditions of polytetrafluoroethylene do not affect the transition kinetics, that is, melt‐crystallized film and as‐polymerized powders show similar transition behaviors, despite largely different crystallinities. © 2001 John Wiley & Sons, Inc. J Polym Sci B: Polym Phys 39: 750–756, 2001     

Computer simulation of the melting kinetics of polymer crystals under condition of modulated temperature

Computer simulation has been applied to the modeling of the melting kinetics of polymer crystals, which we have recently presented to predict the response of the kinetics to a sinusoidal modulation in temperature on heating. The frequency and heating-rate dependencies have been examined with a Gaussian or uniform distribution of the melting points. For both of the distributions, the details of the dependence have been examined on the basis of the analytical results of the modeling. It has also been confirmed that the response of the kinetics has higher harmonics as expected from the formulation of the modeling. This behavior corresponds to the experimental results of temperature-modulated DSC (T-MDSC) in the melting region of polymer crystals.     

Temperature‐modulated differential scanning calorimetry studies on the origin of double melting peaks in isothermally melt‐crystallized poly(L‐lactic acid)

In this work, the melting behaviors of nonisothermally and isothermally melt‐crystallized poly(L ‐lactic acid) (PLLA) from the melt were investigated with differential scanning calorimetry (DSC) and temperature‐modulated differential scanning calorimetry (TMDSC). The isothermal melt crystallizations of PLLA at a temperature in the range of 100–110 °C for 120 min or at 110 °C for a time in the range of 10–180 min appeared to exhibit double melting peaks in the DSC heating curves of 10 °C/min. TMDSC analysis revealed that the melting–recrystallization mechanism dominated the formation of the double melting peaks in PLLA samples following melt crystallizations at 110 °C for a shorter time (≤30 min) or at a lower temperature (100, 103, or 105 °C) for 120 min, whereas the double lamellar thickness model dominated the formation of the double melting peaks in those PLLA samples crystallized at a higher temperature (108 or 110 °C) for 120 min or at 110 °C for a longer time (≥45 min). © 2007 Wiley Periodicals, Inc. J Polym Sci Part B: Polym Phys 45: 466–474, 2007     

Integrated circuit thermopile as a new type of temperature modulated calorimeter

Thermopiles, integrated into a thin silicon membrane, are used in some calorimetric applications (e.g. Setaram SETline120 portable thermal analyzer) that utilize temperature ramps of the thermostat to obtain the caloric response. A new calorimetric method is proposed, which uses an integrated circuit thermopile (ICT) in an oscillating mode. A heater integrated into the membrane drives temperature oscillations and the thermopile senses the temperature gradient across the membrane. ac calorimetry and the 3-omega method do not measure total heat losses and have an inherent quasi-adiabatic low frequency limit. On the other hand, the thermopile setup can measure total heat losses that permits one to monitor enthalpy changes in a sample, e.g. during crystallization. Low frequency measurements are limited only by sensor sensitivity. Preliminary results show that the same experimental setup can be used to make dynamic heat capacity measurements over a frequency range from 1 mHz to 100 Hz. At high frequencies (1 Hz and higher), heat capacity of nanogram samples can be measured.     

Temperature modulated differential scanning calorimetry (TMDSC) in the region of phase transitions. Part 2: some specific results from polymers

By means of four different examples (pressure crystallised, gel crystallised, nascent and highly stretched polyethylenes (PEs)) it is shown that temperature modulated DSC offers advantages against common DSC. It is possible to see dynamic processes inside the sample during melting. This way we found (i) that during melting of high pressure crystallised PE the so-called ₂-process (known from DMA) takes place, (ii) the lamellae doubling in gel crystallised UHMWPE can be seen in TMDSC signals, though no balance heat flow rate is visible in the common DSC, (iii) the same is true for the recrystallisation in nascent and highly stretched PE many degrees before the melting peak appears. To separate these results from the measured curves the knowledge of the heat transport into and within the sample is needed. A simple low pass filter model has proved its worth for this purpose.     

Learning about calorimetry

Calorimetry deals with the energetics of atoms, molecules, and phases and can be used to gather experimental details about one of the two roots of our knowledge about matter. The other root is structural science. Both are understood from the microscopic to the macroscopic scale, but the effort to learn about calorimetry has lagged behind structural science. Although equilibrium thermodynamics is well known, one has learned in the past little about metastable and unstable states. Similarly, Dalton made early progress to describe phases as aggregates of molecules. The existence of macromolecules that consist of as many atoms as are needed to establish a phase have led, however, to confusion between colloids (collections of microphases) and macromolecules which may participate in several micro- or nanophases. This fact that macromolecules can be as large or larger than phases was first established by Staudinger as late as 1920. Both fields, calorimetry and macromolecular science, found many solutions for the understanding of metastable and unstable states. The learning of modern solutions to the problems of materials characterization by calorimetry is the topic of this paper.This work was financially supported by the Div. of Materials Res., NSF, Polymers Program, Grant # DMR 90-00520 and Oak Ridge National Laboratory, managed by Lockheed Martin Energy Research Corp. for the U. S. Department of Energy, under contract number DE-AC05-96OR22464. Support for instrumentation came from TA Instruments, Inc. Research support was also given by ICI Paints, and Toray Industries, Inc.     

Modeling the heat flow and heat capacity of modulated differential scanning calorimetry

Modulated differential scanning calorimetry (MDSC) uses an abbreviated Fourier transformation ?r the data analysis and separation of the reversing component of the heat flow and temperature signals. In this paper a simple spread-sheet analysis will be presented that can be used to better understand and explore the effects observed in MDSC and their link to actual changes in the instrument and sample. The analysis assumes that instrument lags and other kinetic effects are either avoided or corrected for.     

Crystallization and melting of poly(oxyethylene) analyzed by temperature-modulated calorimetry

Temperature-modulated calorimetry, TMC, is used to evaluate the temperature region of metastability between crystallization and melting. While crystals like indium can be made to melt practically reversibly during a TMC cycle of low amplitude so that sufficient crystal nuclei remain unmelted, linear macromolecules cannot, because of their need to undergo molecular nucleation. Modulation amplitudes varying from ±0.2 to ±3.0 K are used to assess the temperature gap between the slow crystallization region and the melting of metastable crystals of poly(oxyethylene) (PEO) of molar mass 1500 Da. This low molar mass PEO serves as a model compound with a metastable gap of melting/crystallization that can be bridged by TMC with a large modulation amplitude. © 1997 John Wiley & Sons, Inc.

1 This article is a US Government work and, as such, is in the public domain in the United States of America.

J Polym Sci B: Polym Phys 35 : 1877–1886, 1997     

Isothermal curing of an epoxy resin by alternating differential scanning calorimetry

The quasi-isothermal curing of a diepoxide resin with a triamine of polyoxypropylene was studied by alternating differential scanning calorimetry (ADSC), which is a temperature modulated DSC technique. The complex heat capacity measurements allows to analyse the vitrification process at curing temperatures (T_c) below the maximum glass transition of the fully cured epoxy (T_g∞=85.8°C). Initially, the modulus of the complex heat capacity, |C^*_p|, increases until a maximum (conversion between 0.42 and 0.56) and then decreases. This step is followed by an abrupt decay of |C^*_p|, due to the vitrification of the system, which allows the determination of the vitrification time. This value agrees well with that determined by the partial curing method. The phase angle and out-of-phase heat capacity show an asymmetric wide peak during the vitrification process. The change in |C^*_p| at vitrification decreases with the increase of T_c becoming zero at temperature T_g∞. This epoxy-triamine system shows a delay of the vitrification process respect to other model epoxy systems probably due to the presence of polyoxypropylene chains in the network.

The decay of |C^*_p| during vitrification may be normalised between unity and zero by defining a mobility factor. This mobility factor has been used to simulate the reaction rate during the stage where the reaction is controlled by diffusion. The observed reaction rate is simulated by the product of the kinetic reaction rate, determined by the autocatalytic model, and the mobility factor.     

Solution calorimetry to monitor swelling and dissolution of polymers and polymer blends

The observed rate of drug release from a polymeric drug delivery system is governed by a combination of diffusion, swelling and erosion. It is thus not a simple task to determine the effects of the polymer on the observed drug release rate, because the swelling characteristics of the polymer are inferred from the drug release profile. Here we propose to use solution calorimetry to monitor swelling. Powdered polymer samples (HPMC E4M, K4M, K15M and NaCMC, both alone and in a blend) were dispersed into water or buffer (pH 2.2 and 6.8 McIlvaine citrate buffers) in a calorimeter and the heat associated with the swelling phenomena (hydration, swelling, gelation and dissolution) was recorded. Plots of normalised cumulative heat (i.e. q_t/Q, where q_t is the heat released up to time t and Q the total amount of heat released) versus time were analysed by the power law model, in which a fitting parameter, n, imparts information on the mechanism of swelling.

For all systems the values of n were greater than 1, which indicated that dissolution occurred immediately following hydration of the polymer. However, while not suitable for determining reaction mechanism, the values of n for each polymer were significantly different and, moreover, were observed to vary both as a function of particle size and dissolution medium pH. Thus, the values of n may serve as comparative parameters. Properties of the polymer blends were observed to be different from those of either constituent and correlated with the behaviour seen for polymer tablets during dissolution experiments. The data imply that solution calorimetry could be used to construct quantitative structure–activity relationships (QSARs) and hence to optimise selection of polymer blends for specific applications.     

Temperature-modulated differential scanning calorimetry of reversible and irreversible first-order transitions

Temperature-modulated differential scanning calorimetry of first-order transitions has led to many new observations. Some of these involve non-linear processes or deal with transformations of practically instantaneous response. The latter may cause serious lags within the calorimeter due to limited thermal conductivity of the sample and the instrument. In both cases the “reversing heat capacity” or a “complex heat capacity” is not a precise representation of the transition since both are computed from abbreviated Fourier transforms, limited to the evaluation of the first harmonic component. One has in these cases to work in the time-domain with the raw output. But even from these analyses in the time-domain many interesting new insights about the transition and the calorimeter performance can be generated.     

Glass transition temperature of thin polycarbonate films measured by flash differential scanning calorimetry

Flash differential scanning calorimetry was used to study the glass transition temperature T_g of polycarbonate ultrathin films. The investigation was made as a function of film thickness from 22 to 350 nm and over a range of cooling rates from 0.1 to 1000 K/s. Polycarbonate spin cast films were floated on a layer of grease on the calorimetric chip. The results show a greatly reduced glass temperature for the thinnest films relative to the macroscopic value. We also observed that the magnitude of the glass temperature reduction decreases as the cooling rate increases with the highest cooling rates showing little thickness dependence of the T_g. Dynamic fragility and activation energy at T_g were found to decrease with decreasing film thickness. The results are discussed in the context of literature reports for supported and freely standing polycarbonate films. © 2014 Wiley Periodicals, Inc. J. Polym. Sci., Part B: Polym. Phys. 2014 , 52, 1462–1468