Frederick I. Mopsik

Molecular dipole electrets

The total polarization due to molecular dipoles in a glassy electret is computed using an Onsager cavity approach. From this result, all the possible contributions to the piezoelectric and pyroelectric coefficients are considered. It is shown that there are major contributions from the variation in dielectric constant and, for pyroelectricity, from thermal motion. These results account well for experimental data for polyvinyl chloride.

Precision time‐domain dielectric spectrometer

A description is given for an automated method for determining dielectric constant and loss by the measurement of the time response of the dielectric to a step voltage. Attention is paid to the circuits necessary to achieve high accuracy (0.1%) and high sensitivity (tan δ=10−5) over audio and subaudio frequencies (104 to 10−4 Hz). These include a 100‐V step generator accurate to 5 ppm, a charge detector with a time‐independent bias current of 30 fA, and a clock that can control sampling time from 5 μs to 10 s. In addition, a numerical Laplace transform, based on a cubic spline, is described that preserves the accuracy of the time data when they are transformed into the frequency domain.

Dielectric Properties of Slightly Polar Organic Liquids as a Function of Pressure, Volume, and Temperature

The dielectric constant and density have been measured simultaneously for carbon tetrachloride, carbon disulfide, isopentane, and toluene. The Clausius–Mossotti polarization for all liquids at constant temperature shows a linear decrease with increasing density. Only isopentane and toluene show a temperature dependence of the Clausius–Mossotti polarization at constant density consistent with the presence of a permanent dipole moment. The dipole moments derived from the constant density data are 0.105 D for isopentane and 0.332 D for toluene. Also, the Bottcher–Onsager equation is shown to be incompatible with the measured data.

Numerical evaluation of the dielectric polarization distribution from thermal‐pulse data

A method for numerically carrying out the Fourier analysis for the thermal‐pulse experiment is given. It is shown that it is possible to obtain the polarization distribution across the thickness of a thin film (25 μm) to within the limits set by the experimental data. For such films, resolution of the distribution to within 0.1 of the film thickness is possible. Results are given for the experiment by using a charge measurement rather than a voltage measurement. The effect of a finite‐width pulse is shown to cut off the Fourier coefficients in such a way as to smooth any distribution. Pulsing the sample alternately on both sides is shown to greatly increase the resolution of the experiment. Results for a PVF2 film and a P(VF2‐TFE) copolymer film show that interesting details can be found by the experiment.

Bulk Modulus and Grüneisen Parameters for Linear Polymers

Expressions are derived for the isothermal bulk modulus BT and Anderson–Gruneisen parameter δ for a simple two‐dimensional bundle‐of‐chains model where parallel nearest‐neighbor chains are assumed to interact with a simple separation‐dependent potential. Bulk modulus data for general linear polymers are examined and shown to be essentially temperature independent at constant volume in accord with the model. The theory is shown to predict quite accurately for reasonable assumed pair potentials both BT and δ of high‐density polyethylene and n‐paraffins with no adjustable parameters. The values of δ obtained for polyethylene from BT data are in excellent agreement with those reported from ultrasonic data. The analysis is also shown to provide a sensitive measure of the form of the effective chain pair potential near its minimum. Also, a close relationship between δ and the Gruneisen constant γ is derived for this model.

Dynamics of polybutadienes with different microstructures. 2. Dielectric response and comparisons with rheological behavior

A series of 1,2-1,4-polybutadienes with varying 1,2 vinyl content was investigated using time-domain dielectric spectroscopy. The time range was 10 Ϫ5-300 s, which can be converted by Laplace transform to a frequency range of 10 Ϫ3-6000 Hz. The samples were the same as those used in a previous rheological study from these laboratories. Therefore, a direct comparison of dielectric and mechanical responses was possible. Within experimental uncertainty, the ␣ relaxation observed by both methods shows the same temperature dependence but there is an offset between the characteristic times of both methods, which increases with increasing vinyl content. This result can be qualitatively understood from the difference of the size of the dipolar groups, viz. cis and vinyl monomeric units, in the context of the DiMarzio-Bishop model. In addition, the question of time-temperature superposition was studied using the dielectric data. In the cases of a vinyl content у0.53 no deviations from time-temperature superposition were detected. Only for the sample with the lowest vinyl content 0.07 does the attempt to construct a master curve from the dielectric loss data fail. In this case a fit with a combination of a Havriliak-Negami and a Cole-Cole functions suggests that this deviation from time-temperature superposition is an intrinsic feature of the ␣ relaxation rather than an effect of its merging with the ␤ relaxation. The absence of indications for such a deviation in the rheological study can be explained by the smaller frequency range of the latter. This stresses the necessity of a large dynamic range in experiments aimed at the examination of the time-temperature superposition principle.

Preexisting polarization and influence of electrode materials on PVF2 electrets as determined by thermal pulse and pyroelectric methods

A number of polyvinylidene fluoride (PVF2) electrets were prepared with different permutations of gold and aluminum electrodes and poled with dc fields up to 160 MV m−1 at room temperature. Polarization distributions were measured by the thermal pulse method and pyroelectric coefficients were determined. Quantitative measurements were made of a significant level of polarization in nominally unpoled PVF2 and a contact electrification mechanism was proposed. No consistent effects of electrode materials on polarization distribution were found. PVF2 poled at room temperature has its highest polarization near the center of the thickness in contrast to the results on samples poled at elevated temperatures and cooled inhomogeneously with the field applied.

A precision capacitance cell for measurement of thin film out-of-plane expansion. II. Hygrothermal expansion

The data reduction techniques necessary for utilizing the capacitance cell described previously by us for thickness measurements [C. R. Snyder and F. I. Mopsik, Rev. Sci. Instrum. 69, 3889 (1998)] in a humid environment are presented. It is demonstrated that our data reduction techniques provide thicknesses that are equivalent to those measured under dry conditions within the expected experimental uncertainty. The utility of this technique is demonstrated by the measurement of the hygrothermal expansion (swelling) of a bisphenol-A based epoxy with novolac hardener and fused silica filler. Our technique is shown to have a higher sensitivity than most current thermomechanical analysis techniques and is readily amenable to humid conditions.

Limitations on Distinguishing Between Representations of Relaxation Data Over Narrow Frequency Ranges

In this article, we examine the ability to distinguish between relaxation functions with data over a limited range of frequency. It is demonstrated that over these limited frequency ranges under a variety of conditions, the Cole–Cole equation can be used to fit data generated by the Havriliak–Negami equation. These results show that discerning between several very different broad relaxation functions fit to data obtained over narrow time or frequency ranges is nearly impossible within experimental accuracy. Therefore, the uniqueness of the fit parameters, and hence the ability to verify model predictions, is brought into question. Furthermore, as this conclusion is drawn from comparison of exact functions that experience no dispersion overlaps or instrumental systematic errors that can mask exact fits, the true situation with experimental data is even worse. The same conclusion can be applied to time domain data.

Piezo- and pyro-electricity in polymer electrets

In 1960 Gubkin1 published an electrostatic theory of piezoelectricity which did not involve the microscopic structure of the electret. This approach would seem to be a good description of the phenomenon of piezoelectricity in polymer materials, many of which are partially or completely amorphous. We applied Gubkins model to a specific set of experimental conditions, which resulted in some simplifications. These conditions are: 1. We measure changes in surface charge at nearly zero potential. 2. We use thin evaporated metal electrodes so that the electrodes change area in accord with changes in specimen area, A. 3. We consider only amorphous specimens which we assume to be elastically isotropic. 4. We assume the free charges to be frozen in the solid and to move in proportion to macroscopic strains, and the polarization charges to be rigid molecular dipoles whose positions move in proportion to macroscopic strains but whose effective total moment remains constant as long as the strains are isotropic.