No abstract

Polyvinylidene chloride (PVDC) co-polymer resins are commonly formulated with a variety of solid additives for the purpose of processing or stabilization. A homogeneous distribution of these additives during handling and processing is important. The Dow Chemical Company developed a process to incorporate solid materials in latex form onto PVDC resin bead surfaces using a coagulation process. In this context, we present a method to characterize the distribution and thickness of these latex coatings. The difference in backscattered electron signal from the higher mean atomic number PVDC core and lower atomic number latex coating in conjunction with scanning electron microscopy (SEM) imaging using a range of accelerating voltages was used to characterize latex thickness and distribution across large numbers of beads quickly and easily. Monte Carlo simulations were used to quantitatively estimate latex thickness as a function of brightness in backscatter electron images. This thickness calibration was validated by cross-sectioning using a focused ion-beam SEM. Thicknesses from 100 nm up to about 1.3 µm can be determined using this method.

Polyvinylidene chloride (PVDC) copolymer resins including SARAN™ resins produced by The Dow Chemical Company are commonly used to produce barrier films for food packaging applications [1]. PVDC polymers are particularly advantageous in these applications because they have excellent barrier properties to the flow of oxygen and water vapor over a wide range of environmental conditions. PVDC resins are commonly formulated with a variety of solid additives, including stabilizers, lubricants, extrusion processing aides, colorants, nucleating agents and the like [1]. These are commonly added as small particles in a blending operation. While such a dry blending process is convenient, it does have its drawbacks. In a dry blend the components exist as individual particles, commonly with different particle sizes and/or densities. As a consequence, dry blended formulations are susceptible to segregation of the components during transportation and handling of the resin blend. Such maldistribution of additives can in turn have adverse effects on resin extrusion or film performance. In response, Dow developed a novel process to incorporate solid materials in latex form onto PVDC resin bead surfaces using a coagulation process [2]. This process gives a very uniform distribution of the additive on bead surfaces. Since it is locked onto the surface of the bead it also prevents segregation of the blend components. We developed microscopy techniques in response to the need to characterize the thickness and distribution of these latex coatings on PVDC resin beads.Using backscattered electron imaging chlorine-bearing material such as PVDC appears brighter than latex, composed predominantly of carbon and hydrogen. The depth reached by backscattered electrons is a function of the accelerating voltage used. As beam electrons enter a sample inelastic scattering events reduce their energy, eventually bringing them to rest. Higher energy electron beams travel deeper than lower energy beams before coming to rest. The depth reached by BSEs follows a similar trend. At a lower voltage of 5 keV the BSE signal comes mostly from the latex on top of the PVDC particle surface. The underlying PVDC can be seen in only a few patches where latex is absent or very thin ( Figure 1A). At a higher voltage of 20 keV, the beam passes through the layer of latex; the underlying bright PVDC is evident ( Figure 1C). The image collected at 10 keV is intermediate between these two; thick latex patches still appear dark, but thinner latex-coated areas appear brighter. Images such as these can be directly compared from sample to sample in order to assess relative differences in latex thickness and coverage.In order to quantitatively calibrate the BSE image brightness to latex thickness on the PVDC particle surface Monte Carlo simulations were used to model electron-sample interactions [3]. The backscatter coefficient is defined as the fraction of beam electrons that escape the sample as BSEs. Figure 2 shows the backscatter coefficient calculated from the Monte Carlo...