Viscoelastic and curing phenomena during roll-to-roll nano imprint lithography

Roll-to-Roll manufacturing aims at scaling ultraviolet- and thermally-cured nanoimprint lithography (UV-NIL, T-NIL) to commercial production speeds and volumes. Winding is the only conve- nient way of storing large quantities of nanoimprinted webs as they await unwinding in sequential R2R processes with distinct transport speeds. At production speeds, the imprinted resin is still chemically evolving when the imprinted web enters the winder, through a phenomenon called dark curing. The viscoelastic resin at various curing stages deforms under the contact pressure due to winding. This study is concerned with the impact of the contact pressure on the imprinted peak heights and potentially the functionality of the nanoimprinted surface. We develop a multiscale numerical model of the winding of the imprinted web. First the evolving properties of the resin through time are characterized, combining the effect of dark curing and viscoelasticity on the time-dependent properties. Second, a finite element model of the imprinted web uses the resin mechanical properties to determine the effective properties of the imprinted web. Finally, the winding model determines the pressure and resulting strain of the imprints in the wound roll. The surface creep is quantified. This prediction establishes how and how long the imprinted materials should be wound.

Viscoelasticity in roll-to-roll lamination

Lamination is commonly used in roll-to-roll manufacturing to create a product with specific properties. During the lamination process, two or more webs are bonded in the nip contact zone between two rollers. A common defect of laminated webs is curl, that is the inability of the web to lie flat under no tension. Curl originates from residual stresses in the laminate, often resulting from strain incompatibilities between the laminae. Complex nip mechanics induce shear and normal strains in the web laminae. Strain incompatibilities and consequent curl can be avoided when the strain in the laminae are identical, which results in a complex inverse problem for operators. A robust predictive tool can avoid using a costly trial-and-error approach. We developed a finite element model of the lamination process to determine the influence of the process parameters on the resulting curl. The model consists of two web laminae and an adhesive layer laminated by a rigid roller and a rubber-covered roller and considers the influence of web tensions, angular velocity of the roller, nip load and torque applied to the rubber-covered roller, adhesive viscoelasticity, and wrap angle. Results show that the lamination model can help predict curl but needs to include a soft adhesive layer to properly represent the physics of the nip mechanics. Curl highly depends on web tensions, as expected, but also on rubber-covered roller torque and entry wrap angle. Curl moderately depends on the nip load as a secondary factor. The properties of the adhesive may impact the laminate strain depending on the lamination speed, which controls the strain rate in the nip contact zone.

Viscoelastic Poisson’s ratio of webs

During web handling operations, the web moves along a processing line, supported by rollers, and is subject to numerous successive processes, for example several printing operations with different colors. Registration errors are caused by a change in web position between two different prints. The displacement between the different prints blurs the printed pattern. Often, products with registration errors are rejected by the customers so industrials try to reduce them as much as possible. Predicting the exact lateral position of the web during the web handling process remains difficult due to the time-dependent behavior of the web. The material property controlling this process parameter is the Viscoelastic Poisson’s Ratio (VPR).

The VPRs of different webs, low-density polyethylene (LDPE oriented and non-oriented) and nonwoven (polypropylene), are measured using Digital Image Correlation during stress relaxation and creep. The heterogeneity of the full field strains and its temporal variations are discussed. The influences on the VPR of the test, the material, its orientation, and the size of the specimen are studied with multiple ANOVAs. Finally, we present the error engendered by considering the Poisson’s ratio constant instead of considering the VPR in a time-dependent model and the consequences on the registration error during an industrial process.

The orientation is a factor influencing the VPR for anisotropic materials. Moreover, the influence of the size depends of the homogeneity of the material. For heterogenous materials, the specimen size influences the long relaxation time of the VPR. Furthermore, the strain fields recorded for the LDPEO and NW present strong heterogeneities. These heterogeneities can increase the registration errors if they occur at a printing location. Finally, the position error engendered by considering the elastic Poisson’s ratio instead of the VPR can reach a few millimeters, leading to noticeable registration erros. In conclusion, the VPR is particularly important for heterogeneous materials such as non-woven webs.