Hydrogen Bonds under Stress: Strain-Induced Structural Changes in Polyurethane Revealed by Rheological Two-Dimensional Infrared Spectroscopy

Abstract

The remarkable elastic properties of polymers are ultimately due to their molecular structure, but the relation between the macroscopic and molecular properties is often difficult to establish, in particular for (bio)polymers that contain hydrogen bonds, which can easily rearrange upon mechanical deformation. Here we show that two-dimensional infrared spectroscopy on polymer films in a miniature stress tester sheds new light on how the hydrogen-bond structure of a polymer is related to its viscoelastic response. We study thermoplastic polyurethane, a block copolymer consisting of hard segments of hydrogen-bonded urethane groups embedded in a soft matrix of polyether chains. The conventional infrared spectrum shows that, upon deformation, the number of hydrogen bonds increases, a process that is largely reversible. However, the 2DIR spectrum reveals that the distribution of hydrogen-bond strengths becomes slightly narrower after a deformation cycle, due to the disruption of weak hydrogen bonds, an effect that could explain the strain-cycle induced softening (Mullins effect) of polyurethane. These results show how rheo-2DIR spectroscopy can bridge the gap between the molecular structure and the macroscopic elastic properties of (bio)polymers.

Figure 1

(A, B) Schematic of rheo-IR and rheo-2DIR. (C) Structure of polyurethane, a block copolymer composed of soft and hard segments. The urethane groups in the hard segments form strong hydrogen bonds that act as physical cross-links. Because of the different polarity and chemical nature, the soft and hard segments separate, leading to the formation of hard domains embedded in a soft phase. (D) Stress–strain curve of a TPU film at a strain rate of 8 s^–1 with a strain step-size of 3%. The stress–strain curve in the second cycle is more compliant than in the first cycle (strain softening). Filled circles indicate the deformations at which rheo-2DIR was performed.

Figure 2

(A) IR absorption spectra of TPU film at 0% deformation. (B) IR spectra of TPU at 0, 200, 500% deformation normalized with respect to the respective total spectrum area to compensate for thinning effect. (C) 2D gradient FT-IR map with respect to the deformation percentage. (D) IR spectra of before deformation and upon recovery of the zero-stress condition.

Figure 3

(A) 2DIR spectrum of TPU at zero stress condition at a T_w = 1 ps. (B) 2DIR spectra and nodal line slopes at zero stress and upon recovery of zero-stress condition after a deformation up to 500% at a T_w = 1 ps. The nodal line slope at zero stress is shifted along the probe axis in the right 2DIR spectrum for comparison.

Figure 4

(A) Nodal line slope during a deformation cycle. (B) Nodal line slopes before and after a deformation cycle in 5 independent experiments. In the thicker samples (Protex 002), we can only measure 2DIR spectra at 200% strain (for lower deformation the IR absorption is too high). (C) Comparison of bleach diagonal slices extracted by 2DIR spectra in 3B. Error bars represent +/– one standard deviation.