NUCLEAR MAGNETIC SPECTROSCOPY

Nuclear magnetic spectroscopy (NMR) is an analytical technique that has been around for a relatively long time. However, apart from conventional chemical analyses, its application to polyurethanes and polymer structure has lagged other methods due to the technical difficulties associated with NMR of solids [91]. The basic concepts of NMR of liquids apply to the NMR of solids generally and polyurethanes specifically. They are well covered in numerous other references [92, 93]. In the solid state, the number of interactions of any nucleus with its environment (i.e., dipolar, quadrupolar, j coupling, etc.) is very inhomogeneous and has a time dependence resulting in highly diffused resonance absorption frequency. The advent of magic angle spinning (MAS) techniques has allowed for narrow resonant peak measurements to be made on solid polymers, which would normally be too broad [94-96]. MAS spin rates are on the order of 5-10 kHz, and the magic angle of the sample placement in the magnet is 54.74°. The advent of MAS techniques has allowed for measurements to be made on solid polymers, which would normally be too broad to evaluate. Typical nuclei investigated are 13C and 1H, with 13C providing superior resolution but 1H requiring much shorter data acquisition times.

In addition to analytical improvements allowing for observations of nuclei and their chemical environments in the solid state by MAS, the physics of nuclear spin and diffusion within an NMR magnetic field has allowed accurate measurement of specific motions and characteristic times of those motions in the solid state. Thus, examples of critical information that can be provided by solid-state NMR are:

(i) Spatial order and connectivity

(ii) Small-molecule interactions

(iii) Surface interactions

(iv) Polymer dynamics and their temperature dependence

(v) Reaction monitoring

(vi) Site activity

(vii) Blend miscibility

Limitations of solid-state NMR are:

(i) Time intensive

(ii) Cost

(iii) Low sensitivity for some nuclei

Since polyurethanes are often multiphasic materials consisting of low glass transition temperature soft segments cross-linked by high glass transition hard segments, the opportunity for developing quantitative understanding of these motions and the resulting polymer properties is apparent. The use of specified pulse sequences such as Hahn spin-echo techniques allows correlation of NMR-determined spin-oriented relaxation times with morphology and mobility [97-99]. Specific to polyurethanes, the mobile soft segments and the rigid hard domains should represent clearly distinguishable environments with substantially different relaxation times. In fact, it is observed that the spin-spin relaxation time (the product of a spin-echo experiment, distinct from a spin-lattice relaxation) for hard segments is 7-20 |us, while for soft segments it is 200-1000 jjs [100]. As a matter of practicality, proton relaxation

Figure 5.29 (a) Representative stack plot of 1H MAS NMR spectra of a polyurethane foam acquired at 38°C with values of the 1H dipolar dephasing time indicated using the Hahn spin-echo technique. The resonance at 3.8 ppm reflects components of propylene oxide and ethylene oxide from the soft segment backbone, while the resonance at ~1 ppm is from the pendant methyl unit of propylene oxide. (b) Representative biexponential fit to data obtained from stack plot data. Reprinted with permission from Ref. [16]. © Elsevier Pub.

spectra are usually obtained for soft segment dynamics due to their higher mobility and accompanying longer dipolar dephasing time constants. Hard segment dynamics will often be probed by 13C NMR due to greater difficulty in resolving individual hard segment 1H peaks appearing often as a broad peak with spinning side bands extending over a wide frequency range [101].

The different scale of hard and soft segment relaxation times in the spin-echo experiment allows for isolation of the two phases and analysis of each phase for its particular characteristics [102]. In the event of each phase existing in a homogeneous environment, the relaxation decay will appear as a single exponential decay. In the event that a single phase comprises more than one environment, each environment results in a distinctive spin-spin relaxation time, and the relaxation may be well fit to a multiexponential decay described by Equation 5.15:

where P. and Tdi are the percentage and dipolar dephasing time constants for the rth component, respectively [103].

An example of experimental data and analysis is shown in Figure 5.29. A time-resolved stack plot of 1H MAS NMR spectra of selected resonances of a polyurethane soft segment is shown with the dipolar dephasing time indicated using the spin-echo technique. The least squares analysis of intensity for a specific resonance versus dipolar dephasing time provides the best fit for the dipolar time constants, which can then be interpreted to infer the mobility of a percentage of soft segment components and their subsequent role in measured polyurethane properties [16]. In the particular case of Figure 5.29, the multiexponential fit is well characterized by two time constants: one fast and one much slower. An interpretation of this observation might be that a percentage of the soft segment is mobile reflecting dynamics typical of an unencumbered soft segment, while another component is highly reflective of hard segment dynamics due to dipolar and hydrogen bonding at the hard segment-soft segment boundary. The temperature dependence of the component percentage and time constants can provide additional insight into the molecular-level dynamics responsible for observed polymer properties.