Our group studies the composition and nanoscale structure as well as dynamics of carbon materials, polymers, nanocomposites, proteins, and natural organic matter, using a variety of quantitative, selective, or two-dimensional nuclear magnetic resonance (NMR) experiments, many of which were developed by us ( link ). We have also introduced new methods for quantitative analysis of scattering data of nanostructured materials. This dual approach has enabled us to "solve" important aspects of the structure of the Nafion fuel-cell membrane, nanodiamond, chars, and the nanocomposite in bone.
We have developed expertise in the detailed and quantitative NMR characterization of highly aromatic carbon materials, some with applications in catalysis. Studies of biochars, carbon-coated mesoporous oxides, and melanoidins formed in the Maillard reaction [J. Agric. Food Chem. 2009, 2011; Chem. Mater. 2014] fall into this category. Solid-state 13C NMR is the best available method for characterizing these materials comprehensively, quantifying the aromaticity and functional group concentrations, and characterizing the size of clusters of fused aromatic rings. When these materials are made from 13C-enriched precursors, distinction of heterocycles and identification of many different linkages is possible, best in spectrally edited two-dimensional 13C-13C experiments without diagonal ridge. We have developed an approach to “translate” the NMR spectra into realistic structural models of the typical repeat unit of complex organic materials. [Carbon, 2014]
Detonation nanodiamond is a fascinating material produced in >20 t quantities annually, consisting of 5-nm diameter carbon particles. Many structural models show a partially aromatic particle surface layer, but 13C NMR documents ≤ 1% of aromatic carbons in pristine purified detonation nanodiamond. Instead, the surface is found to be covered by C-H, C-OH, and C=O moieties. The 12-14% surface carbons (i.e. with bonds to ≤3 carbons) expected for a bulk-terminated diamond particle are thus accounted for. NMR has also revealed a nonaromatic shell and provided the first estimate of the depth of the unpaired electrons from the particle surface. [JACS, 2009; J. Chem. Phys. C, 2014 & 2015]
Knowledge of the chemical structure of organic matter in the environment is required for a true understanding of soil formation, its role in the carbon cycle, or sorption of nutrients and contaminants. Solid-state NMR provides unique opportunities for a comprehensive analysis these otherwise nearly intractable systems. We have developed methods for obtaining both quantitative and selective 13C NMR spectra of organic matter from the environment. Our studies have shown the molecular origin of a yield decline in intensive rice farming [PNAS, 2004] and that oxidized char residues account for up to 50% of all C in fertile Midwestern soils. [Environ. Sci. Technol., 2012]
Semicrystalline polymers such as polyethylene or polypropylene combine stiffness and toughness due to alternating rigid nanocrystallites and soft amorphous layers. As a result of chain connectivity across the crystalline-amorphous interface, the traditional structural models unintentionally but unavoidably produce excess density in the noncrystalline reqions. We have identified the avoidance of such density anomalies as a guiding principle for predicting structural features such as chain ends at the crystal surface or pronounced chain tilt in the crystallites that reduce the chain crowding at the interface. We are using NMR and other techniques to document these predictions.
To determine the shape and size of nanoparticles and other nanostructures by NMR is very challenging. This information is obtained much more easily from microscopy or scattering data, for instance small-angle scattering patterns of oriented samples. Even for unoriented systems, small-angle scattering curves I(θ) or I(q) contain information on particle shape and packing. To take full advantage of this approach, we have introduced an algorithm to calculate the I(q) curve from any model defined on a cubic lattice with periodic boundary conditions. The approach is based on numerical 3D Fourier transformation as used in multidimensional NMR. [J. Appl. Cryst., 2007] As a complementary approach to determine the shapes of nanocrystals, we are also simulating wide-angle diffraction patterns.
The proton-exchange membrane (PEM) is a central, and often performance-limiting, component of all-solid H2/O2fuel cells. Nafion®, the most widely used PEM, consists of a perfluorinated polymer that combines a hydrophobic Teflon-like backbone with hydrophilic ionic side groups. It stands out among polymer materials for its high, selective permeability to water and small cations.
We continue to work on developing radio-frequency pulse sequences for extending the capabilities of solid-state NMR. We have introduced the following techniques:
Protein spectral editing. Overlap of cross peaks, for instance in 13C-13C NMR spectra, of proteins interferes with secondary structure analysis and peak assignment. We have developed spectral editing techniques to obtain signals of just Ile, Val, and Leu based on CH selection, and of Glu and Asp based on COO selection. For the latter, complete suppression of C bonded to N was achieved by asymmetric 13C{15N} REDOR. [J. Biomol. NMR 2013]
Quantitative 13C multiCP NMR. Unlike IR or Raman spectroscopy, NMR is intrinsically quantitative: fractional peak areas are equal to the fractions of carbons in the respective environments, if the experiment is performed properly. The gold standard for quantitative 13C NMR is direct polarization with a sufficiently long recycle delay; this, however, may require waiting for hundreds of seconds between scans. Instead, cross polarization (CP) from 1H is commonly used, where the much faster longitudinal relaxation of 1H allows for recycle delays of a few seconds. However, often underrepresents signals of carbon far from the nearest proton, in particular for complex materials with fast or differential T1r relaxation during CP. We have recently demonstrated that this problem can be overcome by the multiCP approach. [J. Magn. Reson., 2014]
We have established NMR as a tool for characterizing nanocomposites, typically consisting of inorganic nanoparticles dispersed in an organic polymer, not only in terms of composition, but also in terms of the location of various components relative to the organic-inorganic interface. The composition of the surface layers of both the nanoparticles and the matrix, which are particularly important since they are responsible for the interactions between the components, can be determined by selective NMR techniques, often based on heteronuclear dipolar couplings across the interface. This approach has enabled us to demonstrate the presence of citrate strongly bound to apatite nanocrystals in bone, which stabilizes the thin crystals at thicknesses of only 3 – 4 unit cells. [PNAS, 2010)]
The systematic conformation dependence of 13C chemical shifts in proteins provides useful structural insights even for ‘difficult’ proteins that do not crystallize or dissolve. Taking advantage of >1.2 million 13C chemical shifts in the Biological Magnetic Resonance (data)Bank, we have determined the characteristic chemical shift ranges of α-helix, β-sheet and random-coil of the 20 canonical amino acids. This required identification and removal of ~14% incorrectly referenced or otherwise compromised datasets. The purged data reveal unusual structural features, such as distinct chemical shifts associated with “left-handed helix” conformations.
Many discussions of thermodynamics in undergraduate and even advanced textbooks are seriously lacking, with poorly considered concepts and even incorrect equations. We have started to set right some of these problems [J. Chem. Educ., 2014], showing that system work should generally be calculated as the volume integral of system pressure (rather than pext), and pointing out the implicit assumption of uniform pressure and temperature (uPT) processes underlying many equations in thermodynamics.
We have also provided the first explanation of the exothermicity of all combustion reactions, showing why fire is hot regardless of the composition of the fuel and predicting the heat of combustion (-418 kJ/mol times the number of moles of O2) from the elemental composition with an uncertainty of only a few percent. Our analysis overturns some conventional notions of the energetics of molecules, for instance that the heat of combustion arises from breaking high-energy bonds in the fuel, and reveals that CO2 has essentially the same total bond energy as CH4. [J. Chem. Educ., 2015]