Multiple-quantum NMR study of clustering in hydrogenated amorphous silicon

Dynamic Nuclear Polarization Enhanced Multiple-Quantum Spin Counting of Molecular Assemblies in Vitrified Solutions

Crystallization pathways are essential to various industrial, geological, and biological processes. In nonclassical nucleation theory, prenucleation clusters (PNCs) form, aggregate, and crystallize to produce higher order assemblies. Microscopy and X-ray techniques have limited utility for PNC analysis due to the small size (0.5-3 nm) and time stability constraints. We present a new approach for analyzing PNC formation based on 31P nuclear magnetic resonance (NMR) spin counting of vitrified molecular assemblies. The use of glassing agents ensures that vitrification generates amorphous aqueous samples and offers conditions for performing dynamic nuclear polarization (DNP)-amplified NMR spectroscopy. We demonstrate that molecular adenosine triphosphate along with crystalline, amorphous, and clustered calcium phosphate materials formed via a nonclassical growth pathway can be differentiated from one another by the number of dipolar coupled 31P spins. We also present an innovative approach for examining spin counting data, demonstrating that a knowledge-based fitting of integer multiples of cosine wave functions, instead of the traditional Fourier transform, provides a more physically meaningful retrieval of the existing frequencies. This is the first report of multiquantum spin counting of assemblies formed in solution as captured under vitrified DNP conditions, which can be useful for future analysis of PNCs and other aqueous molecular clusters.

An analysis of double-quantum coherence ESR in an N-spin system: Analytical expressions and predictions

Electron spin resonance pulsed dipolar spectroscopy (PDS) has become popular in protein 3D structure analysis. PDS studies yield distance distributions between a pair or multiple pairs of spin probes attached to protein molecules, which can be used directly in structural studies or as constraints in theoretical predictions. Double-quantum coherence (DQC) is a highly sensitive and accurate PDS technique to study protein structures in the solid state and under physiologically relevant conditions. In this work, we have derived analytical expressions for the DQC signal for a system with N-dipolar coupled spin-1/2 particles in the solid state. The expressions are integrated over the relevant spatial parameters to obtain closed form DQC signal expressions. These expressions contain the concentration-dependent "instantaneous diffusion" and the background signal. For micromolar and lower concentrations, these effects are negligible. An approximate analysis is provided for cases of finite pulses. The expressions obtained in this work should improve the analysis of DQC experimental data significantly, and the analytical approach could be extended easily to a wide range of magnetic resonance phenomena.

DNP-NMR of surface hydrogen on silicon microparticles

Dynamic nuclear polarization (DNP) enhanced nuclear magnetic resonance (NMR) offers a promising route to studying local atomic environments at the surface of both crystalline and amorphous materials. We take advantage of unpaired electrons due to defects close to the surface of the silicon microparticles to hyperpolarize adjacent ¹H nuclei. At 3.3 T and 4.2 K, we observe the presence of two proton peaks, each with a linewidth on the order of 5 kHz. Echo experiments indicate a homogeneous linewidth of ∼150-300 Hz for both peaks, indicative of a sparse distribution of protons in both environments. The high frequency peak at 10 ppm lies within the typical chemical shift range for proton NMR, and was found to be relatively stable over repeated measurements. The low frequency peak was found to vary in position between -19 and -37 ppm, well outside the range of typical proton NMR shifts, and indicative of a high-degree of chemical shielding. The low frequency peak was also found to vary significantly in intensity across different experimental runs, suggesting a weakly-bound species. These results suggest that the hydrogen is located in two distinct microscopic environments on the surface of these Si particles.

Nanoscale structure of microvoids in a-Si:H: a first-principles study

In this paper, we have studied the shape, size, and number density of atomic microvoids in hydrogenated amorphous silicon (a-Si:H). By jointly employing experimental infrared data and ab initio simulations, we propose a simple and effective hydrogenation scheme, which is capable of producing large atomistic models of a-Si:H for studying microvoids. Our results suggest that hydrogen atoms in the networks are distributed in sparse (or isolated) and clustered environments. For a-Si:H models with 9-14 at.% hydrogen, we find approximately 3-4 at.% of total hydrogen atoms are distributed in the isolated phase. The density of the clustered phase is found to be between 6-12 at.%, which appears to depend on the amount of hydrogen in the network. The calculation of radii of gyration of atomic microvoids shows that the diameter of the microvoids is distributed from 6 Å to 12 Å. A few hydrogen molecules have also been observed to form inside the microvoids in our study, the concentration of which is about 1 at.% relative to silicon atoms. A comparison of our results with those from small-angle x-ray scattering (SAXS), infrared (IR) absorption, nuclear magnetic resonance (NMR) and calorimetric studies are presented.

Metastable defect formation at microvoids identified as a source of light-induced degradation in a-Si:H

Light-induced degradation of hydrogenated amorphous silicon (a-Si:H), known as the Staebler-Wronski effect, has been studied by time-domain pulsed electron-paramagnetic resonance. Electron-spin echo relaxation measurements in the annealed and light-soaked state revealed two types of defects (termed type I and II), which can be discerned by their electron-spin echo relaxation. Type I exhibits a monoexponential decay related to indirect flip-flop processes between dipolar coupled electron spins in defect clusters, while the phase relaxation of type II is dominated by 1H nuclear spin dynamics and is indicative for isolated spins. We propose that defects are either located at internal surfaces of microvoids (type I) or are isolated and uniformly distributed in the bulk (type II). The concentration of both defect type I and II is significantly higher in the light-soaked state compared to the annealed state. Our results indicate that in addition to isolated defects, defects on internal surfaces of microvoids play a role in light-induced degradation of device-quality a-Si:H.

A statistical approach for analyzing the development of 1H multiple-quantum coherence in solids

A novel statistical approach for analyzing (1)H multiple-quantum (MQ) spin dynamics in so-called spin-counting solid-state NMR experiments is presented. The statistical approach is based on the percolation theory with Monte Carlo methods and is examined by applying it to the experimental results of three solid samples having unique hydrogen arrangement for 1-3 dimensions: the n-alkane/d-urea inclusion complex as a one-dimensional (1D) system, whose (1)H nuclei align approximately in 1D, and magnesium hydroxide and adamantane as a two-dimensional (2D) and a three-dimensional (3D) system, respectively. Four lattice models, linear, honeycomb, square and cubic, are used to represent the (1)H arrangement of the three samples. It is shown that the MQ dynamics in adamantane is consistent with that calculated using the cubic lattice and that in Mg(OH)2 with that calculated using the honeycomb and the square lattices. For n-C20H42/d-urea, these 4 lattice models fail to express its result. It is shown that a more realistic model representing the (1)H arrangement of n-C20H42/d-urea can describe the result. The present approach can thus be used to determine (1)H arrangement in solids.

A study of hydrogen microstructure in amorphous silicon via inversion of nuclear magnetic resonance spectra

We present an inverse approach for studying hydrogen microstructure in amorphous silicon. The approach consists of generating a prior distribution (of spins/hydrogen) by inverting experimental nuclear magnetic resonance (NMR) data, which is subsequently superimposed on a network of amorphous silicon. The resulting network is then relaxed using a total-energy functional to obtain a stable, low-energy configuration such that the initial spin distribution is minimally perturbed. The efficacy of this approach is demonstrated by generating model configurations that not only have the correct NMR spectra but also satisfy simultaneously experimental structural, electronic and vibrational properties of hydrogenated amorphous silicon.

Hydrogen Equilibration and Metastability in Amorphous Silicon

We review evidence for the vital role of H in metastability and defect equilibration in hydrogenated amorphous silicon. H pairs act both as an H reservoir that equilibrates with DB's and dopants and as the metastable H state of light-induced metastability. With ab initio pseudopotential theory, we calculate energies and configurations for two novel H pairing sites associated with small hydrogenated vacancies. We describe microscopic models of equilibration and metastability based on our calculations.

Hydrogen Microstructure in Amorphous Semiconductors

The effects of deposition parameters on the H microstructure of plasma deposited amorphous silicon (a-Si:H) and amorphous silicon carbide (a-SiC:H) are measured via multiple quantum nuclear magnetic resonance (MQ NMR). These studies indicate clusters of 5 to 7H atoms exist in a-Si:H films prepared at temperatures ranging from 113 to 324°C. In the range from 270 to 324°C, only these small clusters exist, but lower temperature films also contain larger clusters. Annealing studies indicate H rearranges in a-Si:H prior to evolution. Deposition temperature and annealing temperature have similar effects on H concentration in a-Si:H, but deposition temperature control the density and microstructure of the film. The addition of dopant atoms also affects the H microstructure, with phosphorous causing larger H clusters to form, and boron reducing clustering in a-Si:H films. This perturbation of the film's microstructure suggests that the effects of dopant addition are more complex in amorphous than in crystalline semiconductors. The concentration of carbon atoms also effects H microstructure of a-SiC:H in a complex way. We surmise that H microstructure, rather than H content, determines amorphous semiconductors properties.

Carbon Local Bonding Configurations in Amorphous Hydrogenated Silicon-Carbon Alloys

We present the results of a carbon-13 nuclear magnetic resonance (NMR) study of well-characterized thin films of amorphous hydrogenated silicon carbide. The NMR data detail the distribution of carbon local bonding configurations in films which have carbon-to-silicon ratios less than one. In particular, we show data which clearly identify and quantify non-hydrogenated sp2, or unsaturated, carbon bonding environments.

Vacancies, microstructure and the moments of nuclear magnetic resonance: the case of hydrogenated amorphous silicon

Recent experiments on hydrogenated amorphous silicon using infrared absorption spectroscopy have indicated the presence of mono- and divacancies in samples for concentrations of up to 14% hydrogen. Motivated by this observation, we study the microstructure of hydrogen in two model networks of hydrogen-rich amorphous silicon with particular emphasis on the nature of the distribution (of hydrogen), the presence of defects and the characteristic features of the nuclear magnetic resonance spectra at low and high concentrations of hydrogen. Our study reveals the presence of vacancies, which are the built-in features of the model networks. The study also confirms the presence of various hydride configurations in the networks, from silicon monohydrides and dihydrides to open chain-like structures, that have been observed in the infrared and nuclear magnetic resonance experiments. The broad and the narrow line widths of the nuclear magnetic resonance spectra are calculated from a knowledge of the distribution of spins (hydrogen) in the networks.

The Staebler-Wronski Effect: New Physical Approaches and Insights as a Route to Reveal its Origin

In the recent years more and more theoretical and experimental evidence have been found that the hydrogen bonded to silicon in dense hydrogenated amorphous silicon (a-Si:H) predominantly resides in hydrogenated divacancies. In this contribution we will philosophize about the option that the small fraction of divacancies, missing at least one of its bonded hydrogen, may correspond to some of the native and metastable defect states of a-Si:H. We will discuss that such defect entities are an interesting basis for new and alternative views on the origin of the SWE.

Nuclear hyperpolarization in solids and the prospects for nuclear spintronics

Nuclear hyperpolarization can be achieved in a number of ways. This article focuses on the use of coupling of nuclei to (nearly) pure quantum states, with particular emphasis on those states obtained by optical excitation in bulk semiconductors. I seek an answer to this question: "What is to prevent the design and analysis of nuclear spintronics devices that use the extremely long-lived hyperpolarized nuclear spin states, and their weak couplings to each other, to affect computation, memory, or informational technology schemes?" The answer, I argue, is in part because there remains a lack of fundamental understanding of how to generate and control nuclear polarization with schemes other than with rf coils.

Materials modeling by design: applications to amorphous solids

In this paper, we review a host of methods used to model amorphous materials. We particularly describe methods which impose constraints on the models to ensure that the final model meets a priori requirements (on structure, topology, chemical order, etc). In particular, we review work based on quench from the melt simulations, the 'decorate and relax' method, which is shown to be a reliable scheme for forming models of certain binary glasses. A 'building block' approach is also suggested and yields a pleading model for GeSe(1.5). We also report on the nature of vulcanization in an Se network cross-linked by As, and indicate how introducing H into an a-Si network develops into a-Si:H. We also discuss explicitly constrained methods including reverse Monte Carlo (RMC) and a novel method called 'Experimentally Constrained Molecular Relaxation'. The latter merges the power of ab initio simulation with the ability to impose external information associated with RMC.

Constants of motion in NMR spectroscopy

We present a general method for constructing a subset of the constants of motion in terms of products of spin operators. These operators are then used to give insight into the multi-spin orders comprising the quasi-equilibrium state formed under a Jeener-Broekaert sequence in small, dipolar-coupled, spin systems. We further show that constants of motion that represent single-quantum coherences are present due to the symmetry of the dipolar Hamiltonian under 180 degrees spin rotations, and that such coherences contribute a DC component to the FID which vanishes in the absence of the flip-flop terms and is only present for spin clusters with an odd number of spins.

Multiple-quantum NMR spin dynamics of inhomogeneous one-dimensional systems in solids

Multiple-quantum NMR spin dynamics of inhomogeneous one-dimensional systems in solids is investigated by analytical and numerical methods. A fermion approach for MQ spin dynamics of one-dimensional inhomogeneous systems is developed in the approximation of the dipole-dipole interactions (DDI) of nearest neighbors. It is shown that only MQ coherences of the zeroth and plus/minus second orders appear in the approximation of the DDI of the nearest neighbors even in inhomogeneous one-dimensional systems. We also investigate MQ dynamics of inhomogeneous chains numerically. Intensities of MQ NMR coherences for a linear chain consisting of 3000 spins are calculated.

Multiple quantum spin counting techniques with quadrupolar nuclei

Phase incremented and continuous irradiation multiple spin correlation methods are applied to spin [Formula: see text] nuclei with small quadrupole couplings such as (7)Li in LiCl and are shown to successfully produce a coherently coupled dipolar spin network. Application to the analogous Na salt shows successful spin correlation evolving at a slower rate due to the weaker homonuclear dipolar coupling strength between Na nuclei. The results are analysed using a statistical approach. Spin counting is non-trivial as not only multiple quantum coherences between spins are generated but also within the quadrupolar spin levels. Na(2)C(2)O(4) is investigated as a material with non-negligible quadrupole coupling and it is in this limit that the spin correlation techniques are found to break down.

Line shapes of multiple quantum NMR coherences in one-dimensional quantum spin chains in solids

General formulae for intensities of multiple quantum (MQ) NMR coherences in systems of nuclear spins coupled by the dipole-dipole interactions are derived. The second moments of the MQ coherences of zero- and second orders are calculated for infinite linear chains in the approximation of the nearest neighbor interactions. Supercomputer simulations of intensities of MQ coherences of linear chains are performed at different times of preparation and evolution periods of MQ NMR experiments. The second moments obtained from the developed theory are compared with the results of the supercomputer analysis of MQ NMR dynamics. The linewidth information in MQ NMR experiments is discussed.

Sensitivity enhancement in multiple-quantum NMR experiments with CPMG detection

We present a modified multiple-quantum (MQ) experiment, which implements the Carr-Purcell-Meiboom-Gill (CPMG) detection scheme in the static MQ NMR experiment proposed by W. S. Warren et al. (1980, J. Chem. Phys.73, 2084-2099) and exploited further by O. N. Antzutkin and R. Tycko (1999, J. Chem. Phys.110, 2749-2752). It is demonstrated that a significant enhancement in the sensitivity can be achieved by acquiring echo trains in the MQ experiments for static powder samples. The modified scheme employing the CPMG detection was superior to the original MQ experiment, in particular for the carbonyl carbon with a very large chemical shift anisotropy.

Diffusion mechanism of hydrogen in amorphous silicon: ab initio molecular dynamics simulation

The mechanism of H migration in amorphous Si has remained an unresolved problem. The main issue is the small activation energy (1.5 eV) relative to the known strength of Si-H bonds (2-3.5 eV). We report first-principles finite-temperature simulations which demonstrate vividly that H is not released spontaneously, as proposed by most models, but awaits the arrival of a floating bond (FB). The "migrating species" is an FB-H complex, with H jumping from Si to Si and the FB literally floating around it. Migration stops when the FB veers away.

Hydrogen above saturation at silicon vacancies: H-pair reservoirs and metastability sites

We propose that hydrogen-passivated multivacancies which appear to be fully saturated with H can actually capture additional H in electrically inactive sites. In silicon, first-principles total energy calculations show that splitting an (m>or=2) multivacancy into a mono- and an (m-1) vacancy provides a low-strain pairing site for H, 0.4 eV per H lower than any known bulk pairing site. This monovacancy ejection mechanism is an excellent candidate for the H reservoir found both in crystalline and amorphous Si. A distinct H pairing on the fully saturated m vacancies, by forming an internal surface Si-Si dimer, provides the final state of light-induced metastable degradation of hydrogenated amorphous silicon.

Biomolecular solid state NMR: advances in structural methodology and applications to peptide and protein fibrils

Solid state nuclear magnetic resonance (NMR) methods can provide atomic-level structural constraints on peptides and proteins in forms that are not amenable to characterization by other high-resolution structural techniques, owing to insolubility, high molecular weight, noncrystallinity, or other characteristics. Important examples include peptide and protein fibrils and membrane-bound peptides and proteins. Recent advances in solid state NMR methodology aimed at structural problems in biological systems are reviewed. The power of these methods is illustrated by experimental results on amyloid fibrils and other protein fibrils.

Selection rules for multiple quantum NMR excitation in solids: derivation from time-reversal symmetry and comparison with simulations and (13)C NMR experiments

New derivations of selection rules for excitation and detection of multiple quantum coherences in coupled spin-1/2 systems are presented. The selection rules apply to experiments in which the effective coupling Hamiltonian used for multiple quantum excitation is both time-reversal invariant and time-reversible by a phase shift of the radiofrequency pulse sequence that generates the effective couplings. The selection rules are shown to be consequences of time-reversal invariance and time-reversibility and otherwise independent of the specific form of the effective coupling Hamiltonian. Numerical simulations of multiple quantum NMR signal amplitudes and experimental multiple quantum excitation spectra are presented for the case of a multiply (13)C-labeled helical polypeptide. The simulations and experiments confirm the selection rules and demonstrate their impact on multiple quantum (13)C NMR spectra in this biochemically relevant case.

Counting spins with a new spin echo double resonance

In traditional spin echo double resonance (SEDOR), the echo amplitude M is decreased when the observed spins S are flipped by pi together with the pi refocusing pulse on the observed spins I; the dependence on tau is then determined. In the new version of SEDOR, the echo amplitude is measured as a function of the S spin flip angle theta at a constant pulse spacing tau. The analysis is simple and powerful for long tau, where the strong collision limit applies. There, the variation of M with theta can be fit, yielding the number n of spins S to which each spin I is coupled. Data from amorphous silicon with 1H and 2D show the described effect. A MAS version of the new method is used on multiply labeled alanine and urea, with results in good agreement with the predictions for n = 2, as expected. By Fourier transforming M with respect to the flip angle theta, a stick spectrum results; the largest numbered non-vanishing stick yields the number n of spins S coupled to each spin I. Simulations are presented for an n = 2 system. The present technique is compared to the multiple-quantum spin-counting method. Copyright 1998 Academic Press.

One- and two-dimensional 1H magic-angle spinning experiments on hydrous silicate glasses

Applications of various one- and two-dimensional 1H magic-angle spinning nuclear magnetic resonance (MAS NMR) techniques for the elucidation of structural properties of hydrous silicate glasses are described. Advantages and limitations of one-dimensional experiments [MAS and combined rotation and multiple-puls spectroscopy (CRAMPS)] and of two-dimensional approaches (spin-exchange experiments, "MAS-CRAMPS" correlation plus extended versions) are discussed. The various 1H MAS NMR techniques are illustrated by practical examples of spectra obtained from a hydrous silicate glass of Na2O.4SiO2.0.7H2O composition.

Multiple-quantum nuclear magnetic resonance spectroscopy of coupled 1/2 spins in solids. Combination with cross-polarization and magic-angle spinning

Proton multiple-quantum coherence spectroscopy has been combined with magic-angle spinning (MAS) and cross-polarization (CP). This enables the detection of the proton (or any other abundant spin) multiple-quantum coherence spectrum via the high-resolution 13C (any other spin) nuclear magnetic resonance (NMR) spectrum. For this purpose multiple-quantum pulse sequences synchronised to sample rotation have been designed, and the average Hamiltonians of these sequences have been analysed. The analysis allows the design of optimal experimental conditions. As a demonstration of the technique, it has been applied to a mixture of adamantane and hexamethylethane. From earlier 13C spin diffusion experiments it was known that these two molecules form a mixed crystal. With our technique we detected two different phases with different molecular translational self-diffusion coefficients in this system.