Espectroscopia Mössbauer foi usada para investigar as propriedades magnéticas dos compostos antiferromagnéticos CO1-xFexCO3 com x= 0,01, 0,05, 0,1 e 0,3 à temperatura de 4,2 °K, abaixo da temperatura de Néel. As interações quadripolares medidas à temperatira ambiente, acima da temperatira de Néel, e a 4,2 °K mostram pouca dependência com a concentração. / Mössbauer spectroscopy has been used to investigated the magnetic properties of the antiferromagnetics compounds CO1-xFexCO3 com x= 0,01, 0,05, 0,1 and 0,3 at the temperature of 4,2 °K, below the Néel temperature. The quadrupole interaction measured at room temperature, above the Néel temperature, and at 4,2 °K show little concentration dependence.

Amostras de aço AISI H13 foram nitretadas utilizando uma fonte DC por 5 horas, em atmosfera de 80%N2/20%H2, em duas densidades de corrente (alta e baixa) e em diferentes temperaturas. As fases formadas nas amostras nitretadas foram correlacionadas com o comportamento do coeficiente de atrito, do desgaste e da dureza. As intensidades de emissão luminescente relacionadas aos íons N2+ e moléculas N2 no plasma foram medidas e observou-se que são maiores nas condições em que as densidades de corrente também são maiores. Como esperado de resultados da literatura, quanto maior a temperatura de nitretação maior é a espessura das camadas nitretadas. Foi observado que em temperaturas iguais, maiores densidades de corrente provocam a formação de camadas de compostos mais espessas. As amostras foram analisadas usando difração de raios X por incidência rasante, espectroscopia Mössbauer de conversão de elétrons, análise por reação nuclear, espectroscopia de emissão óptica por descarga luminescente, microscopia eletrônica de varredura, análise por microssonda eletrônica e perfilometria. As amostras foram testadas em um ultramicrodurômetro e em um tribômetro do tipo pino-sobre-disco. Análises Rietveld foram aplicadas aos difratogramas do material massivo e da zona de difusão para discriminar entre as fases ferrita e martensita. A composição de fases das amostras nitretadas utilizando diferentes parâmetros mostraram a presença de diferentes fases contendo nitrogênio (e, y' e martensita ∝'-Fe(N)) e carbono (martensita ∝' -Fe(C) e cementita). Na maior temperatura usada neste trabalho, a quantidade relativa das fases e e y' mostrou dependência da densidade de corrente. A correlação entre o conteúdo das fases e as propriedades mecânicas e tribológicas mostrou que nas amostras onde a segregação de carbono foi associada com a formação de cementita, o coeficiente de fricção foi menor do que nas regiões onde as fases e ,y' e martensita foram observadas. Nas amostras com maior conteúdo de martensita e menos e e y', o desgaste foi menor. A distribuição em profundidade da dureza e do conteúdo de martensita y' Fe(N) apresentaram o mesmo comportamento. / AISI H13 steel samples were nitrided using a DC source for 5 hours, under 80% N2/20% H2 atmosphere, at two current densities (high and low), and different temperatures. The phase content on the plasma nitrided samples was correlated to coefficient of friction, wear, and hardness. The plasma emissions were measured and it was observed that the intensity of N2+ ions and N2 molecules was higher at high current density. As expected from previous results in the literature, higher nitriding temperatures resulted in thicker layers. It was observed that at the same temperatures, higher current densities induced thicker compound layers as well. The samples were analysed with grazing incidence x-ray diffraction, conversion electron Mössbauer spectroscopy, nuclear reaction analysis, glow discharge optical emission spectroscopy, scanning electron microscopy, electron microprobe analysis, and perfilometry, and were tested in an ultramicrohardness tester and on a pin-on-disc tribometer. Rietveld analysis was applied to discriminate between ferrite and martensite in the diffraction patterns of the bulk samples and of the diffusion zone. The phase composition of the nitrided samples with different parameters showed the presence of different phases containing nitrogen (e, y', and martensite) and cementite. At the highest temperature used in this work, the relative amount of the phases e and y'showed a dependency on the current density. The correlation of the phase content with the mechanical and tribological properties showed that in samples with carbon segregation and formation of cementite, the friction coefficient was lower than in in regions where the phases e, y', and martensite were observed. In the samples with higher martensite content and less e,y' the mechanical wear was lower. Hardness presented the same trend as the content of martensite ∝'-Fe(N) in the depth profile.

Apresentamos um modelo teórico para explicar as tran sições de valéncia do Sm nas ligas Smi _ x M xS, onde M é um metal de transição (por ex. Y ou La), como uma função da concentração x. O sistema é descrito por um modelo de duas bandas: uma banda de largura nula e de energia E0 (o nível 4f do Sm) e uma banda de condução, a qual é tratada na aproximação do potencial coerente (CPA). Os elétrons 4f interagem entre si via uma repulsão coulombiana finita U, que separa a banda f em duas sub-bandas de energias E0 e Eo+U. Consi dera-se que as bandas f e de condução se hibridizam, sendo V o parâmetro de hibridização. Com este mo delo é possível obter uma transição continua de um estado funda mental não-magnético a um estado não magnético ou magnético, con forme se varia a razão U/V. A variação de valência como uma função da pressão nos monocalcogenetos de Sm (x = O) é também calculada. O modelo usa do é uma extensão do descrito acima, ao qual nOs incluimos uma repulsão coulombiana G entre elétrons f e de condução, mantendo finito o valor de U. Para diferentes valores de G/V obtemos tran sições de primeira ou de segunda ordem de um estado semicondutor a um estado metálico. A resistividade elétrica em ambas as fases é também calculada como uma função da temperatura e os re sultados teoricos obtidos estão em boa concordância com os experimentais. / We present a theoretical model to explain the valence transitions of Sm in Sm1-xMxS alloys, where M is a transition metal (e.g. Y or La) as a function of the concentration x. The system is described by a two-band model: a zero width band of energy E0 (the 4f-level of Sm) and a conduction band, which is treated in the coherent potencial approximation (CPA). The 4felectrons interact between them via a finite intra-atomic coulomb repulsion U, which splits the f-band in two sub-bands at energies E0 and Eo+U. The f -and conduction bands hybridize, being V the hybridization parameter. Within this model it is possible to obtain a continuous valence transition from a non- -magnetic to a non-magnetic or a magnetic ground-state, varying the ratio U/V. The change of valence as a function of pressure in the Sm monochalcogenides (x =0) is also computed. The model used is an extension of the one described above, in which we include a Coulomb repulsion G between f - and conduction electrons, preserving the finite value of U. For different values of G/V we obtain first or second order transitions from a semiconducting to a metallic state. The electrical resistivity in both fases as a function of temperature is also computed and the theoretical results obtained are in good agreement with the experimental ones.

Amostras de aço AISI H13 foram nitretadas utilizando uma fonte DC por 5 horas, em atmosfera de 80%N2/20%H2, em duas densidades de corrente (alta e baixa) e em diferentes temperaturas. As fases formadas nas amostras nitretadas foram correlacionadas com o comportamento do coeficiente de atrito, do desgaste e da dureza. As intensidades de emissão luminescente relacionadas aos íons N2+ e moléculas N2 no plasma foram medidas e observou-se que são maiores nas condições em que as densidades de corrente também são maiores. Como esperado de resultados da literatura, quanto maior a temperatura de nitretação maior é a espessura das camadas nitretadas. Foi observado que em temperaturas iguais, maiores densidades de corrente provocam a formação de camadas de compostos mais espessas. As amostras foram analisadas usando difração de raios X por incidência rasante, espectroscopia Mössbauer de conversão de elétrons, análise por reação nuclear, espectroscopia de emissão óptica por descarga luminescente, microscopia eletrônica de varredura, análise por microssonda eletrônica e perfilometria. As amostras foram testadas em um ultramicrodurômetro e em um tribômetro do tipo pino-sobre-disco. Análises Rietveld foram aplicadas aos difratogramas do material massivo e da zona de difusão para discriminar entre as fases ferrita e martensita. A composição de fases das amostras nitretadas utilizando diferentes parâmetros mostraram a presença de diferentes fases contendo nitrogênio (e, y' e martensita ∝'-Fe(N)) e carbono (martensita ∝' -Fe(C) e cementita). Na maior temperatura usada neste trabalho, a quantidade relativa das fases e e y' mostrou dependência da densidade de corrente. A correlação entre o conteúdo das fases e as propriedades mecânicas e tribológicas mostrou que nas amostras onde a segregação de carbono foi associada com a formação de cementita, o coeficiente de fricção foi menor do que nas regiões onde as fases e ,y' e martensita foram observadas. Nas amostras com maior conteúdo de martensita e menos e e y', o desgaste foi menor. A distribuição em profundidade da dureza e do conteúdo de martensita y' Fe(N) apresentaram o mesmo comportamento. / AISI H13 steel samples were nitrided using a DC source for 5 hours, under 80% N2/20% H2 atmosphere, at two current densities (high and low), and different temperatures. The phase content on the plasma nitrided samples was correlated to coefficient of friction, wear, and hardness. The plasma emissions were measured and it was observed that the intensity of N2+ ions and N2 molecules was higher at high current density. As expected from previous results in the literature, higher nitriding temperatures resulted in thicker layers. It was observed that at the same temperatures, higher current densities induced thicker compound layers as well. The samples were analysed with grazing incidence x-ray diffraction, conversion electron Mössbauer spectroscopy, nuclear reaction analysis, glow discharge optical emission spectroscopy, scanning electron microscopy, electron microprobe analysis, and perfilometry, and were tested in an ultramicrohardness tester and on a pin-on-disc tribometer. Rietveld analysis was applied to discriminate between ferrite and martensite in the diffraction patterns of the bulk samples and of the diffusion zone. The phase composition of the nitrided samples with different parameters showed the presence of different phases containing nitrogen (e, y', and martensite) and cementite. At the highest temperature used in this work, the relative amount of the phases e and y'showed a dependency on the current density. The correlation of the phase content with the mechanical and tribological properties showed that in samples with carbon segregation and formation of cementite, the friction coefficient was lower than in in regions where the phases e, y', and martensite were observed. In the samples with higher martensite content and less e,y' the mechanical wear was lower. Hardness presented the same trend as the content of martensite ∝'-Fe(N) in the depth profile.

No description available.

This thesis describes the synthesis and characterisation of metal complexes that contain chemically reduced bipyridyl ligands. The crystal structures of twenty-six novel complexes are reported alongside detailed discussions on the electronic and spectroscopic effects and trends associated with the different oxidation states within these species. Chapter One introduces the isomers of bipyridine and their redox chemistry, the concept of non-innocent ligands and the spectroscopic techniques that are currently used in determining ligand oxidation states. Subsequently, examples of main group, transition metal, lanthanide and actinide species that contain or may contain reduced forms of the ligand are discussed. Chapter Two details the synthesis and structural characterisation of alkali metal salts of singly and doubly reduced forms of the three commercially available bipyridine isomers. The effects of this reduction are investigated with the aim of developing diagnostic fingerprints for each of the ligand oxidation states. Chapter Three discusses the synthesis of an homologous series of compounds of the form [M(2,2'-bipy)(mes)₂]^n– where M = Cr, Mn, Fe, Co, Ni and n = 0, 1. Trends in magnetism, bonding and electronic structure are investigated with reference to theoretical calculations and the diagnostic fingerprints identified in the previous chapter. Chapter Four describes the synthesis and characterisation of three compounds containing the isostructural motif [Zn₂(4,4'-bipy)(mes)₄]^n– where n = 0, 1, 2. Structural and spectroscopic changes are discussed and com- pared to theoretical calculations. Chapter Five contains descriptions of the spectroscopic techniques employed in the above research and synthesis routes to all compounds featured in this thesis.

Des films de nitrure de zinc et d’étain (ZnSnN2) ont été élaborés par co-pulvérisation magnétron réactive à des températures proches de l’ambiante. La composition chimique des revêtements a été optimisée en ajustant les paramètres de dépôt comme la tension appliquée aux cibles métalliques, la pression de travail et la composition du gaz plasmagène. Dans les conditions optimisées, les films sont fortement cristallisés sur les différents types de substrats testés. Une étude approfondie sur la structure des films a été entreprise par microscopie électronique en transmission. Nous avons ainsi pu montrer que nos films de ZnSnN2 cristallisent dans le système hexagonal. Toutefois, cette structure diffère de celles présentées dans la littérature pour le nitrure de zinc et d’étain. Des études sur l’environnement chimique des éléments constitutifs des revêtements ont également été menées par spectrométrie Mössbauer et par photoémission X. Elles montrent que l’étain est présent dans nos films sous forme de Sn4+ en configuration tétraédrique. Nous avons également pu montrer que l’oxygène présent dans nos films est principalement localisé dans les zones inter-colonnaires. Enfin, les propriétés optiques et électriques de nos films ont été estimées en fonction de leur composition chimique. L’ensemble des résultats obtenus durant ce travail démontre la pertinence de ZnSnN2 pour des applications futures en tant que couche absorbante dans les cellules photovoltaïques / Zinc tin nitride (ZnSnN2) thin films have been deposited by reactive magnetron co-sputtering at room temperature. The stoichiometry of the films has been controlled by optimizing the deposition conditions such as the voltage applied to the metallic targets, the deposition pressure and the composition of the gas mixture. By using the optimized parameters, the deposited films are highly crystallized on the different used substrates. A special attention has been devoted to the determination of the film structure. Among the various structures reported in the literature, we have shown by transmission electron microscopy that the films crystallised in a hexagonal structure. Nevertheless, the structure of our films does not fit with that reported in the literature for the hexagonal ZnSnN2 material. In addition to this structural study, we have performed fine characterization using conversion electron Mossbauer spectrometry and X-ray photoemission spectroscopy. Both methods show that the optimized films contain Sn4+ ions in tetrahedral configuration. Nevertheless, oxygen contamination at the column boundaries has been evidenced. The electrical and optical properties of the films have been determined has a function of the film composition. The results obtained in this PhD work clearly evidence that ZnSnN2 is a suitable material for photovoltaic applications

This thesis deals with the studies of magnetization and electrical transport properties of iron nanoparticles embedded in the carbon matrix. The synthesis and characteristics of the nanoparticle systems studied, are also presented. Carbon-iron (C-Fe) based systems are of growing interest due to their improved magnetic properties as well as in their potential application as sensors, catalysts, and in various other applications. In particular, nanocomposites of iron carbide, such as the cementite phase Fe3C, are further suited to diverse technological exploitations due to their enhanced mechanical properties and importance in ferrous metallurgy. The recent interest in magnetic nanostructures lies in the emergence of novel magnetic and transport properties with the reduction of size. As the dimension approaches the nanometer length scale, interesting size-dependent properties like enhanced coercivity, enhanced magnetic moment, super paramagnetism etc. are seen. Thermal assisted chemical vapour deposition (CVD) is used to decompose and chemically react the introduced precursors, maleic anhydride and ferrocene. This method provides relative size control over the individual particles by varying C/Fe concentration in precursors and the pyrolysis temperature during the co-deposition process. Ferrocene has been used actively for the production of nanoparticle composites and in the production of nanostructured carbon. The temperature of preparation, reaction rate, and the time duration of annealing directly effects the nanoparticle compositions. The catalytic effect of transitional elements are well documented in literature. This thesis is an effort to understand the growth of ferromagnetic nanocrystallites in carbon matrix, which undergo partial graphitization due to the catalytic effect of transitional elements. The effect of transitional metal on the degree of graphitization of the carbon matrix, morphology of the nanoparticle and the carbon matrix are studied. The phase of the ferromagnetic iron nanoparticles and the structural investigation forms part of the study. Here X-Ray diffraction (XRD) is employed to study the presence of different phases of iron in the partially graphitized carbon matrix. The matrix morphology and the particle size distribution were studied using Transmission Electron Microscopy (TEM) and High-Resolution TEM (HRTEM). The ferromagnetic states of the iron nanoparticles are investigated using Mossbauer spectroscopy. The results from these studies, are used to correlated the macroscopic properties to the microscopic studies. The enhanced magnetization, coercivity and the temperature dependence of the magnetization value is understood within the frame work of ferromagnetic Bloch law and surrounding carbon spins. The logarithmic temperature dependence of conductivity of the nanoparticle composites is analyzed in the framework of interference models as well as the many-body Kondo interaction effect. This thesis contains seven chapters: In chapter 1, a brief introduction to mesoscopic physics and the size-dependent phenomenon are given. Special attention is paid to magnetic nanoparticle and its composites, and the various finite-size effects exhibited by them are discussed in detail. The relevance of carbon matrix and its importance on the growth of iron nanoparticles with high thermal stability is also discussed. The ballistic and diffusive transport phenomena observed in low-dimensional systems are briefly discussed. The interplay of localization and various interaction effects at nanoscale are examined. In disordered metals the low temperature conductivity is dominated by the interference effects. A brief discussion is made on the conductivity in disorder systems, with the presence of magnetic impurities and how the classic many-body Kondo problem, is effected by various interactions. Chapter 2, mainly deals with the experimental techniques employed in the thesis. The thermal-assisted chemical vapour deposition setup used to decompose and chemically react the introduced organometallic precursors, for the preparation of C:Fe composites are discussed and its advantage over other preparation methods are emphasized. The method is optimized to provide relative size control over the nanoparticles composites and the phase compositions by varying C/Fe concentration in precursors and the pyrolysis temperature, during the co-deposition process. The various structural characterization tools used in the present study are summed up concisely in this chapter. The SQUID magnetometer system; its working principle and the various protocol used for the low temperature magnetization measurements are elaborated. Further, details regarding superconducting magnetic cryostat, utilized for the low temperature conductivity and magneto resistance measurements, are discussed. Films of C:Fe composites are grown on substrates to study the effect of disorder and sample size on the conductivity behaviour of the composites at low temperature. Chapter 3, presents the outcome of the structural studies undertaken on the C:Fe composites using XRD, TEM, and HRTEM. X-ray diffraction measurements performed on the powder composites reveal that, in addition to the presence of sharp diffraction peak from nanographite, peaks corresponding to the different phases of Fe are also seen. The effect of preparation temperature on the matrix morphology is revealed from the estimation of degree of graphitization. Iron carbide is the predominant phase in all the prepared composites. For low concentration of iron, iron carbide alone is present but as the percentage of iron in the samples increased other phases of iron are also seen. The microscopic studies on the prepared compositions revealed the presence of nanosized iron particles well embedded in the partially graphitized matrix. Here again, with the increase in iron percentage, agglomeration of ferromagnetic nanoparticles are seen. The kinetics of the particle growth and the filamentous nature of the carbon matrix are also discussed. Mossbauer investigation on C:Fe composites are presented in chapter 4. The measurements revealed the iron atom occupation in the crystal lattice. In the lower Fe concentration samples, the room temperature Mossbauer spectrum revealed the presence of sextet from Fe3C (cementite) phase. As the percentage of iron increased, sextet from α-Fe, Fe3O4 are also seen in some of the prepared compositions. Effect of carbon atoms on the structure and magnetic properties of the nanoparticle species are obvious from the isomer shift measurements. Chapter 5 comprises of the various magnetic properties and interactions present in small particle system such as magnetic anisotropy, coercivity, enhanced magnetization, inter-and intra-particle interactions etc. Magnetization measurements carried out in SQUID magnetometer on the C:Fe composites and carbon flakes (prepared from organic precursor, maleic anhydride alone) are presented. The enhanced magnetic properties of the nanoparticle assembly is discussed in detail. The hysteresis loops trace, with a finite coercivity at room temperature, indicates the ferromagnetic nature of the samples. At room temperature the magnetization value saturates at high magnetic field, indicating negligible effect from super paramagnetic particles on the hysteresis loop. The squareness ratio, saturation magnetization, coercivity and remanence magnetization values are analyzed in detail. The temperature dependence of magnetization shows a combination of Bloch law and Curie-Weiss behaviour, consistent with the picture of ferromagnetic clusters embedded in a carbon matrix. The Bloch’s constant is found to be larger by an order of magnitude compared to the bulk value, implying stronger dependence of magnetization with temperature. Effort to understand the enhanced magnetic moment in the light of magnetism in carbon was taken up. The proximity effect of ferromagnetic metal on the carbon and the hydrogen bonding with the dangling bonds, both studied in detail in literature, in connection with the induced magnetic moments in carbon, are invoked. In chapter 6, the different conductivity regimes are identified, to study the conduction mechanisms in composites and films. For the transport measurements pelletized samples are used for the resistivity and magneto resistance measurements. The conductivity data are analyzed based on the interplay of localization and Kondo effect in the ferromagnetic disordered system. In order to understand the effect of disorder and thickness on the Kondo problem, transport measurements are carried on thin films of C:Fe composites grown on quartz and alumina substrate. Disorder induced metal-insulator transition is observed in the prepared samples. The zero-field conductivity and magneto resistance data is fitted to variable range hopping (VRH) in strong localization regime. Chapter 7 summarizes the thesis and presents some perspectives for the future.

[1] Four specimens of zoned tourmaline from granitic pegmatites are characterised in detail, each having unusual compositional and/or morphologic features: (1) a crystal from Black Rapids Glacier, Alaska, showing a central pink zone of elbaite mantled by a thin rim of green liddicoatite; (2) a large (~25 cm) slab of Madagascar liddicoatite cut along (001) showing complex patterns of oscillatory zoning; and (3) a wheatsheaf and (4) a mushroom elbaite from Mogok, Myanmar, both showing extensive bifurcation of fibrous crystals originating from a central core crystal, and showing pronounced discontinuous colour zoning. Crystal chemistry and crystal structure of these samples are characterised by SREF, EMPA, and 11B and 27Al MAS NMR and Mössbauer spectroscopies. For each sample, compositional change, as a function of crystal growth, is characterised by EMPA traverses, and the total chemical variation is reduced to a series of linear substitution mechanisms. Of particular interest are substitutions accommodating the variation in [4]B: (1) TB + YAl ↔ TSi + Y(Fe, Mn)2+, where transition metals are present, and (2) TB2 + YAl ↔ TSi2 + YLi, where transition metals are absent. Integration of all data sets delineates constraints on melt evolution and crystal growth mechanisms. [2] Uncertainty has surrounded the occurrence of [4]Al and [4]B at the T-site in tourmaline, because B is difficult to quantify by EMPA and Al is typically assigned to the octahedral Y- and Z-sites. Although both [4]Al and [4]B have been shown to occur in natural tourmalines, it is not currently known how common these substituents are. Using 11B and 27Al MAS NMR spectroscopy, the presence of [4]B and [4]Al is determined in fifty inclusion-free tourmalines of low transition-metal content with compositions corresponding to five different species. Chemical shifts of [4]B and [3]B in 11B spectra, and [4]Al and [6]Al in 27Al spectra, are well-resolved, allowing detection of very small (< ~0.1 apfu) amounts of T-site constituents. Results show that contents of 0.0 < [4]B, [4]Al < 0.5 apfu are common in tourmalines containing low amounts of paramagnetic species, and that all combinations of Si, Al and B occur in natural tourmalines.

[1] Four specimens of zoned tourmaline from granitic pegmatites are characterised in detail, each having unusual compositional and/or morphologic features: (1) a crystal from Black Rapids Glacier, Alaska, showing a central pink zone of elbaite mantled by a thin rim of green liddicoatite; (2) a large (~25 cm) slab of Madagascar liddicoatite cut along (001) showing complex patterns of oscillatory zoning; and (3) a wheatsheaf and (4) a mushroom elbaite from Mogok, Myanmar, both showing extensive bifurcation of fibrous crystals originating from a central core crystal, and showing pronounced discontinuous colour zoning. Crystal chemistry and crystal structure of these samples are characterised by SREF, EMPA, and 11B and 27Al MAS NMR and Mössbauer spectroscopies. For each sample, compositional change, as a function of crystal growth, is characterised by EMPA traverses, and the total chemical variation is reduced to a series of linear substitution mechanisms. Of particular interest are substitutions accommodating the variation in [4]B: (1) TB + YAl ↔ TSi + Y(Fe, Mn)2+, where transition metals are present, and (2) TB2 + YAl ↔ TSi2 + YLi, where transition metals are absent. Integration of all data sets delineates constraints on melt evolution and crystal growth mechanisms. [2] Uncertainty has surrounded the occurrence of [4]Al and [4]B at the T-site in tourmaline, because B is difficult to quantify by EMPA and Al is typically assigned to the octahedral Y- and Z-sites. Although both [4]Al and [4]B have been shown to occur in natural tourmalines, it is not currently known how common these substituents are. Using 11B and 27Al MAS NMR spectroscopy, the presence of [4]B and [4]Al is determined in fifty inclusion-free tourmalines of low transition-metal content with compositions corresponding to five different species. Chemical shifts of [4]B and [3]B in 11B spectra, and [4]Al and [6]Al in 27Al spectra, are well-resolved, allowing detection of very small (< ~0.1 apfu) amounts of T-site constituents. Results show that contents of 0.0 < [4]B, [4]Al < 0.5 apfu are common in tourmalines containing low amounts of paramagnetic species, and that all combinations of Si, Al and B occur in natural tourmalines.