Front Matter Pages Crystallo-Chemistry of Carbon Intercalation Compounds. Electronic Properties of Graphite Intercalation Compounds. Applications of Intercalation Compounds. Stanley Whittingham, Lawrence B. Back Matter Pages About this book Introduction Materials with layered structures remain an extensively investigated subject in current physics and chemistry.
Most of the promising technological applications however deal with intercalation compounds of layered materials. Graphite intercalation compounds have now been known for a long time. Intercalation in transition metal dichalcogenides, on the other hand, has been investigated only recently. Learn more.
Layered compounds exhibit various properties based on their layered structures. The most typical chemical property is the intercalation reaction, which most layered oxides and hydroxides possess and which is important for the charge—discharge of secondary batteries and the preparation of new layered compounds. Layered materials have been a focus in the fields of superconductors, ferromagnetics, thermoelectrics, cosmetics, and photocatalytic materials, where investigations have been conducted for a wide range of applications.
Therefore, although it is not possible to introduce all of the layered oxides and hydroxides here, their fundamental properties and reactions are often dependent on the transition metal elements in the layered structure.
Handbook of Layered Materials
In this chapter, the fundamental structures and properties of layered perovskite oxides, layered oxides cobalt, manganese, copper, titanium, niobium , and layered double hydroxides are described. In addition, exfoliation of the layered structures and the formation of layered structures using nanosheets are introduced. The full text of this article hosted at iucr. If you do not receive an email within 10 minutes, your email address may not be registered, and you may need to create a new Wiley Online Library account. If the address matches an existing account you will receive an email with instructions to retrieve your username.
It shows the expected 6-fold symmetry. The peaks are labelled with the corresponding Miller-Bravais hkil indexes. The analysis of the edges also gives reliable information on the number of layers and can be used to investigate a large number of flakes, from zoomed-in high-resolution edge images. If SLG folds or several SLGs stack one on the other, selected area diffraction is used to distinguish contentious cases, in the case of graphene.
This method is therefore the preferred route to establish the thickness distribution in terms of the number of layers of the flakes derived from layered material. Thus, our ink has higher SLG yield with respect to previous works, but lower c than Reference Khan et al This higher c was achieved by long time up to h ultrasonication. However Reference Khan et al reported defect formation and reduction of size as a result. For graphene and nanotubes, AS mix is small.
Therefore, for dispersion and stabilization of graphene in solvents, AH mix needs to be very small. This can be achieved by choosing a solvent whose surface energy is very close to that of graphene. The surface energy of NMP satisfies this requirement and allows efficient exfoliation of graphite. Graphite can also be efficiently exfoliated in water with the use of bile salt surfactants. More detail on this is set out below. The viscosity of NMP at room temperature 1. Larger flakes dispersed in a higher viscosity medium such as NMP experience higher frictional force and sedimentation coefficient, making it more difficult for them to sediment during ultracentrifugation.
The G peak corresponds to the E 2g phonon at the Brillouin zone centre. The D peak is due to the breathing modes of sp 2 rings and requires a defect for its activation by double resonance DR. The 2D peak is the second order of the D peak. The 2D peak is always seen, even when no D peak is present, since no defects are required for the activation of two phonons with the same momentum, one backscattering from the other. DR can also happen intravalley, i. This gives the D' peak. The 2D' is the second order of the D' peak. We assign the D and D' peaks to the edges of the submicrometer flakes, rather than to the presence of a large amount of disorder within the flakes.
This is further supported by the plot of the G peak dispersion, Disp G Fig. Also, Disp G is nearly zero for all samples, compared to the values larger than 0. The distribution of 2D peak position, Pos 2D , Fig. We estimate p of about 1. This gives Z of about 1 1. We use an Epson Stylus inkjet printer with a S cartridge under a constant nitrogen flow. A high acquisition speed camera Sony XCD-X, with s " acquisition rate captures the dynamics of droplet formation.
Shown in Fig. Notably, even if Z is about 24, we do not detect any satellite droplet. Hence, although Z is out of the conventionally assumed stable inkjet printing range, Fig. Thus, we will focus on the pristine graphene ink in the subsequent sections. We note that LPE is a viable technique to achieve liquid dispersion of a range of layered materials e.
Therefore, this approach provides a range of printable inks based on layered materials.
Porous heterostructured clays - recent advances and challenges - revisão
These can then be mixed or printed to form hybrid heterostructures with novel properties. Further details on such structures are set out below. Inkjet-Printed Features. The nozzle of our printer is about 1 mm above the substrate. The final layout of printed nanoparticle inks depends on the substrate SE, as well as ink viscosity and surface tension. The 0 2 plasma is generated at W and 4 x 10 '1 Torr for 2min.
We use optical micrographs with dark-field imaging to visualize the inkjet-printed drops. As discussed above, we use NMP as solvent to reduce the coffee ring effect compared to low boiling point solvents e. However, we still observe coffee rings when printing on pristine Si0 2 , while there is flake uniformity and no coffee rings on HMDS-treated Si0 2. Thus, HMDS appears to prevent coffee rings. To understand this, we measure the substrates' SE and investigate the printed stripes' morphology, before and after surface treatment. We utilize contact angle analysis to estimate the substrate surface tension, and SE.
We have shown the viability of inkjet printing to fabricate complex layouts. Indeed, our S L correspond to SEs of about A small Q C results in the rapid drop spreading on the substrate, as for 0 2 -treated Si0 2. When inkjet printing stripes, the interdrop i. When the distance is large, individual drops are deposited.
As the interdrop distance decreases, these merge into a line.
Thus, in order to obtain a continuous line, we need an interdrop distance smaller than the drop diameter. On the other hand, some workers have reported that a very small interdrop distance can result in particle aggregation on the substrate, thus a nonuniform stripe i. We select an interdrop distance suitable to have continuous lines, avoiding at the same time nonuniformities and irregular edges. AFM confirms the presence of voids and irregular flake distribution, with Rz of about nm.
AFM confirms a more homogeneous network with Rz of about 15 nm. Rz is lower when Q c is higher, because the poor wettability of drops with higher Q c reduces the stripe width, confining the flakes onto a smaller area. The uniformity of stripes printed on the HMDS-treated substrate corroborates the above considerations on the SE changes. In fact, the presence of silane groups in the molecular structure of HMDS acts as promoter of metallic particle adhesion to the substrate.
Analogously, HMDS may promote the adhesion of graphene flakes to the substrate, thus favouring the formation of a regular network. The data show that the first stripe has very similar characteristics to the ink, as expected. Note, however, that the 2D peak shape, even for the 90 nm stripe, remains distinctly different from that of graphite. A similar aggregation of flakes was previously observed for thick films derived from graphene dispersions.
In all cases Disp G remains similar and very low, again showing the lack of large amounts of defects within the flakes. Transparent and Conductive Patterns. We now investigate the viability of our ink to print transparent and conductive patterns. Thus, from Fig. Thus, stripes on HMDS treated glass have a higher o, combined with a more regular network of flakes, compared to the other two substrates.
A similar trend has been reported by others for CNT films on Si0 2 produced by vacuum filtration , inkjet printed CNT patterns on Si0 2 , graphene films on Si0 2 , and polyethylene-terephthalate, as well as Ag nanowire films, produced by vacuum filtration on Si0 2. Such regimes can be explained considering that our films stop behaving like bulk materials below a critical thickness t min , entering the percolation region.
This indicates formation of a more uniform network on HMDS-treated and pristine glass compared to 0 2 -treated glass. We also determine the minimum concentration necessary to achieve the bulk conductivity regime. Given our c of about 0. We assign this to the percolative regime, with o DC deviating from a bulk-like behaviour. Here we report preliminary work on non-optimised devices. It is anticipated that the skilled person will be able to work from this disclosure in order to produce devices with superior performance without undue effort.
Inkjet-printed TFTs based on organic semiconducting polymers have been widely investigated. Several inkjet-printed TFTs using various carbon nanomaterials have been reported. Contacts were formed from chromium-gold source and drain pads. A layer of poly[5,5'-bis 3-dodecyl thienyl -2,2'-bithiophene] PQT is printed on top. The output characteristics shown in Fig. This is consistent with the intuitive idea that optimized devices are obtained when the interdrop distance is roughly equal to the average drop diameter. This behaviour is expected, since we have shown in Fig. On the other hand, the field effect modulation becomes less effective when the channel is too thick.
However, we do not wish to increase too much the post-annealing temperature to avoid possible sample damage. This is yet again expected considering the coffee ring effects of nonoptimized substrates. The overall charge conduction in crystalline organic semiconducting thin films is determined by both intrachain and interchain charge transport.
The former is much faster than interchain hopping. Many groups have tried to improve interchain hopping, e. We combine our graphene ink with one of the most common organic polymers in inkjet printing, Poly 3,3"'-didodecyl quarter thiophene PQT , in order to investigate graphene's viability as an interchain hopping enhancer similarly to Au nanoparticles and CNTs. PQT is widely used due to its higher environmental stability up to days at room conditions , with respect to other organic semiconducting inks.
For each V gs , V ds is swept from 0 to V in steps of 2 V. Thus, the combination of graphene and organic semiconducting inks is promising for high- performance printed electronics. To summarize, liquid phase exfoliated graphene is an ideal and low-cost material to make printable inks.
This demonstrates its viability for flexible and transparent electronics. Our ink preparation technique can be generalized to a wide range of layered materials e. More detail is given here relating to the process of LPE and centrifugation of graphene in order to provide the required thickness distribution for the graphene flakes. As will be apparent to the skilled person, these techniques can be adapted to provide the required thickness distribution for other layered materials.
Graphene dispersions can be purified following different strategies based on preparative ultracentrifugation. This separation technique was historically used for the separation of biological materials in a uniform or density gradient medium DGM. Sedimentation based separation SBS separates various particles into fractions on the basis of their sedimentation rate, which determines how particles in dispersion sediment out of the fluid in which they are dispersed, in response to a centrifugal force acting on them.
During the process, they move along an ultracentrifuge cell, dragged by the centrifugal force, until they reach the corresponding isopycnic point, i. Such a process depends only on the buoyant density of the particles and is also known as isopycnic separation.
- Prejudice in the Modern World: Primary Sources!
- Layered Double Hydroxides (LDHs)?
- 3 Layered Materials: Oxides and Hydroxides;
- Reliques of Ancient English Poetry?
- Algebraic Transformation Groups: An Introduction?
- Hidden Star (Stars of Mithra, Book 1).
If the ultracentrifugation is stopped before the particles achieve their respective isopycnic points, a zonal separation ZS is attained. The latter is dependent on the sedimentation rates of the particles themselves. SBS was introduced in in the field of carbon-based materials for the separation of individual nanotubes from bundles [Reference O'Connell et al ]. On the other hand, while ZS permits the separation of nanotubes by length, isopycnic separation is used for the sorting of nanotubes by diameter, metallic vs semiconducting, and chirality. Bile salt surfactants also allow the isolation of flakes with controlled thickness, when combined with DGU.
High yield of exfoliated graphene can be also obtained via low power sonication of graphite in a di-hydroxy Sodium Deoxycholate SDC aqueous solution followed by a separation in a centrifugal field. Both SBS and isopycnic separation can be used to obtain dispersions highly enriched in monolayer and few layer graphene. On the other hand, isopycnic separation, besides sorting flakes by number of layers, allows us to obtain larger flakes with respect to SBS. This is also evidenced through micro-Raman. We also show that SDC is a suitable surfactant for sorting graphite flakes by number of layers.
We show that this separation is strongly affected by the coverage and clustering aggregation properties of the surfactant molecules. Basic principles of separation in centrifugal field - sedimentation based separation When a graphene-surfactant complex GSC is dispersed in a solvent and subjected to a centrifugal field, three forces act on it. In general, a particle of known volume and density in a medium of constant density will be accelerated under a centrifugal field, until the net force equals the force resisting its motion through the medium, f depends on shape and size of the particles and increases as the particle geometry moves away from a spherical shape, which means that large or elongated particles experience more frictional drag than compact spherical ones of same mass.
Both F and F f act in the opposite direction to sedimentation. The rate of sedimentation in a centrifugal field, is described by the Svedberg equation: where s is the sedimentation coefficient, commonly reported in Svedberg S unit, where 1 S corresponds to 10 "13 sec. S depends on the properties of the particle and is proportional to the buoyant effective molar weight of the particle, while it is inversely proportional to f.
In general, molecules with different weights, shapes or sizes, will move with different velocities in a given centrifugal field. They will thus have different S values. As a function of ultracentrifugation time, the GSC begin to pile up at the bottom of the cuvette. As the concentration of GSC at the cuvette bottom increases, the diffusion process opposes further sedimentation. Equilibrium is reached when sedimentation and diffusion are balanced and the concentration of flakes along the cuvette no longer changes with time.
From equation 1 , at the equilibrium the sedimentation of GSC only depends on the frictional coefficient and molecular weight. Basic principles of separation in centrifugal field - isopycnic separation. Isopycnic separation exploits subtle density variation between objects in order to obtain a spatial separation inside an ultracentrifuge cell under a centrifugal force.
Reference Arnold et al first applied this separation technique in the field of carbon-based nanomaterials for the diameter sorting of SWNTs. In this case, the effective density p of a nanotube, with radius R, shell thickness t and containing a fraction F of liquid filling, can be calculated from Equation 2: where pi is the density of graphene and p 2 that of the liquid.
Equation 2 shows that the density of the nanotube is intrinsically related to its radius and in first approximation is inversely proportional to the radius itself. Thus in principle, the relationship between pure nanotube density and diameter would be sufficient for the discrimination of SWNTs by their diameters in a density gradient. However, the density of SWNTs, at least in the diameter range 0.
This means that when nanotubes are centrifuged in water at high g-force they should pellet at the bottom of the cell.
Services on Demand
However, surfactant shells around SWNTs provide additional buoyancy that keeps individual nanotubes and small bundles aloft whereas larger bundles sediment out, permitting the separation of SWNTs by their diameter in a density gradient. Moreover, the use of co-surfactants mixtures permitted the metallic vs semiconducting separation and the single chirality enrichment. Contrary to nanotubes, where in principle a separation of bare nanotubes is possible, in the case of graphitic flakes we have to induce a density difference between SLG, BLG, etc.
The density difference is then provided by surfactant covering the flakes. The surfactant, besides stabilizing the repulsion between the hydrophobic surface of graphitic flake and water, provides a variation in buoyant density, according to the flake thickness. Considering a uniform layer of surfactant, the buoyant density of the GSC increases with the number of graphene layers.
The graphite powder used Sigma Aldrich: product number consists of graphite flakes with average size in the order of few mm 2. SDC is an amphiphilic molecule, with a hydrophobic and a hydrophilic side, which disperses graphite in aqueous solution by physical adsorption on its surface. SDC molecules are steroids, organic compounds such as cholesterol, composed of seventeen carbon atoms arranged in four rings, the cyclopentenophenanthrene nucleus, and are shaped like a flattened ellipsoid with dissimilar sides.
A short aliphatic chain protrudes from one end of the steroid nucleus and terminates in a hydrophilic group. High HI indicates stronger hydrophobicity, a key requirement in the adsorption of flat molecules onto the hydrophobic graphitic surface whose structure is composed of carbon atoms. The resultant dispersion is ultrasonicated for minutes in a low power bath sonicator Decon, W. During ultrasonication, hydrodynamic shear-forces, associated with cavitation, induce exfoliation. Cavitation is the formation, growth, and collapse of small bubbles or voids in liquids, as a result of pressure fluctuation.
The resulting flakes are then surrounded by the SDC molecules, preventing re-aggregation. The obtained dispersion contains both graphene sheets and thicker graphite flakes. SBS in a centrifugal field is thus requires to remove the thicker graphitic material. Experimental - isopycnic separation.
The preliminary step of the graphene sorting is the ultrasonication of graphite powder in a water-surfactant solution as described above followed by a short pre-ultracentrifugation, 10 min at RPM Sorvall WX ultralOO ultracentrifuge. The supernatant is then extracted and used for the isopycnic separation. The DGM choice is fundamental. Due to low viscosity of the DGM, density gradients produced by salts are less stable with respect to those produced with Sucrose and Optiprep. Moreover, salts induce strong aggregation on the hydrophobic solutes that may affect the separation process itself.
Additionally, the percentage of sucrose used as DGM can have a significant impact on the separation. Sucrose has high viscosity, exponentially increasing at high concentrations and is mainly used in DGU for ZS rather than for isopycnic separation. It is known to use sucrose as DGM, in ZS, to separate GO sheets by size and surface chemistry obtaining almost monodisperse chemically reduced graphene oxide sheets, with relatively small differences in sheet diameter.
On the other hand, Optiprep is better suited for isopycnic separation due to its higher viscosity, with respect to salts, and to density tunability, with respect to sucrose. Moreover, it has an almost constant viscosity as function of the gradient density. By diluting Optiprep, the density profile can be shaped in different ways: linear, nonlinear, or step. Step gradients, formed by stacking layers of different density, are most effective for the separation of molecules with large density differences.
In nonlinear gradients, the DGM density changes nonlinearly along the cell. In principle, nonlinear gradients are the most sensitive, since a variety of depth-density profiles can be produced according to the density variation, enabling trapping of particles of different densities along the cell length. However, often there is no advantage over a linear gradient in the final separation, since the particles take a long time to reach the equilibrium.
However, the initial shape of the gradient is virtually never maintained, because diffusion tends to make it less steep. The density gradient is formed directly in a cuvette Seton, ultra clear open-top, 14x89 mm, capacity We insert 1. On top of this, we place 1. The cuvette is then filled with two more layers: 2. This creates a step gradient. A linear gradient is then produced by diffusion.
The cuvette is capped and tilted horizontally for 2 hours and then vertically for another 2 hours. The average and maximum accelerations are about ,g and about ,g, respectively. During the process, GSC move along the cuvette, dragged by the centrifugal force, until they reach their isopycnic points. The successful separation is indicated by the appearance of grey bands along the cuvette. After isopycnic separation, the sorted flakes are extracted following the fractionation procedure developed in References Arnold et al and Crochet et al for the extraction of SWNTs.
We use upward displacement fractionation exploiting a syringe pump. Fluorinert FC p of about 1. The distance between the top of the dispersion and the upper enriched band is carefully measured using a slide caliper and the corresponding volume calculated. This is then extracted and discarded by injecting the same volume of Fluorinert at the bottom of the cuvette.
We use OAS in order to evaluate the concentration, c, of graphitic material in dispersion. The presence of residuals is not surprising given a considerably thick film, hindering the complete surfactant removal. The knowledge of graphitic mass in the film allows us to determine the final concentration of the initial stock dispersion. A linear fit of the absorption values measured at nm for each diluted dispersion gives a of about Lg " m "1. A reference solution containing only solvent and surfactant is used for background correction.
For the former, the dispersions are drop cast onto holey carbon TEM grids mesh and rinsed with Dl water. The number of graphene layers in a flake can be identified by analyzing the electron diffraction patterns as explained above. We again use this criterion to identify and distinguish SLG from multilayers. Note that, despite being one of the very few techniques able to reliably prove the existence of SLGs, electron diffraction is time consuming. It requires careful sample tilt series for both X and Y directions in order to achieve perfect normal-incidence of the electron beam.
Off-axis deviations result on diffraction spot intensities that cannot be interpreted directly. The number of layers can often be counted from zoomed-in high resolution images of the edges. Often true SLG edges, produced from sheared interatomic bonds, can be recognized by their more random trajectories. If a FLG folds on itself, a scroll is observed along that edge. If a SLG folds on itself, usually electron diffraction patterns will show a random restacking.
Selected area diffraction can then be used as a final proof to distinguish contentious cases. The deposited samples are then rinsed with Dl water in order to remove the excess of surfactant and DGM. Raman spectra are collected with a Renishaw spectrometer at nm excitation by focusing the laser spot on the boundary region of the evaporated drop, using a X objective, and about 1 mW power. The asymmetry of the UV peak, with a high-wavelength tail, is attributed to excitonic effects. The graphene dispersion mostly contains nm flakes, as shown by the TEM analysis in Fig.
This allows us to label the peaks with the Miller-Bravais hkil indices. The inner peaks, and - , are about 1. To have a reliable statistics we measured flakes. We get a very large population of FLG flakes with less than four layers. Statistical analysis, see Fig. The flakes' surface area can be estimated taking into account the approximate geometrical shape of the objects i. The flake surface area statistical analysis in Fig.
Statistical analysis of the micro-Raman spectra shows that the 2D peak has, on average, pos 2D of cm "1 see Fig. Although the pos 2D is upshifted and FWHM 2D larger with respect to that of graphene flakes produced by micromechanical cleavage, the 2D peak still shows a lorentzian lineshape. This implies that the flakes are mostly monolayers or that they are electronically almost decoupled in the case of FLG and behave, to a first approximation, like a collection of single layers. The very large intensity of the D peak is not due to the presence of a large amount of structural defects, otherwise it would be much broader, and G, D' would merge in a single band.
We rather assign it to edges of our sub-micrometer flakes. Thus, due to possible chemical doping induced by surfactant residual atop graphene flakes, it is the shape of the 2D peak that is the most effective way to identify a SLG. Similar results are obtained when trapping individual flakes in solution with the ROT apparatus.
As shown from the statistical analysis of the spectra obtained from individual trapped flakes, the 2D peak has an average position at cm "1 see Fig. Indeed the ROT is intrinsically probing individual flakes, while in micro Raman analysis of drop cast samples it is likely to collect the Raman signal from several different flakes in the laser spot.
Finally Figs. These values are, on average, slightly higher with respect to that obtained with microRaman. Statistical analysis shown in Fig. Results and discussion - exfoliation mechanism. To predict the aspect ratio of graphene flakes obtained by sonication, we use a similar approached developed by Ahir et al in order to investigate the mechanical scission of nanotubes. The graphite flakes are considered as a multilayered cylinder composed of N number of stacked, circular graphene layers, with radius r. During sonication, the bubble implosion imposes an inward radial fluid flow, which induces viscous forces on the flakes.
Such forces must be sufficiently high for exfoliation, but sufficiently low to avoid in-plane fracture, in order to produce large graphene flakes. However, this condition should balance the condition that takes into account the intrinsic fracture of the graphene layer in order to avoid the in plane fracture. To estimate the exfoliation strength o e , i. This model estimated the van der Waals force between two parallel surfaces by using the Hamaker constant H, a parameter resulting from the theory of the pair-wise summation of the London dispersion energies between atoms.
In general, the van der Waals interactions include the force between two permanent dipoles Keesom force , the force between a permanent dipole and a corresponding induced dipole Debye force , and the force between two instantaneously induced dipoles London dispersion force.
Carbon atoms are non-polar materials and therefore London dispersion force and van der Waals force are equivalent. The force between two parallel graphene layers, averaged over different relative orientations, is estimated as Eq. From the force between two parallel surfaces we get:. Using Eq. B2, the average statistical distribution of graphite flakes composed of different layers can be estimated. The graphite exfoliation via sonication produces on average flakes with lateral sizes that increase with the number of layers. Thus the exploitation of SBS permits to obtain a separation based on the number of layers other than on their mass.
However, the separation process will be more effective for SLG with a decrease in efficiency for flakes with a higher number of graphene layers. Results and discussion - isopycnic separation. In order not to lose such flakes and maintain a high SLG percentage, a sorting mechanism which selects also the larger flakes is necessary. Thus, we demonstrate that isopycnic separation can overcome the drawback of SBS, because it separates flakes by density instead of mass, as explained above.
TEM analysis provides evidence of effective sorting of graphite flakes by number of layers. HRTEM microscopy reveals graphene sheets with dimensions of several hundred nanometers. A more definitive identification of graphene can be made by analysing the electron diffraction patterns.
The inner peaks, and , are about 5 times more intense than the outer ones, and , indicating that the flake is SLG.
Intercalated Layered Materials
On the other hand, flakes extracted at higher buoyant density fraction 12 are composed of a higher number of layers with respect to those extracted in fraction 1. The fact that the flake is multilayer is also demonstrated by the diffraction pattern Fig. In this case, the outer peaks, and , are more intense than the inner ones, 10 and , indicating that the flake is multilayer see Fig.
At even higher buoyant densities, we find flakes with a higher number of layers with respect to fractions 1 and The statistical analysis of the flake surface area reported in Fig. These graphene flakes show a surface area at least two orders of magnitude larger with respect to those reported by Reference Green and Hersam Flakes in fraction 12 are composed of a higher number of layers in comparison with that in fraction 1 , with neither SLGs nor BLGs. In particular the average surface area is about 2 times larger than that of fraction 1.
In fraction 12 the flakes have an average area of 1. These spectra show systematic changes both in shape and intensity of the D, G, D' and 2D peaks. In order to quantify these variations, in Figs. We note a softening of Pos G and a stiffening of Pos 2D going from fraction 1 to fraction G and 2D peaks have different doping dependence, however, in this case all the fractions are extracted from the same dispersion having the same surfactant and DGM, which means that the starting material has been subjected to the same possible unintentional doping process.
The aggregation properties of the surfactant molecules atop the flake surface depend on the surfactant type and concentration. Isopycnic separation permits to separate selectively flakes of a given surface to volume ratio. We use Eq. B3 to calculate the variation of buoyant density of GSCs composed of a different number of layers for different surfactants or different c s. The graphene-SDC complex has a lower buoyant density and a higher ability to separate graphite flakes that differ by number of layers with respect to SC.
This relies in the different aggregation properties of the two bile salts.
Indeed, SDC and SC differ only in the number of -OH groups, which was demonstrated by others to be a key parameter for the efficient sorting of SWNTs, due to a different aggregation property of the two surfactant molecules. Thus, while in nanotubes the formation of bigger surfactant clusters around the sidewalls is a drawback for the effectiveness of the separation, in the case of graphitic flakes this is an advantage.
- Comparative Veterinary Pharmacology, Toxicology and Theraphy: Proceedings of the 3rd Congress of the European Association for Veterinary Pharmacology and Toxicology, August 25–29 1985, Ghent, Belgium Part II, Invited Lectures.
- Something Only We Know.
- Homers Katalog der Schiffe;
- About this book.
Indeed, large surfactant aggregate tend to reduce the buoyant density of the GSC, thus permitting the separation of flakes with a higher number of layers with respect to the same process carried out with SC. This difference relies on the micellization behaviour of the two bile salts. While dihydroxy bile salts have elongated structures, the trihydroxy ones form highly hydrated globular micelles. This has been attributed to the total number of -OH groups and to their position on the steroid backbone, which changes the hydrocarbon-water contact area.
Bile salt micelles form in two stages: the primary micelle appears first. Subsequently the interaction between the hydrophilic surfaces of the salt molecules forms bigger aggregates, called secondary micelles. The lack of the -OH group in position a7 of the SDC backbone increases the hydrophobic area, initiating the formation of secondary micelles. On the contrary, the presence of the third -OH group in SC reduces the formation of secondary micelles. Because the driving force for secondary micelle formation is the hydrophobic interaction between water molecules and the surface of the monomer or dimer, the -OH group in position a7 on the hydrophilic part of SC reduces the hydrocarbon-water contact area in comparison with SDC.
This results in an overall reduction of the buoyant density of the GSC after isopycnic separation, permitting the separation of flakes with a higher number of graphene layers, compared to other surfactants. In conclusion of this section, we demonstrated a scalable and efficient process to prepare graphene dispersions produced by low power sonication of graphite in aqueous solution.
A fracture mechanics model was developed to demonstrate that the maximum aspect ratio of a graphene flake radius over thickness obtained by sonication is related to the balancing between the condition for exfoliation and that for fracture of the graphene flake. We achieved high percentage of SLGs after separation in preparative ultracentrifugation. We have also shown that isopycnic separation, in addition to separate graphene by number of layers, produces flakes of micrometer size that are orders of magnitude larger with respect to the ones produced by SBS. Tuning the type and concentration of surfactant used for the separation, it is possible to improve the control on the number of layers of the sorted graphene flakes.
As for nanotubes, the surfactant clustering is thus a key factor for graphene separation.