CC BY
4УГЛ Е Р ОД rJ Ь J Е ГАГ О СТРУКТУРЫ
CARB ON NAN OSTRU CTURES
ON SOME TECHNIQUES AND EXPERIMENTAL RESULTS: RELEVANCE FOR NANOTECHNOLOGY APPLICATIONS
Yu. S. Nechaev ^
Member of the International Editorial Advisory Board
G. V. Kurdjumov Institute of Metals Science and Physics, I. P. Bardin Central Scientific-Research Institute for Ferrous Metallurgy, Vtoraya Baumanskaya st., 9/23, Moscow, 105005, Russia E-mail: Yuri1939@inbox.ru
PART I On some experimental proofs of the hydrogen multilayer intercalation with carbonaceous nanostructures: Relevance for development of superadsorbents for fuel-cell-powered vehicles
The analytical consideration of some recent experimental and theoretical data on the hydrogen on-board storage problem shows the necessity and economical expediency of carrying out further basic studies and initiating a constructive discussion on the physical key-note aspects («open questions») of the hydrogen sorption by carbon-based nanomaterials: especially, on the hydrogen multilayer intercalation in carbonaceous nanostruc-tures, their relevance for the development of super-adsorbents for fuel-cell-powered vehicles, i. e. storage materials, satisfying most of the U.S. DOE targets. It is consistent with the U.S. National Academies' recent recommendations and manifestations of the critical situation of the hydrogen storage problem.
Introduction
As is known [1, 2], hydrogen (H2) has recently been intensively investigated as an ideal secondarily derived renewable energy source. Out of the problems to be solved for the industrial utilisation of hydrogen energy, the development of an effective hydrogen storage system (or technology) is the most important problem [1, 2]. The U.S. Department of Energy (DOE) has established different targets and requirements for on-board hydrogen storage systems. Their strategic objectives for the year 2010 include a minimum «gravi-
metric» capacity (weight of stored H2/system weight) of 6 wt. % and a «volumetric» capacity (density) of 45kgH2 m-3. These values are referred to the whole system, including the storage medium, the vessel, refuelling infrastructures, any regulators, electronic controllers, sensors, et cetera (www.eere.energy.gov/hydrogenandfuelcells). Therefore, it is important to emphasize that for the achievement of the system-level targets, the gravimetric and volumetric capacities of the storage material (medium) alone must clearly be higher than the established system-level improvements. In addition to weight, volume and cost, there are also several important DOE targets defined by vehicular requirements: namely, hydrogen charging and discharging rates, durability, safety and op-erability over temperatures and pressures. For last about 10 years, various storage strategies and technologies have been proposed and tested, but to date none of the approaches have fulfilled all of the DOE requirements and targets for either transportation or utility use. As concluded in a paper on the U.S. Department of Energy National hydrogen storage project [1], DOE agrees with the National Academies' recent recommendation [2] that it should be continued to elicit new concepts and ideas, because success in overcoming the major stumbling block on on-board storage is crucial for the future of fuel cells in transportation systems. The development of hydrogen-fueled vehicles and portable electronics will require new materials, and especially, nanomaterials that can store large amount of hydrogen at ambient temperature and relatively low pressures, and provide small volume, low weight, and fast kinetics for recharging.
Статья поступила в редакцию 06.10.2007 г.
The article has entered in publishing office 06.10.2007.
I
189
For last about 10 years, graphite, carbon na-notubes, and nanofibers have been both theoretically and experimentally investigated as potential adsorbent structures (cf. reviews in [3, 4]). Even though some reports claim very high storage capacity, such findings have not been supported by the majority of investigators. Recently, however, hydrogen storage in nanostructured carbon has attracted renewed interest because of new developments in nanotechnology research, and the significant advantage of this light weight and reasonably inexpensive material.
Herein, some analytical study results are presented on the hydrogen multilayer intercalation with carbonaceous nanostructures and their relevance for the development of super-adsorbents for fuel-cell-powered vehicles, i. e. storage materials satisfying most of the DOE targets.
1.1. On the specific intercalation of atomic hydrogen into graphene layers
A real possibility of hydrogen intercalation into (between) near-surface graphene layers of highly oriented pyrolytical graphite (HOPG) has been shown in a series of experimental studies [5-8]. Study [5] of atomic hydrogen accumulation in HOPG samples and etching their surface on hydrogen thermal desorption (TD) have been performed using scanning tunneling microscope (STM) and
Fig. 1.1. STM images of the untreated HOPG sample (under ambient conditions) for areas (a) 60.8-60.8 nm, and (b) 10.910.9 nm, the high resolution image at position of the square in image (a), AFM image (tapping mode, under ambient conditions) of the HOPG sample for 1,000-1,000 nm area, subjected to atomic hydrogen dose (D) of 1.8-1016 H°/cm2(c). Along the line shown in (c), the surface height profile (d). The STM tunnel bias voltage and current were 50-100 mV and 1-15 mA, respectively [5]
atomic force microscope (AFM). The surface morphology of untreated reference HOPG samples revealed by STM was found atomically flat (Fig. 1a), with a typical periodic structure of graphite (Fig. 1b). Exposure (treatment) of the reference HOPG samples (30-125 min at the pressure of about 1 Pa and near-room temperature) to different atomic hydrogen doses (D) has drastically changed the initially flat HOPG surface into a rough surface, covered with bumps-blisters (Fig. 1c), with the average height (for the shown case) of about 4 nm (Fig. 1d).
On TD heating the HOPG samples (under mass spectrometer control), desorption of hydrogen was found (Fig. 2a). As is shown in Fig. 2a, the des-orbed hydrogen amounts (Q) increase with increase of the total hydrogen dozes (D) to which HOPG samples were exposed; percentage of D which retained in samples (Q) goes on decreasing towards a saturation stage. After TD, the HOPG surface there was no bumps anymore, rather the graphite surface was atomically flat, and it was covered with some etch-pits of nearly circular shapes (Fig. 2b), and (as was also shown) of monolayer (or sometimes, two layers) depth. This implies that after departure of captured hydrogen gas, the bumps-blisters became empty of hydrogen and HOPG surface restores back flat surface morphology under action of Van der Waals forces [5]. It was also found [5] that bumps-blisters on HOPG surface containing hydrogen gas were removed after 12-14 times of successive STM scanning (in ambient), leaving behind flat graphite surface with smaller (in comparison with the TD case) irregular etch-pits of monolayer (or sometimes, two layers) depth over there. It's supposed [5] that during successive STM scanning process, holes (in some bumps) created by STM tip becomes bigger when hydrogen gas escapes through them, because some of the carbon atoms from holes edges accompany the hydrogen gas, contributing to bigger sizes of the holes (etch-pits). These results are consistent with results [9] on the «protuberances» disappearance under successive STM scanning of a graphite (HOPG) surface exposed to atomic hydrogen. It is also supposed in [5] that during TD process there is formation of small amount of hydrocarbons and their release together with hydrogen accounts for bigger sized etch-pits of nearly circular shapes.
These and some other observations of net flat HOPG surface in the TD and successive STM scanning cases made the author of [5] to construct a model as shown in Fig. 3. According to this mod-
el, bumps found on HOPG surface after atomic hydrogen exposure (Fig. 1c) are simply graphite blisters, containing inside hydrogen gas in molecular form. As is supposed, atomic hydrogen intercalates in between layers in graphite net through holes in graphene hexagons (due to a small diameter of atomic hydrogen in comparison with the hole size) and latter on being converted to H2 gas form makes it captured inside graphene blisters (due to a relatively large kinetic diameter of hydrogen molecules). But on heating (TD process), some blisters are broken (punctured) due to rise of pressure, up to the tensile limit (strength) for graphite layer forming blister [5]. The number of etch-pits after TD (Fig. 2b)) is considerably smaller than the number of previously existing blisters (Fig. 1c), showing that each blister was not broken to leave behind an etch-pit, and that it took place (during TD heating) some blisters merging and/or coalescence.
Taking data from Fig. 1 (and some others), author [5] found an average blister radius of 25 nm and a height of 4 nm. Then considering the blister as a semi-ellipse, the blister area (Sb ~ 2.010-11 cm2) and its volume (Vb~ 8.410-19 cm3) were found. The amount of retained hydrogen in this sample was Q~ 2.81014 H2/cm2 (Fig. 2a) and the number of hydrogen molecules captured inside the blister turned out as QSb ~ 5.5103 [5]. Thus (within the ideal gas approximation) the pressure for the single bluster at room temperature is of PH2 ~
~ k(QSb )/Vb ~ 2.5 107 Pa, the estimated accuracy is not higher than the order of magnitude. During TD heating, for instance, at 1000 K the pressure can reach a value of PH2 ~ 8.5107Pa (within the ideal gas approximation), which can be enough for some blisters to get punctured (or tensile ruptured), maybe due to some defects in their roofs (walls). In paper [5] the estimated quantities (pressures) are put lower by one order of magnitude (probably, it's a slip of the pen); particularly, it is confirmed by the fact that in previous similar studies [6-8] the pressure values of PH2 ~ 30-50 MPa were declared.
It is relevant to note that estimates [10] of hydrogen fugacity as a function of pressure (up to 1.9108 Pa) and temperature (for 223 1000 K), with using the Abel-Noble equation of state, show that the ideal gas approximation (used above) is available for [5] conditions, within order-of-mag-nitude accuracy.
In [5] the pressure values are compared with known experimental values of tensile and compres-sive strengths for graphite — 107Pa and 3107Pa, respectively. But it seems more reasonable to use recent data on elasticity, strength and toughness
Fig. 2.1. (a) Hydrogen storage efficiency of HOPG samples, desorbed molecular hydrogen (Q) versus doze (D) of atomic hydrogen exposure. (b) STM image for 600x600 nm area of the HOPG sample subjected to atomic hydrogen doze of 1.8- 10te H°/cm2, followed by hydrogen thermal desorption (TD) [5]
Fig. 3.1. Model showing hydrogen accumulation in HOPG, forming blister-like structures. (a) Pre-atomic hydrogen interaction stage. (b) After interaction stage, H2 captured inside grapheme blisters. Sizes are not drawn exactly to scale [5]
of carbon nanorods and nanotubes, for instance, [11-13], data [14] on stress-strain state of multiwall carbon nanotube under internal pressure, and, for instance, data [15] on carbon onions as nano-scopic pressure cells for diamond formation. In these studies [11-15] much higher values (by several orders of magnitude, in comparison with graphite) of modulus of elasticity, modulus of elongation and tensile strength are declared. Hence, it follows: (i) that the blister formation at room temperatures (Figs. 1-3) can occur within the elastic deformation conditions, (ii) that an counteraction blister pressure (of the fugacity order) should be taken into account, and (iii) that the hydrogen pressure (or fugacity) in the blisters [5] can be much higher than the above estimated values (in the least, within the experimental and approximation errors).
It is consistent with the thermodynamic estimation of the equilibrium hydrogen fugacity (fH2) in the blister which can be performed by using the
«acting masses law» for the «reaction» of 2H( ,hH„ . , as follows:
(gas) 2(gas_in_blisters)'
IhJP° - (Ph/P0 )2 X
xexp{[AHdls -TASdls - fH2 (AV/n^Rl), (1-1)
where PH - 1 Pa is the atomic hydrogen pressure in the atomizer [5], P0 = 1 Pa is the standard pressure, AHdis = 448 kJ/mol (H2) is the known experimental value of the dissociation energy (enthalpy) of one mole of gaseous hydrogen (at room temperatures), ASdis - 0 (in the used approximation) is the dissociation entropy, AV - Vb, n - (QSb)/NA, NA being the Avogadro number, R is the gas constant, T - 300 K [5]; hence, fH2 > 109 Pa.
It is necessary to emphasize that the adsorbed hydrogen amount Q (localised, mainly, in the blisters in the graphite monolayer (Figs. 1-3)) corresponds to a relatively low average hydrogen/carbon atomic ratio in the monolayer (i. e., between the two graphene layers), namely: (H/C)~(2Q/NC)-0.1, where NC being the number of carbon atoms per 1 cm2 of two graphene layers.
On the other hand, the hydrogen volumetric (mass) density in the blisters [5] can be estimated
as p- (QMHiSb)/vb - 45 kg/m3, where MH2 being the
hydrogen molecule mass; this is close (within the experimental errors) to the liquid hydrogen mass density (70.8 kg/m3 at 21.2 K and 0.1 MPa).
By using the experimental and estimates' results, one can conclude on some three-dimensional clustering of hydrogen molecules in the graphite monolayer (between the two graphene layers), that is on formation and growth of liquid-like three-dimensional nanoclusters (Figs. 1-3) under [5] conditions.
In this connection, it is relevant to point, for instance, to results [16] of molecular dynamic simulations of liquefaction of hydrogen molecules upon exterior (deformed) surfaces of single-walled carbon nanotube (SWNT) bundles (at 80 K and 10 MPa). Those studies were carried out in relevance to the very known experimental finding for SWNT bundles [17].
It is also relevant to compare data [5] with related results [9] of studying the graphite surface modifications induced by interaction of hydrogen atoms (respectively deuterium) with perfectly crystalline HOPG surfaces. Surface properties were probed [9] with high-resolution electron-energy-loss spectroscopy (HREELS) revealing the formation of C-H units with different vibrational energies. Comparison with density functional theory (DFT) calculations [18] led authors [9] to establish the models of the hydrogen adsorption processes at the graphite surface. It was shown in [9] that the vibration at 295 meV was due to a single H atom bonding to graphite (C-H), while vibrations at 331 meV
and 345 meV (and higher energy losses) were respectively related to formation dimmer and quartet, or more generally, a higher number of clustering atoms (i. e., hydrogen clusters formation). Subsequently, studies were performed [9], with using scanning tunneling microscopy (STM). On the electronic point of view [9], as hydrogen locally disturbed the electronic density near the Fermi level, charge density confinement was observed between the hydrogen clusters, particularly when using low tip-sample bias voltage. It was also observed that when the tip iteratively scanned the same surface area, protuberances which were attributed to the hydrogen presence, eventually disappeared, evidencing hydrogen desorption. Authors [9] assumed that the desorption phenomenon derived from mechanical interaction between the tip and the graphite surface.
It is also necessary to take into account experimental data [6-8] on hydrogen thermodesorp-tion (TD) from highly oriented pyrolytic graphite (HOPG) exposed to atomic hydrogen (as in [5]) at near-room temperatures, and their constructive (thermodynamic) analysis [3, 4]. Such a treatment resulted in emergence of surface nano-hill-ocks (nano-blisters) with heights of 3-5 nm and diameters of 40-75 nm, the most of which disappeared after hydrogen thermodesorption that proceeded as a first-order reaction [7, 8]. The TD measurements (at a heating rate of u = 25K/s) revealed two TD peaks (processes): a peak a centered at Ta- 1123 K (ATa- 180 K — its width at half of its height, Sa/Sx- 0.45 — its fraction of the total spectrum area, Qa- 230 kJ/mol — the activation energy of the process) and a TD peak p centered at Tp- 1523 K (ATp- 250 K, Sp/Sx- 0.55, Qp - 385 kJ/mol).
Using results of analyses [3, 4] and results of the above consideration, one can attribute TD process (peak) a with process III of dissociative chemisorption of hydrogen between graphene layers (Table 1, model F* in Fig. 4). The rate-controlling stage of the process (peak) a can be attributed to diffusion of hydrogen atoms (between the two surface graphene layers (Fig. 3b)) from the nearest graphene blisters to a «punctured» one (accompanying with the diffusant reversible trapping, i.e., C-H bonding at chemisorption centers in the graphene layers (model F* in Fig. 4)). The diffusion characteristics are as Dm, D0III and Qin - Qa (Table 1); the diffusion length (DmATa/u)1/2 is of order of 1-10 nm, i.e., as the separation between the walls of the neighboring blisters (Fig. 1d, Fig. 3b). It can be related with study results [9] (considered above) on the vibration contribution at 295 meV due to a single H atom bonding to graphite (C-H).
In the same way, one can attribute the TD process (peak) p with process IV of dissociative chemisorption of hydrogen between graphene lay-
International Scientific Journal for Alternative Energy and Ecology № 11(55) 2007
© 2007 Scientific Technical Centre «TATA»
Table 1
Characteristics of hydrogen chemisorption and diffusion in isotropic graphite and related carbon nanostructures (herein, Refs., Figs. &
Eqns. are corresponded to those of [3])
Hydrogen chemisorption in sp2 carbon materials Chemisorption and diffusion models and the energies of formation of chemical bonds linking the hydrogen atoms to the material Characteristics of the processes Type of sorption isotherm
Process III in isotropic graphite [51, 53] (Figs. 5 and 7a, TPD peak III), in GNFs [12] (Fig. 6, peak y (III)), and in nano structured graphite [14, 52-56] (Figs. 7b and c, III) Dissociative chemisorption of hydrogen between grapheme layers (reaction (l)-(4)). Bulk diffusion of hydrogen atoms with a reversible diffusant capture at chemisorption centers in graphene layers (Fig. 8, model F*)\ AHim = -243 + 3 kj mol"1 (H) AHmiI « 1/2 WJb + AH(ml «-19 ± 1 kJ mol"1 (H), ASmil/R « -14.7 (-15.4) at A;, « 0.5 (1.0), Eqns. (5)-(7). Aii = £>oiiiexp(-0ni/m Am ~ 3T0"3 cm2 s"1, Qm «Q1- A/km = 250 + 3 kJ mol"1 (H), Ql « 7 + 4 kJ mol"1 (H), Eqns. (8), (9) Sieverts - Langmuir Eqns. (5), (5a)
Process II in isotropic graphite [51, 53] (Figs. 5 and 7a, II), in GNFs [12] (Fig. 6, peak ß (II)), and in defective single-wall [61] andmultiwall [62] nanotubes Dissociative-associative chemisorption of H2 in intergranular of defective (surface) regions (reaction (10)-(13)). Diffusion of H2 in these regions with reversible diffusant dissociation and capture at sorption centers in (Fig. 8, model H)\ Aflfi2 hi « -560 + 10 kJ mol"1 (2H) Aff(M « AHdis + AHam « -110 kJ mol"1 (H2), AS(im/R « -30 atXIIm « 0.5 (0.25), Eqns. (14)—(16). Ai = A>nexp(-fti/ÄD, Doii^l-STO^nrs"1, On « QM- Afffuo « 120 + 2 kJ mol"1 (H2), çfef» 10 + 5 kJ mol"1 (H2), Eqns. (17), (18) Henry - Langmuir Eqns. (14), (14a)
Process I in isotropic graphite [51, 53] (Fig. 5, TPD, peak I), in single-wall nanotubes [26, 63, 64], and in multiwall nanotubes [62] (Section 5.1) Dissociative-associative chemisorption of H2 in siu'face layers of the material (reaction (10)—(13) and (12a)). Diffusion of H2 in these layers with reversible diffusant dissociation and capture at chemisorption centers in (Fig. 8, model G and F); AH(12,12a ii « -460 + 10 kJ mol"1 (2H) AHfoii « AHdis + AHax 12ali « -10 + 7 kJ mol"1 (H2), AS(13lI/Ä « -20 at Al„ « 0.5 (0.25), Eqns. (14)-(16). A = A>iexp(-ßi/ÄD, Ai ~ 3T0"3 cm2 s"1, Ql « ff - AfffUB « 20 + 2 kJ mol"1 (H2), Çf « 10 + 8 kJ mol"1 (H2), Eqns. (17), (18) Henry - Langmuir Eqns. (14), (14a)
Process IV in isotropic [51, 53] (Fig. 5, peak IV) and pyrolytic [60] and nano structured [52, 53] (Fig. 7a, peak IV) graphite Dissociative chemisorption of H2 in defective regions of the graphite lattice (reaction (l)-(4)). Bulk diffusion of hydrogen atoms in defective regions with reversible diffusant capture by chemisorption centers (Fig. 8, model C and D)\ AH(3)W « -364 + 5 kJ mol"1 (2H) AH(4)W « 1/2AHdis + AH(3)W « -140 + 5 kJ mol"1 (H), Eqns. (5)-(7). Dw = Dowexp(-Qw/RT), Dow « 6102 cm2 s"1, Qw « -AH(3)N « 365 + 50 kJ mol"1 (H), Eqns. (8), (9) Sieverts - Langmuir Eqns. (5), (5a)
Note. -D0III, Dou, Dol, and -D0IV are the pre-exponential (entropic) factors of hydrogen diffusivities (Dm, Du, Dv and DIV) for carbon materials corresponding to the respective processes.
ers with some defects, for instance, as dislocation loops (Table 1, models C and/or D in Fig. 4). The rate-controlling stage of the process (peak) p can be attributed with diffusion of hydrogen atoms (between the two surface graphene layers (Fig. 3b)) from the available graphene blisters to a «punctured» one (accompanying with the diffusant reversible trapping, i. e., C-H bonding at chemisorp-tion centers in the defects' regions of the graphene layers (models C and/or D in Fig. 4)). The diffusion characteristics are as DIV, D0IV and QIV -- Qp (Table 1); the diffusion length (DIVATp/v)1/2 is of order of 10-100 nm, i.e., as the separation between the neighboring etch-pits (Fig. 2b); the defects of the dislocation loop type can be formed (created) during (and/or due to) «shrinking» and/ or disappearing a number of the blisters.
The in [5] considered mechanism of etching (i. e., creating etch-pits on the surface (Fig. 3b)) due to departure together with some hydrogen accompanying carbon atoms (obviously, as hydrocarbons complexes) from the holes edges (in punctured graphene blisters), resulting in bigger sizes of the holes, can be attributed [3, 4] only with process II of dissociative-associative chemisorption of hydrogen molecules (Table 1, model H in Fig. 4). As has been shown in [3, 4], only process II (from
H H H D
Fig. 4.1. Theoretical models of hydrogen atom chemisorption on graphite [19]; it's Fig. 8 in [3], which is referred in Table 1
the considered ones I-IV in Table 1) is characterized by an accompanying (initiated by the process) occurrence of a fairly small amount of hydrocarbons (CH4 and others) in the thermal desorption (TD) spectra. The explanation [3, 4] of this phenomenon is that the energy (-AH(12)II, Table 1) of desorption (detachment) of two hydrogen atoms from the carbon atom of the sorption center (model H in Fig. 4) is much higher than the energy (-AHC-C, -485 kJ/mol) of detachment of this carbon atom from its the two nearest carbon neighbors. It can be related to study results [9] (considered above) on the vibration contribution at 331 meV due to a dimmer of hydrogen atoms bonding to graphite. Probably, such type of contributions can be also attributed to process I (Table 1, model F in Fig. 4).
As has been noted in [5], the results lead to the assumption that atomic hydrogen could be accumulated in closed graphite nanotubes, through graphene sheet walls of the nanotubes, in H2 gas form and this storage would be stable.
In this connection it is relevant to note results [20-23] on interactions of low-kinetic-energy hydrogen atoms (0.5-30 eV) with single-walled carbon nanotubes (SWNT) based on the molecular dynamics and, particularly, ab initio calculations. According to their finding, hydrogen atoms with an energy of 16-25 eV are characterised by a high probability of penetrating through the side faces of the closed SWNTs and accumulating inside these capsules in the form of hydrogen molecules (Fig. 5). Owing to the high mechanical strength of nano-tubes, hydrogen molecules can be concentrated therein up to volume densities exceeding by far that observed upon capillary condensation (Fig. 5). According to estimates in [21], pressure of molecular hydrogen embedded into SWNT can reach a value as high as 60 GPa. This hydrogen condensate inside individual SWNT, depending on hydrogen volume density and pressure, can undergo several phase transitions giving rise to different crystal lattices built up of hydrogen molecules [22]. If the pressure of hydrogen inside a (5,5) SWNT is 37.4 GPA, the bulk sorption capacity can reach 63 kg/m3, which meets the DOE requirements [1] of the automotive industry for the year 2015. In [23, 24] potential ways for practical achieving of such high hydrogen sorption levels by bundles of closed SWNTs have been suggested as well as methods for hydrogen withdrawal through the walls of the nanotubes. As a matter of fact, recently, it was experimentally proven [25, 26] that 5.1 wt. % of hydrogen storage was obtained by hydrogenation of single-wall carbon nanotubes with atomic hydrogen using core-level photo electron spectroscopy and X-ray absorption spectroscopy.
It is relevant also to add that instead of the atomic atomizer technique one could try to use the
194
International Scientific Journal for Alternative Energy and Ecology № 11(55) 2007
© 2007 Scientific Technical Centre «TATA»
S,. à r
Fig. 5.1. Arrangement of H2 molecules inside a (5,5) singlewalled carbon nanotube (SWNT) at different volume densities of hydrogen [20]. Volume (mass) density of hydrogen/kg m-3: (a) 20, (b) 51, (c) 90, (d) 125, (e) 142
technique of the electrolytic hydrogen charging of materials in question.
2.1. On the hydrogen intercalation vs. chemisorption mechanisms of the spillover enhancement of the sorption capacity of carbonaceous nanomaterials with metals-catalysts nanoparticles
Recently, a number of data, for instance, [2735] has been obtained on the spillover enhancement (of about an order of magnitude (Fig. 6)) of the hydrogen storage capacity (at room temperatures and technological pressures) of carbonaceous nanomaterials containing some amount of nano-particles of transition metals (as H2 dissociation catalysts, e. g., Pd, Pt, or Ni).
In this connection, the renewed interest to carbonaceous nanostructures relevance to the onboard storage problem has been emphasised [30].
It is supposed in [27-35] that the presence of highly dispersed metal nanoparticles (Pd or others) in carbon nanomaterials facilitates the initial adsorption of molecular hydrogen and its subsequent dissociation into hydrogen atoms. These molecules may then spill onto (over) nearby carbon sites by crossing the interface between metal and carbon, migrating (maybe via activated surface diffusion) across the carbon surface, and finally being stored in some new «carbon sinks» — carbon defect sites and/or carbon defective structures (both edge and in-plane sites on graphene
Fig. 6.1. High pressure hydrogen isotherms at 298 K for a pure metal organic framework IRMOF-8 (a), Pt-doped activated carbon Pt/AC and IRMOF-8 physical mixture (1:9 weight ratio) (*), and for bridged sample of Pt/AC-bridge-IRMOF-8: first adsorption (o), desorption (p) and second adsorption (O), [34]
sheets, or a minor expansion of local inter-layer distances), which are obviously developed in the neighbourhood of the metal nanoparticles. In some instances [30, 31], the amount of hydrogen adsorbed by the Pd-containing carbon fibres was many times the value that was expected on the basis of formation of Pd hydride alone. Therefore, the storage mechanisms (the hydrogen intercalation vs. chemisorption), as well as the carbon structural changes have been discussed in [30]. The chemisorption contributions have been found, for instance, in [28] by using Fourier transform infrared spectroscopic analysis. Some different models are considered, for instance, in [36, 37]. It is relevant to note that a negligible (i.e., by about one order less than in [27-35]) spillover effect has been found in [37] for hydrogen uptake of Pt-doped graphite nanofibers; it has been «surprising» for authors themselves [37]. As has been noted in [27-37], the spillover enhancement mechanisms are not well understood and further studies are necessary.
There are bases [3, 4] to suppose that the micromechanisms of the observed phenomenon [2737] can be (in some extent) related to the above considered results [5-9] on the specific intercalation and chemisorption of atomic hydrogen into graphene layers, and also to the discussed below results [17, 38, 39] on the hydrogen intercalation through graphene layers induced (initiated) by the chemisorption.
3.1. On the hydrogen intercalation (multilayer physical adsorption) in graphite nanofibers (GNFs) and SWNT bundles initiated by monolayer chemisorption
Analysis [3] of a number of related studies has revealed a real possibility of the hydrogen
■ 1
195
intercalation (multilayer physical adsorption) in graphite nanofibers and carbon SWNT bundles initiated (induced) by the monolayer chemisorp-tion of I-III types (Table 1).
Particularly, it has been shown [3] that the above noted distinguished data [17] (Fig. 7) can be related to the hydrogen intercalation (multilayer physical adsorption or hydrogen liquefaction) initiated by the monolayer chemisorption of I type (Table 1). It is relevant to note that the non-conventional adsorption isotherms in Fig. 7 are in some extent similar to the isotherms in Fig. 6; hence, some similarity in the adsorption micro-mechanisms can be supposed.
As is noted in [3], a direct experimental proof of the multilayer intercalation (or condensation) of hydrogen between graphene sheets in graphite nanofibres (GNFs) bundles with Pd catalyst (Fig. 8) has been obtained in [38], with declaring of an anomalously high value of their hydrogen storage capacity (17 wt. % (hydrogen weight to sum of carbon and hydrogen weights)) for charging at 300 K and 8 MPa of H2 gas). From these data (Fig. 8)) a rather high value of the adsorbate volume (mass) density can be estimated with order-of-magnitude accuracy. It can be also shown [3] that the observed multilayer intercalation (Fig. 8) was induced by the monolayer chemisorption of III and II types (Table 1).
As a matter of fact, experimental data [38] are adequate (in principle, relevance to the multilayer intercalation phenomenon) to the super-adsorption sensational data [39] for GNFs (in total, about 40 wt. % (hydrogen weight to sum of carbon and hydrogen weights) for charging at 300 K
and 11 MPa of H„
, those (according to analy-
sis [3]) can be attributed with the multilayer inter-
Fig. 8.1. Micrograph of dehydrogenated graphite nanofibres (GNFs); the arrows indicate some of the formed slit-like nanopores (due to the hydrogen multilayer intercalation) [38]
Fig. 7.1. Hydrogen adsorption isotherms at 80 K for the initial untreated SWNTs (1), for SWNTs treated with ultrasound prior to hydrogen charging (2), for samples (2) after a second round of hydrogen charging, for Saran activated carbon (4), and for the Saran material after introducing a correction for the ratio (3:16) of the specific surface areas (5), [17]
Fig. 9.1. Temperature-programmed desorption curves [39] of the residual hydrogen desorption (with TD peaks f and y those can be attributeded [3, 4] to processes II and III, Table 1, respectively from GNF samples charging at 298 K at 11 MPa (1) and 0.1 MPa (2); the total amount of the desorbed residual hydrogen (f and y peaks) is of order of 10 wt.%
calation (or condensation) of hydrogen between graphene sheets in GNFs induced by the monolayer chemisorption of III and II types (Fig. 9). Comparing this Fig. 9 with Fig. 1 from [40] and using analytical results [3, 4], one can explain the absence of such a phenomenon both for GNFs [40] and for a number of other related experiments.
4.1. Conclusion
1. There are some experimental proofs (considered in Sections 1, 2, 3) of occurrence of the hydrogen multilayer intercalation (physisorption of a condensation or clustering type) with carbonaceous nanostructures initiated by (and/or accompanied with) chemisorption.
2. Hence, one can conclude that there is a real possibility of developing the carbonaceous «super» adsorbents [3] of hydrogen for vehicular and other applications.
3. Further studies of the hydrogen multilayer intercalation with carbonaceous nanostructures can allow solving the on-board hydrogen storage problem for the fuel-cell-powered vehicles, in particular, relevance to the DOE requirements with respect to the gravimetric and volumetric capacities.
196
International Scientific Journal for Alternative Energy and Ecology № 11(55) 2007
© 2007 Scientific Technical Centre «TATA»
II
5.1. References
1. Satyapal S., Petrovic J., Read C., Thomas G., Ordaz G. // Catal. Today. 2007. Vol. 120. P. 246.
2. Committee on Alternatives and Strategies for Future Hydrogen Production and Use, Hydrogen Economy: Opportunities, Costs, Barriers, and R&D Needs, National Academy Press, Washington, 2004.
3. Nechaev Yu. S. // Phys. Usp. 2006. Vol. 49. P. 563.
4. Nechaev Yu. S., Alekseeva O. K. // Russ. Chem. Rev. 2004. Vol. 73. P. 1211.
5. Waqar Z. // J. Mater. Sci. 2007. Vol. 42. P. 1169.
6. Denisov E. A., Kompaniets T. N., Maka-renko I. V., Vakar Z., Titkov A. N. // Materi-alovedenie (in Russian). 2003. Vol. 2. P. 45.
7. Denisov E. A., Kompaniets T. N., Kurdyu-mov A. A. // Hydrogen Recycling at Plasma Facing Materials / Ed. by C. H. Wu. Dordrecht: Kluwer Acad. Publ., 2000. Vol. 1. P. 273.
8. Denisov E. A., Kompaniets T. N. // Tech. Phys. 2001. Vol.46. P. 240.
9. Thomas C., Layet J.-M., Angot T. // Extended Abstracts, Carbon 2006, The British Carbon Group (2006). The Robert Gordon University, 1A2 Thomas, pdf. Document (CD-R).
10. San Marchi C., Somerday B. P., Robinson S. L. // Int. J. Hydrogen Energ. 2007. Vol. 32. P. 100.
11. Wong E. W., Sheehan P. E., Lieber Ch. M. // Science. 1997. Vol. 277. P. 1971.
12. Kim H., Kaufman M. J., Sigmund W. M., Jacques D., Andrews R. // J. Mater. Res. 2003. Vol. 18. P. 1104.
13. Kim H., Sigmund W. // J. Cryst. Growth.
2004. Vol. 267. P. 738.
14. Galanov B. A., Galanov S. B., Gogotsi Y. // J. Nanopart. Res. 2002. Vol. 4. P. 207.
15. Banhart F., Ajayan P.M. // Nature. 1996. Vol. 382. P. 433.
16. Han S. S., Kang J. K., Lee H. M., van Duin A. C. T., Goddard W. A. // Appl. Phys. Lett.
2005. Vol. 86. P. 203108.
17. Ye Y., Ahn C. C., Witham C., Fultz B., Liu J., Rinzier A. G., Colbert D., Smith K.A., Smalley R. E. // Appl. Phys. Lett. 1999. Vol. 74. P. 2307.
18. Allouche A., Ferro Y., Angot T., Thomas C., Layet J.-M. // J. Chem. Phys. 2005. Vol. 123. P. 124701.
19. Yang F. H., Yang R. T. // Carbon. 2002. Vol. 40. P. 437.
20. Ma Y., Xia Y., Zhao M., Ying M. // Chem. Phys. Lett. 2002. Vol. 357. P. 97.
21. Xia Y. Y., Zhao M. W., Ma Y. C., Ying M. J., Liu X. D., Liu P. J., Mei L. // Phys. Rev. B. 2002. Vol. 65. P. 155415.
22. Xia Y. Y., Zhao M. W., Ma Y. C., Liu X. D., Ying M. J., Mei L. // Phys. Rev. B. 2003. Vol. 67. P. 115117.
23. Ma Y. C., Xia Y., Zhao M., Ying M. // Phys. Rev. B. 2002. Vol. 65. P. 155430.
24. Lee S. M., An K. H., Lee Y. H., Seifert G., Frauenheim T. // J. Am. Chem. Soc. 2001. Vol. 123. P.5059.
25. Nikitin A., Ogasawara H., Mann D., Denecke R., Zhang Z., Dai H., Cho K., Nilson A. // Phys. Rev. Lett. 2005. Vol. 25. P. 225507.
26. Lupu D., Biris A. R., Misan I., Jianu A., Holzhuter G., Burkel E. // Int. J. Hydrogen Energy. 2004. Vol. 29. P. 97.
27. Yang F. H., Lachawiec A. J., Yang Jr. and R. T. // J. Phys. Chem. B. 2006. Vol. 110. P. 6236.
28. Back C.-K., Sandi G., Prakash J., Hran-isavljevic J. // J. Phys. Chem. B. 2006. Vol. 110. P. 16225.
29. Mu S.-C., Tang H.-L., Qian S.-H., PanM., Yuan R.-Z. // Carbon. 2006. Vol. 44. P. 762.
30. Gallego N. C., Baker F. S., Contescu C. I., Wu X., Speakman S.A., Tekinalp H., Edie D.D. // Extended Abstracts, Carbon 2006, The British Carbon Group (2006). The Robert Gordon University, 10A2_Gallego, pdf. Document (CD-R).
31. Gallego N. C., Contescu C. I., Baker F. S., Basova Y. V., Edie D. D. // Extended Abstracts, Carbon 2005 (Gyeoungju, Korea); Korean Carbon Society, 2005.
32. Lueking A. D., Yang R. T. // Appl. Catal. A-Gen. 2004. Vol. 265. P. 259.
33. Li Y. W., Yang R. T. // J. Am. Chem. Soc. 2006. Vol. 128. P. 726.
34. Yang R.T., Li Y., Lachawiec A. J. Jr. // DOE Hydrogen Program-2006 Annual Progress Report, U.S. Department of Energy, 2006. P. 483.
35. Lachawiec A. J. Jr., Qi G., Yang R. T. // Langmuir. 2005. Vol. 21. P. 11418.
36. Zacharia R., Rather S., Hwang S. W., Nahm K. S. // Chem. Phys. Lett. 2007. Vol. 434. P. 286.
37. Jain P., Fonseca D. A., Schaible E., Luek-ing A. D. // J. Phys. Chem. C. 2007. Vol. 111. P.1788.
38. Gupta B. K., Tiwari R. S., Srivastava O. N. // J. Alloy. Compd. 2004. Vol.381. P. 301.
39. Park C., Anderson P. E., Chambers A., Tan C. D., Hidalgo R., Rodriguez N. M. // J. Phys. Chem. B. 1999. Vol. 103. P. 10572.
40. Reichenauer G., Rzepka M., Bauer E., Schliermann T., Bernhardt B., Bohmhammel K., Henneberg E., Knoll U., Maneck H. E., Braue W. // Extended Abstracts, Carbon 2006 (Aberdeen, Scotland, UK); The British Carbon Group (2006). The Robert Gordon University, 10A1_Reichenauer, pdf. Document (CD-R).
■ 1
197
PART II
On the nanotechnology applications of the technique of protective diffusion coatings (TiO2, ZrO2, TiN, ZrN, etc) of refractory metals
(Ti, Zr, etc) and some other techniques
The nanotechnology applications of some developments and results of studying the regularities and micromechanisms of the hydrogen fluoride gas activator influence on oxidation of titanium, zirconium and zirconium-based alloys with niobium, and also — on nitriding, boriding and carbiding a series of refractory metals (Ti, Zr, Nb, Mo, W, Ta) are elaborated. The possible nan-otechnology applications of the techniques of creating a compound-like nanosegregation or the liquid-like nanosegregation at grain boundaries in nanostructured metals, i. e. creating of specific nanocomposites, are considered as well.
Introduction
Materials on the basis of refractory metals and alloys (Ti, Zr, Nb, Ta, Mo, W, etc.) are used in different branches of modern engineering (aircraft, spacecraft, atomic-power, chemical, and some other ones) and possess the necessary mechanical properties at elevated temperatures. However, they are intensively oxidized already at 850 1050 K, which is close to temperatures of brazing the material constructions. Protective diffusion coatings (of the refractory metal nitrides, oxides, silicides, borides, carbides, aluminides, etc.) are widely used for providing higher heat-resistance, wear-resistance and corrosion-resistance of metal articles and brazed constructions working in aggressive gas and liquid environments at elevated temperatures. Deposition temperatures of the protective coatings are obviously to be lower than temperatures where considerable changes of structures and the designated physical-mechanical properties of the metals occur. This is especially important for the case of nanostructured metals (particularly, to protect their recrystallysation during the deposition).
For a number of constructional materials used in modern engineering, the maximum possible decrease of the temperature-time parameters of the processes of deposition of protective coatings can only be achieved (according to results [1-5]) by using species-activators evolving hydrogen fluoride (HF) during the chemical-thermal treatment.
In this contribution, some related developments and results of studying the regularities and mi-cromechanisms of the HF-gas activator influence on oxidation of titanium, zirconium and zirconium-based alloys with niobium (Zr +1.0 % Nb; Zr + 2.5 % Nb) are presented. In addition, the influence on nitriding, boriding and carbiding a series of refractory metals (Ti, Zr, Nb, Mo, W, Ta) are presented, and their relevance for nanote-chnology applications is elaborated. Possible nan-
otechnology applications of the techniques of creating a compound-like nanosegregation or the liquid-like nanosegregation at grain boundaries in nanostructured metals, i. e. creating of the specific nanocomposites, are considered, as well.
1.2. Techniques of the chemical-thermal treatment
It is shown [1 — 5] (Table 1) that hydrogen fluoride evolving by the activator results in a several order enhancement of diffusion kinetics of low temperature oxidation of titanium and zirconium. Firstly, due to the formation and non-rate-controlling growth of an external oxide layer with high content of open porosity (in rutile, baddeley-ite). Secondly, due to fining off the microstructure of a relatively thin internal oxide layer with closed porosity (in anatase, ruffite, that is the meta-stable modifications), the thickness of which practically does not change under real expositions.
It is also noted that the rate-controlling stage of the process is the enhanced diffusion of oxygen preferably through grain boundaries of the thin internal oxide layer of near-constant thickness; the high porosity of the external oxide layer and the fining off the microstructure of the internal oxide layer are caused by internal (topo-chem-ical) reactions within lattice defect regions.
Finally, techniques [1-5] (Tables 1-3) for a large decrease of the temperature-time parameters of the chemical-thermal treatment of refractory metals and alloys, the maximum possible fining off the microstructure of the protective diffusion coatings (of oxide, nitride or other compound compositions), and a new (chemical-thermal) method for the production of ultra-dispersion powders of oxides and nitrides of refractory metals (Russian patents [4, 5]) are shown. According to results [1-5], the metal samples can be completely oxidized (or nitridized) and nanostructured up to their «spreading» into nano-crystalline (<100 nm) oxide (nitride) powders.
The obtained results [1-5] (Tables 1-3) are rather available for nanotechnology applications, and particularly, they can be used for the technology developments of the protective diffusion nano-coating of nano-structured refractory metals and production of the nano-crystalline powders of the metals' oxides or nitrides.
2.2. Techniques of creating the compound-like nanosegregation at grain boundaries
Such techniques can be based on results of a number of long-term studies [6-18] of the compound-like nanosegregation at dislocations and grain boundaries in some technologically important dilute solutions of interstitial and substitu-tional impurities in metals (H2 in Pd, Ni, Fe, steels, Nb, V, Ta; C, N2 and H2 in Fe (steels); Ni, Ti and H2 in maraging steel; O2 and Fe in Cu; N2 and Cr
198
International Scientific Journal for Alternative Energy and Ecology № 11(55) 2007
© 2007 Scientific Technical Centre «TATA»
II
Table 1.2
Growth rates at different temperatures of oxide coatings of Ti and Zr (with and without activator), and characteristics of the oxygen diffusion through the oxides
T, K D*, cm2/s Q*, kJ/mol Z*/t, ^m/h D*, cm2/s Q*, kJ/mol Z*/t, ^m/h D*, cm2/s Q*, kJ/mol Z*/t, ^m/h
Ti, khf2 Ti, [9] Ti, [10]
720 8.810 10 128 ± 6 9 1.510 13 123 0.04 2.510 14 188 0.015
770 3.510 9 35 5.910 13 0.08 1.910 13 0.043
820 1.210 8 120 1.910 12 0.14 1.210 12 0.10
Zr, KHF2 Zr, [9] Zr, [10]
670 4.510 10 76 ± 7 16.5 7.610 13 76 0.165 4.010 13 151 0.12
720 1.210 9 48 2.010 12 0.265 2.7-10 12 0.31
770 2.610 9 103 4.510 12 0.40 1.410 11 0.71
N o t e. D* — the apparent diffusion coefficient, and Q* — the apparent activation energy. Herein, references are corresponded to those of [2]. * — Similar results were also obtained for alloys (Zr +1.0 % Nb, Zr + 2.5 % Nb).
in Ni; Fe, Mn, Cr, Ti, V, Zr, Sc in Al) and their influence on duiffusion assisted processes (the internal oxidation, internal nitridation, precipitate coarsening, recrystallization, the thermal-hydrogen treatment, cracking, blistering, delayed fatigue (fracture), plastification or embritlement).
For instance, as has been shown in [6], the structure, composition, diameter (up to several nm), contribution to electrical resistance, thermodynamic and diffusion characteristics of the hydride-like nanosegregation at dislocations in Pd can vary in a wide diapason depending on concentration of hydrogen dissolved in the normal lattice metal (that is the so called lattice hydrogen). The formation of the hydride-like nanosegregation at dislocations in Pd takes place at large degrees (several orders) of the undersaturation of the solid solutions with respect to the hydride precipitates within the normal lattice of palladium. It means that a specific phase diagram can be considered for the system of «hydrogen - near-dislocation segregation nanore-gions of Pd», in comparison with the generally used conventional phase diagram of «hydrogen -palladium».
Hence, it follows a technological possibility for the development of nanosructured Pd due to hydride-like nanosegregation at the grain boundaries (i. e. a specific nanocomposite).
A similar consideration of results [7-18] can show a technological possibility for the development of such specific nanocomposites of other na-nostructured metals (Fe, steels, Al, etc.) with complex hydride-like, carbide-like, nitride-like, oxidelike or intermetallide-like nanosegregation at grain boundaries.
A technological possibility [19] for the diffusion transport enhancement is the intensive deformation of metallic materials (both for the metal nanostructuring, and for the impurity fast redistribution between the nanograin bulk and boundary regions).
Table 2.2
The diffusion-controlled Ti nitriding (at PN, = 120 kPa), [1]
Activator T, K D, cm2 s-1 D0, cm s Q, kJ mol-1
Without activator 870 970 1070 8.8-10-13 5.8-10-12 2.7-10-11 8.5-10-5 133 i 6
khf2 870 920 970 5.8-10-8 9.6-10-8 1.5-10-7 6.1-10-4 67 i 7
N o te . D — the diffusion coefficient, D0 — the per-exponential factor of D, and Q — the activation energy). * — Similar results were also obtained for zirconium.
Table 3.2
The diffusion-controlled Nb boriding, [1]
T, K Saturation D, 10-11 Do, Q, 1
medium cm s cm s kJ mol
1173 0.7 150 300
1223 Boron carbide 2.3
1273 (B4C) and 7.3
1323 carbon 21
1373 58
1073 0.5 230 280
1123 Boron carbide 2.2
1173 7.8
and carbon with AlF3 activator
1223 25
1273 75
1323 200
1373 510
N o t e. D — the diffusion coefficient, D0 — the per-exponential factor of D, and Q — the activation energy.
3.2. Techniques of creating the liquid-like nanosegregation at grain boundaries
Several technological possibilities [20-25] of creating the liquid-like state in the grain boundary nanoregions in metals can be used. But it seems to be more perspective to use another technologi-
Fig. 1.2. Penetration of liquid Bi through grain boundary nanoregions in Cu; the penetration depth is about 7 mkm after exposition of 8 h at 723 K, it is about 12 mkm after exposition of 8 h at 773 K, and it is about 30 mkm after exposition of 4h at 823 K, [26]; the temperature dependence of the penetration depth corresponds to the diffusion apparent activation enthalpy (AH) of about 2 eV
cal possibility which is based on using experimental data, for instance [26-31], on liquid grooving (deep local wetting), liquid-metal very deep etching of the grain boundary regions (Fig. 1), and crack formation induced by boundary wetting in metallic systems (i. e. Cu-Bi, Al-Sn, Ni-Bi, Al-Ga, Al-In, Cu-In, W-Ni).
An analysis of experimental data [26-31] for the Cu-Bi system (at different temperatures) has been performed: (1) on the liquid deep grooving in grain boundary regions (LDGGB) of a «finger» form canal of micrometers' (near constant) thickness and parabolic time dependence of the canal length growth, (2) on the liquid-metal etching in
Fig.2.2. Penetration (discontinuous «nets») of liquid Bi through grain boundary nanoregions in Cu; the penetration depth (L) is about 17 ßm after exposition of 8 h (t) at 723 K (T), it is about 70 ßm after exposition of 8 h at 773 K, and it is about 120 ßm after exposition of 4 h at 823 K, [26]; the Arrhenius type temperature dependence of the quantity of (L?/t) corresponds to the diffusion apparent activation enthalpy (AH) of about 2 eV
Fig. 3.2. Penetration (through the triple junctions) of liquid Bi in grain boundary nanoregions in Cu; the penetration depth (L) is about 30 /m after exposition of 8 h (t) at 723 K (T), it is about 240 /m after exposition of 8 h at 773 K, and it is about 320 /m after exposition of 4 h at 823 K, [26]; the Arrhenius type temperature dependence of the quantity of (L2/t) corresponds to the diffusion apparent activation enthalpy (AH) of about 2 eV
grain boundary regions (L-MEGB) of near constant nanometers' canal thickness and parabolic time dependence of the length (much deeper than LDGGB), (Fig. 2) and (3) on the liquid-metal etching in the triple junction grain boundary regions (L-METJ), (Fig. 3) of near constant nanometers' canal diameter and parabolic time dependence of the canal length (much deeper than L-MEGB).
The data treatment [26-31] resulted in anomalously high values (of order ~2 eV) of the apparent activation energies of the diffusion mass transfer, and anomalously low values of the apparent diffusion coefficients (Figs. 1-3) in comparison with the related known data on the diffusion characteristics for the melts, grain boundaries and triple junctions, correspondently (in metallic systems without grain boundary liquid grooving).
It should be also taken into account some experimental data [26] (for instance, for Bi-Cu system on the diffusion mass transfer of copper from the solid metal phase (Cu) through the grain boundary Bi-Cu liquid regions (canals) to the Bi-based melt, which is in contact with Cu-phase.
The above facts point to existence of the stage of dissolution of solid materials (Cu), mainly, at the canals' tips (because of the fact of near-constant canals' thickness or diameter), and the stage of its (Cu) diffusion through the (Bi-Cu) melt canals to the contacting (Bi-based) melt phase.
The treatment results (the main features and/ or aspects of the phenomenon in question) can be interpreted and quantitatively described on the basis of using the concept of the atomic diffusion mobility and the Onsager approximation of the ther-modynamic force. The simplest model of the anomalous diffusion mass transfer can be developed by
International Scientific Journal for Alternative Energy and Ecology № 11(55) 2007
© 2007 Scientific Technical Centre «TATA»
using the model (similar to one [6]) of diffusion accompanying with reversible trapping of the melt diffusant (Cu in Bi-Cu liquid canals) by the involved nano-surroundings of the solid metal (Cuphase). The possibility of the atomic scale smoothing of the canals' walls and its influence on the solution-dissolution trapping stage should be taken into account in the kinetic model in question.
In such a model, within the local equilibrium approximation for the diffusant (Cu) distribution between the liquid Bi-Cu nanophase (in the grain boundary canals) and the solid Cu-based nano-phase of the canal surroundings, the apparent diffusion coefficient of Cu can be described (by using Eq. (11) from [6]), as follows:
D*--
[(1 - По )D,lq + Попы (soildCi,q )] [i -Пы, +Пы (o,/dCi,q )]
(1.2)
hsol hsol + hli
Dsol <<-
(1 -nsol )Dliq
and
dC
sol
dC,
liq
■■A exp (AH/kT )
On the basis of the above noted facts, one can suppose that the canals' tips' nanoregions of the solid metal (Cu) are not so much of the atomic scale smoothing (in comparison with the Cu canals' walls). Therefore, one can suppose (i) that the main dissolution process is localized at the canals' tips, (ii) that the dissolution process for the canals' walls is negligible (in comparison with the canals' tips), and (iii) that the majority of Cu atoms come to the to the contacting (Bi-based) melt phase from canals' tips' nanoregions (resulting in the near-constant canals' thickness).
Hence, by using the concept of the atomic diffusion mobility (with taking into account the apparent diffusion coefficient of D*) and the Onsag-er approximation of the thermodynamic force, one can describe the noted above parabolic time dependence of the canal length growth, as follows:
where Dliq is the true diffusion coefficient of Cu atoms in the liquid (melt) Bi-Cu nanophase (in the grain boundary canals), Dsol is the true diffusion coefficient of Cu atoms in the solid Cu-based nano-
L - D * (ACiiqICiiq )t,
phase of the canal surroundings, nsoi
"sol + "iiq
the local volume fraction of the solid Cu-based nan-ophase, hsol is the solid nanophase thickness, hliq is the liquid nanophase thickness, Csol is the atomic fraction of Cu in the solid nanophase, Cu is the atomic fraction of Cu in the liquid nanophase.
The above noted results of the data treatment [26] (Figs. 1-3) can be described with the help of
Eq. (1) in the case of [nsoi (soi/dCllq )]>> 1,
nsol (soildCl,q)
where AH - 2 eV, A - const, k — the Boltzmann constant, T — the temperature in K; such a value of the enthalpy change (AH) can be caused by the atomic scale smoothing of the canal walls. Hence, it follows that
D* - {[(1 -no )/Msoi ]Diiq exp (-AH/kT )}, (2.2)
where AH corresponds to the diffusion apparent activation enthalpy (energy).
In such a model, the quantity of AH - 2 eV corresponds to enthalpy change for the transition (dissolution) of copper atoms from the solid Cu-based nanophase of the canal surroundings (obviously, with the atomic scale smoothing of the Cu canals' walls) to the liquid Bi-Cu nanophase in the grain boundary canals.
A standard value of the dissolution enthalpy of a solid copper phase (without an atomic scale smoothing of its surface) in Vi-Cu melt can be evaluated from the Cu-Bi constitution (phase) diagram [26] as AHdiss - 0.5 eV.
(3.2)
where ACiiq — is the difference in the diffusant (Cu) concentration in the liquid Bi-Cu nanophase in the grain boundary canals at the near-tips' regions and the near-contacting (Bi-based) melt phase regions, Ciiq — is the average concentration.
The possible influence of the local volume changes (due to the «solid-liquid» and/or «liquid-solid» phase transition) and the induced local stresses and local strains can be considered by using results [16-18], for the interpretation of the nano-crack formation data.
Such results can be obviously applied for na-notechnology applications, and particularly, for creating of nanostructured metals with liquid-like nanosegregation and/or nanophase at grain boundaries, in order to stabilize the nanostructures and to obtain specific physical and mechanical (as super plasticity or embrittlement) properties of such nanocomposites.
4.2. Conclusion and proposal on cooperation
Firstly, one can use the techniques [1-5] (Tables 1-3) of a large decrease of the temperature-time parameters of the chemical-thermal treatment of nanostructured refractory metals and alloys, particularly, in order to stabilize nanostructures and to produce specific nanocomposites.
Secondly, techniques [1-5] of the extreme microstructure fining of the protective diffusion coatings (of oxide, nitride or other compound compositions) can be used.
Thirdly, a new (chemical-thermal) method [5] for the production of ultra-dispersion powders of oxides and nitrides of refractory metals can be used.
Techniques [6-19] can be used for the creation of nanostructured metals with compound-like na-nosegregation at grain boundaries, in order to stabilize the nanostructures and to obtain the specific physical and mechanical properties of such nano-
composites. It can be, for instance, the complex hydride-like, carbide-like, nitride-like, carbide-nitridelike, oxide-like or intermetallide-like nanosegrega-tion at grain boundaries and/or dislocations.
Techniques [20-31] can be used for the creation of nanostructured metals with liquid-like na-nosegregation at grain boundaries, in order to stabilize the nano-structures and to obtain the specific physical and mechanical (superplasticity or embrit-tlement) properties of such nano-composites. The known numerous data [26-31] (Figs. 1-3), and the above presented interpretation, on the liquid-metal deep etching of the grain boundary nanore-gions in metallic systems (i. e. Cu-Bi, Al-Sn, Ni-Bi, Al-Ga, Al-In, Cu-In, Ni-W) can be used.
Acknowledgements
This study was made possible by the financial support of the Russian Foundation for Basic Research (Grant 05-08-50222-a).
5.2. References
1. Kostikov V. I., Nechaev Yu. S., Kul'-gaG. Ya. // Dokl. Phys. 2001. Vol.45. P. 161.
2. Kostikov V. I., Nechaev Yu. S., Kul'-gaG. Ya. // Russ. J. Phys. Chem. 2002. Vol. 76. P. 1429.
3. Kostikov V. I., Nechaev Yu. S., Kul'-gaG. Ya. // Perspect. Mater. 2001. Vol.4. P. 85 (in Russian).
4. Russian patent license 2198955 C2 (2003). Kostikov V. I., Nechaev Yu. S., Kul'ga G. Ya. A method of production of the protective diffusion coatings on metals and alloys.
5. Russian patent license 2212983 C2 (2003). Kostikov V. I., Nechaev Yu. S., Kul'ga G. Ya. A method of production of nano-crystalline powders of chemical compounds.
6. Nechaev Yu. S. // Phys. Usp. 2001. Vol. 44. P. 1189.
7. Nechaev Yu. S., Ephimenko S. P. // Met. Phys. Adv. Tech. (Reprint). 2001. Vol. 19. P. 225.
8. Nechaev Yu. S., Filippov G. A. // Defect Diffus. Forum 2001. Vol. 194-199. P. 1099.
9. Nechaev Yu. S., Egziabher K. H. G. // Phys. Stat. Sol. A. 1988. Vol. 106. P. 399.
10. Nechaev Yu. S. // Defect Diffus. Forum. 1989. Vol. 66-69. P. 881.
11. Ephimenko S. P., Nechaev Yu. S., Karelin F. P., Portnaya O.N. // Phys. Chem. Mater. Treat. 1997. Vol.5. P. 101.
12. Nechaev Yu. S. // Defect Diffus. Forum. 2006. Vol. 251-252. P. 111.
13. Nechaev Yu. S., Yewondwossen M. H. // J. Phys. F. Met. Phys. 1987. Vol. 17. P. 1081.
14. Nechaev Yu. S., Leont'ev V. G., Mezhen-nyy Yu. O., Bubenchikov A. S. // Phys. Met. Metall. 1985. Vol. 59. P. 106.
15. Nechaev Yu. S., Krupin Yu. A., Mezhen-nyy Yu.O., Pustov Yu.A., Rastorguev L.N. // Phys. Met. Metall. 1984. Vol. 58. P. 115.
16. Nechaev Yu. S. // Defect Diffus. Forum. 2001. Vol. 194-199. P. 1713.
17. Nechaev Yu. S. // Met. Phys. Adv. Tech. (Reprints). 2001. Vol. 19. P. 533.
18. Nechaev Yu. S., Kodentsov A.A., van Dal M., van Loo F. J. J. // Met. Phys. Adv. Tech. (Reprints). 2001. Vol. 19. P. 107.
19. Nechaev Yu. S. // Defect Diffus. Forum. 2006. Vol. 251-252. P. 123.
20. Vishnjakov Ya. D., Nechaev Yu. S., Vla-dimirov S.A., Pustov Yu.A. // Metall. Trans. Acad. Sci. 1980. Vol. 4. P. 174.
21. Nechaev Yu. S. // Phys. Chem. Mater. Treat. 1992. Vol. 26. P. 610.
22. Nechaev Yu. S. // Colloq. Phys., Colloque C1. 1990. Supplément au № 1, Vol. 51. P. 287.
23. Nechaev Yu. S. // T. Russ. Acad. Sci., Phys. Ser. 2001. Vol. 65. P. 1399.
24. Nechaev Yu. S., Iourtchenko D. V., Hirschberg J. G., Veziroglu T. N. // Int. J. Hydrogen Energ. 2004. Vol.29. P. 1421.
25. Nechaev Yu.S., Burzhanov A.A., Filippov G. A. // Adv. Mater. Sci. 2007. Vol. 7. P. 166.
26. Petelin A. L. Dr. Sc. Thesis, Moscow Steel and Alloys Institute, 2007.
27. Bokstein B., Kostel'tseva N., Peteline A., Rakov S., Rodin A. // Defect Diffus. Forum. 2005. Vol. 237-240. P. 756.
28. Apikhina I., Gulevskii S., Dolgopolov N., Peteline A., Rakov S., Rodin A. // Ibid. P. 855.
29. Dolgopolov N., Petelin A., Rakov S., Rodin A. // Defect Diffus. Forum. 2006. Vol.249. P. 227.
30. Gulevskii S., Kostel'tseva N., Petelin A., Podgornii D., Rodin A., Smirnov A. // Izvestija VUZov. Tsvetnaja Metallurgija. 2005. Vol. 3. P. 71.
31. Apikhina I., Bokstein B., Gulevskii S., Kozlova O., Peteline A., Rakov S., Rodin A. // Defect Diffus. Forum. 2003. Vol. 216-217. P. 181.
202
International Scientific Journal for Alternative Energy and Ecology № 11(55) 2007
© 2007 Scientific Technical Centre «TATA»
II
