Maria A.
Gunawan
ab,
Didier
Poinsot
a,
Bruno
Domenichini
c,
Céline
Dirand
c,
Sébastien
Chevalier
c,
Andrey A.
Fokin
bd,
Peter R.
Schreiner
*b and
Jean-Cyrille
Hierso
*ae
aInstitut de Chimie Moléculaire de l'Université de Bourgogne (ICMUB), UMR-CNRS 6302, Université de Bourgogne, 9 avenue Alain Savary, 21078 Dijon Cedex, France. E-mail: hiersojc@u-bourgogne.fr
bInstitut für Organische Chemie, Justus-Liebig-Universität, Heinrich-Buff-Ring 58, 35392 Giessen, Germany. E-mail: prs@uni-giessen.de
cLaboratoire Interdisciplinaire Carnot de Bourgogne (LICB), UMR-CNRS 6303, Université de Bourgogne, 9 avenue Alain Savary, 21078 Dijon Cedex, France
dDepartment of Organic Chemistry, Kiev Polytechnic Institute, Pr. Pobedy 37, 03056 Kiev, Ukraine
eInstitut Universitaire de France (IUF), 103 Bd. Saint Michel, 75005 Paris Cedex 5, France
First published on 8th December 2014
We detail herein readily accessible processes to control previously unobserved robust self-assemblies of nanodiamonds (diamondoids) in micro- and nanocrystals from their mild vapor deposition. The chemical functionalization of uniform and discernible nanodiamonds was found to be a key parameter, and depending on the type of functional group (hydroxy, fluorine, etc.) and its position on the diamondoid, the structure of the discrete deposits can vary dramatically. Thus, well-defined anisotropic structures such as rod, needle, triangle or truncated octahedron shapes can be obtained, and self-assembled edifices of sizes ranging from 20 nm to several hundred micrometers formed with conservation of a similar structure for a given diamondoid. Key thermodynamic data including sublimation enthalpy of diamondoid derivatives are reported, and the SEM of the self-assemblies coupled with EDX analyses and XRD attest the nature and purity of nanodiamond crystal deposits. This attractive method is simple and outperforms in terms of deposit quality dip-coating methods we used. This vapor phase deposition approach is expected to allow for an easy formation of diamondoid nanoobjects on different types of substrates.
![]() | ||
Fig. 1 Fully hydrogenated hydrocarbon cages adamantane 1 and diamantane 2, and functionalized diamantane analogues 3–5. |
Another attractive and hitherto unmet challenge in this field is the construction of organic materials and organohybrids based on nano- and microcrystals of functionalized diamondoids. Based on such edifices, carbon nucleation followed by growth may also pave the way to access continuous diamond thin films of better defined structure and surface.14 Diamondoids may also serve as precursors in “bottom-up” strategies to build organohybrids,15 and to devise diamond structures under conditions that are possibly milder than the chemical vapor deposition (CVD) conditions currently used for diamond growth.16 Methods have been reported akin to CVD that yielded higher diamondoids (more than four carbon fused-cages) and even diamond from diamondoid seeds.17
However, for applications of nanodiamond-based materials built “bottom-up” there is still a need for better fundamental understanding and mastering of the conditions of nucleation, growth and self-assembly of diamondoid units.18 Additionally, general interest in dispersed micro- and nanocrystals stems from the fact that their properties dramatically change with their size and shape,19 and many applications from optical (plasmon absorption) to medical (transportation through biological barriers) illustrate this point. We thus report on readily accessible processes for the mild vapor deposition of functionalized diamondoids that provide robust self-assembly of nanodiamonds micro- and nanocrystals. Key thermodynamic data including sublimation enthalpy of several diamondoid derivatives are reported from a new measurement protocol at solid–vapor thermodynamic equilibrium state. Easily accessible processes for self-assembled edifices of sizes ranging from 20 nm to several hundred micrometers are reported with remarkable conservation of the general geometry for a given diamondoid, chemical analysis with microscopy ascertaining the full preservation of the nature of nanodiamonds.
Clausius–Clapeyron equations | ΔH0 (kJ mol−1) | r 2 | |
---|---|---|---|
a Pressure in Pa, T in K (see ESI for measurement full details). | |||
1 | ln![]() |
60.0 ± 5 | 0.9991 |
2 | ln![]() ![]() |
92.1 ± 5 | 0.9992 |
3 | ln![]() ![]() |
102.1 ± 5 | 0.9960 |
4 | ln![]() ![]() |
108.8 ± 5 | 0.9983 |
5 | ln![]() ![]() |
96.4 ± 5 | 0.9965 |
The enthalpies of sublimation obtained from these equations are ΔH0 = 60.0 kJ mol−1 for 1, 92.1 kJ mol−1 for 2, 102.1 kJ mol−1 for 3, 108.8 kJ mol−1 for 4, and 96.4 kJ mol−1 for 5, indicating the highest volatility for adamantane, and the lowest for hydroxydiamantane derivatives. The enthalpy of sublimation obtained for adamantane 1 following equilibrium state pressure measurements, ΔH0 = 60.0 kJ mol−1 (14.3 kcal mol−1), is consistent with the values previously reported (14.2 ± 0.3 kcal mol−1).21a This was also the case for diamantane 2 measurements (ΔH0 = 92.1 kJ mol−1), with the values obtained using calorimetry bomb combustion,21a and gas-saturation temperature scanning (95.9 kJ mol−1).23 As expected, the fluorodiamantane 5 was found to be fairly volatile with a ΔH0 = 96.4 kJ mol−1 approaching the one we found for pristine diamantane. Due to a noticeable hygroscopic behaviour of the hydroxydiamantanes a good reproducibility required cautiously dried finely divided pure hydroxydiamantane powders to exclude water traces. It was determined that 4-hydroxydiamantane 4 is significantly less volatile, ΔH0 = 108.8 kJ mol−1, than 1-hydroxydiamantane 3, ΔH0 = 102.1 kJ mol−1.24
The vapor deposition of 1–5 was then conducted, under tuneable conditions of pressure and temperature, on silicon and mica substrates using a simple sublimation apparatus (see ESI†). The deposition experiments were conducted first under air or argon atmosphere 1.0 bar (105 Pa), and then either under static vacuum (initial pressure 5.3 mbar) or dynamic vacuum (5.3 mbar). Temperature conditions ranging between 50 and 110 °C were investigated. Higher sublimation temperatures favour a higher vapor supersaturation regime, which in turn favours effective nucleation in comparison with crystal growth. A good nucleation rate, in comparison with growth rate, then will quickly generate numerous small crystals of critical size that will no longer disappear and serve as “nutrients” for the growth of bigger crystals (Ostwald ripening).25 Deposition times were varied between a few minutes to 16 h. The vapor deposition of a wide variety of functionalized diamondoids was conducted (see ESI†), including compounds 1–5 (Fig. 2).
Other functionalized adamantane and diamantane derivatives, such as the 1-aminoadamantane, 1-adamantyldiphenylphosphinite, 1-adamantyldiphenyl-thiophosphinite, 1-ethynyladamantane, 1-hydroxyadamantane, 4,9-dihydroxydiamantane, 4,9-difluorodiamantane were also deposited (see details in ESI†). Optical microscopy of the deposits was conducted for visual inspection of the obtained deposits. Optimization of the deposition conditions was conducted for the diamondoids that were found most suitable for generating from vapor deposition well-defined crystals self-assembly. As illustrated in Fig. 2, crystals of a micrometer size with very different shapes and structures grew depending on the nature of the diamondoid. The deposition of 1 was difficult due to the high volatility of this compound which easily re-evaporates after condensation. The vapor deposition of 2, although easier, resulted in very fragile assemblies that easily broke on handling (see microstructures in ESI†). Conversely, nice crystalline deposits of well-dispersed homogeneous microcrystals were obtained in less than one hour from the vapor deposition of hydroxyl- and fluoro-functionalized nanodiamonds 3–5. Interestingly, not only the nature of the functionalization shows an overwhelming influence on the self-assembly and thus microstructure of the deposits, but also the position of the functional group on the cage. Thus, from the vapor deposition of 3–5, hexagonal rod, rectangular needle, and triangular plate microcrystals were observed, respectively. The size of the crystals can be adjusted through the duration of the deposition with conservation of excellent dispersion and size homogeneity. The high crystallinity of the deposits was confirmed by successful single crystal X-ray diffraction analyses conducted on all these samples. For instance, the X-ray diffraction analysis of microcrystals of 3 (Fig. 17S in ESI†),26 shows the presence of discrete tetramers for which the units are in a cyclic network of O–H⋯O hydrogen bonding. These tetramers are arranged in 1-D columns. A similar structure was found for 4.
Scanning electron microscopy (SEM) analysis was conducted to examine surface morphologies in further details (Fig. 3). Much to our surprise the SEM images showed that the first deposits elaborated under atmospheric pressure of air or argon do not tolerate the low pressure, 10−6 mbar, and the voltage imposed by SEM analysis. For instance, deposits of 5 were completely volatilized, and while self-assemblies of 3 and 4 were more resistant, they also showed a tendency toward shrinkage during the analysis. Microcrystals of 3 that appeared as well-defined hexagonal rods in optical microscopy turned into hollow rectangular structures under the SEM conditions (Fig. 3a). The formation of hollow needles for 4 is shown in the SEM micrographs (Fig. 3b) with shrinkage also evidenced for the needle crystals from collapsed edges shown in the enlarged picture (right). Hollow micro- and nanostructures are studied in areas which include catalysis, cosmetics, drug and gene delivery, hydrogen production and storage, photonics, photovoltaics, and rechargeable batteries.27 Mostly metal-based and ceramic hollow micro-/nanostructures have been prepared, commonly from template-removing procedures. To the best of our knowledge, the formation of non-polymeric hollow structures of pure hydrocarbons has not previously been documented. We hypothesized that their formation was either due to some unlikely selective internal volatilization and/or rearrangement of the self-assembly under the SEM conditions, or better to some gas capture –either air or argon– inside the structures during the growth process. To support the latter hypothesis, we tried to protect the diamondoid particles grown under atmospheric pressure by covering them with a 15–20 nm thin layer of gold metal before SEM analysis using sputtering by high resolution ion beam coater. This metallization was also achieved under high vacuum (10−5 mbar) and consequently the crystal morphologies were modified before the gold layer could fully protect the deposit. However, these experiments established that during the vapor deposition process under atmospheric pressure some gas was trapped inside the self-assembly of diamondoids. As shown in Fig. 3c and d respectively, shrinkage of the polyhedral particles based on triangle motifs for 5, and hexagonal rods for 3 were observed, while the gold protective layer roughly conserved the original shape of the edifices.
With the goal of growing crystals in which no gas trapping occurs and that would also be resistant to the SEM conditions, we modified the vapor deposition apparatus used under atmospheric pressure to allow sublimation under 5 mbar reduced pressure (Fig. 16S in ESI†). We were delighted to confirm the success of this strategy. As illustrated in Fig. 3e and f, the deposition of microcrystals of about 40 μm was achieved, with the SEM revealing three-dimensional structures fully consistent with the optical microscopy observations. The triangular shapes observed in optical microscopy thus correspond to the faces of regular octahedra which result from the self-assembly from the vapor phase of fluorinated diamondoid 5. The hexagonal rods observed for hydroxylated diamantane 3 revealed an apparent surface roughness that was not observed for 5. EDX microanalysis conducted during the SEM experiments on various areas of the deposited samples confirmed elemental composition and nature of the deposits. This is illustrated in Fig. 4 that presents a deposit of 5 on silicon (area-Spect 3, EDX CKα1 = 0.280 keV, FKα1 = 0.532 keV) and compares it to a pristine zone of the silicon substrate (area-Spect 2, EDX SiKα1 = 1.739 keV). Thus, consistent with the single crystal X-ray diffraction analysis that attested high crystallinity and purity of the deposits, the peaks corresponding to the electron binding energies Kα1 for C, F and Si were detected depending on the focus area of the beam. The size distribution in Fig. 4 confirmed an average size of 40 μm with the major part of the particles having a size between 25 and 55 μm. More than 90% of the visible aggregates have pyramidal and octahedral shapes, and less than 5% edifices experienced intense Ostwald ripening for reaching sizes above 60 μm.
Heavier plasma-hydrogenated detonation nanodiamond developed by Arnault and co-workers have shown a tendency to self-assembling as aggregates in water solution, depending on counter ions present.28 Our attempts to generate accurate and reproducible self-assemblies such as obtained from vapor phases, but by crystallization from our functionalized diamondoids in solution (dip-coating, dichloromethane, water) were unsuccessful, giving mainly a large area covered with isotropic structurally poorly-defined coated material (see optical microscopy and SEM images in ESI†).
After having determined practical vapor deposition conditions for growing well-dispersed self-assembly of diamondoids as microcrystals – with various shapes depending on the diamondoid functionalization and on the position of the function on the cage – we addressed the more challenging task of growing crystals of nanometer size by a similar vapor deposition technique. A particularly pertinent issue was whether shapes and structures previously obtained from functionalized diamondoids vapor self-assembly would be conserved at the nanoscale. The very first step of deposition process could not be controlled by the simple evaporation apparatus we designed for fast and easy self-assembly of diamondoids as microcrystals. Instead we thus used a three-chamber high vacuum physical vapor deposition (PVD, 10−9 mbar) apparatus incorporating in situ XPS analysis (see Fig. SXPS2 in ESI,† from 3 a typical C 1s peak at 284.0 eV appears). The vapor deposition of 3 under suitable vapor pressure was performed using this PVD apparatus, and led to the growth of well-defined crystalline hexagonal rods. Diamondoid vapours were brought very close to the substrates through a heating tube placed at 2 mm distance. Deposition on a silicon surface that was treated following a RCA protocol to remove carbon contamination was achieved from 20 mg of 3 heated at 100 °C for 3 h under a low pressure of 1.4 × 10−6 mbar. SEM images of the resulting deposits are provided in Fig. 5. As expected, these deposits were resistant to the SEM analysis conditions. Fig. 5 (bottom right) provides a representative histogram of the nanocrystals dispersion. Generally, sizes ranging between 20 and 600 nm were obtained under these conditions. Discrete 30–100 nm nanocrystals were obtained in majority (>70% of the particles). We were glad to observe that the self-assembly of 3 mostly exhibits hexagonal shapes with well-defined growth directions. These growth directions were not induced by the Si(111) substrate since an amorphous SiO2 thin layer is always present at the surface of the substrate before deposition as confirmed by XPS (with O 1s peak of SiO2 found at 532.5 eV, see Fig. SXPS1 in ESI†). Accordingly, owing to non-preferentially oriented macroscopic growth of nanocrystals some were positioned perpendicularly to the surface.
In the SEM picture Fig. 5 (left) these diamondoid particles can be easily identified and the images allowed measuring their thicknesses, which were remarkably regular around 20 nm. This regularity suggests a fairly homogeneous growth rate of the diamondoid particles. EDX analysis with a beam focusing on the edifices (beam energy 5 keV, size 1 μm3) showed that the deposit of 3 exclusively contains carbon and oxygen, in full agreement with the chemical nature of the deposit (see Fig. SEDX1 in ESI†). We therefore demonstrated that particles of 3 of approximately 20 nm thickness and size can be deposited from a vapor phase under strictly anaerobic and low-pressure conditions (1.4 × 10−6 mbar), showing self-assembly into hexagonal structures similar to the microstructures formed at micrometer scale under few mbar pressure. This could be straightforwardly extended to the other nanodiamonds 4 and 5.
The relation between the self-assemblies morphology and the functional groups in the molecular precursor is for now difficult to rationalize. This is related to the recognized general non-predictability of molecular structure and morphology of crystals. Accordingly, complexity in polymorphism of functionalized diamondoids has been recently illustrated for adamantane halide derivatives.29 Nevertheless, it is worth mentioning the excellent reproducibility which is obtained from this approach in the formation of robust well-defined self-assemblies. This attractive method is very simple and in our case outperforms in terms of quality dip-coating popular methods (see ESI†). Accordingly, this mild temperature vapor phase deposition approach is expected to allow for an easy formation of diamondoid nanoobjects on different types of substrates, including sensitive ones.
Footnote |
† Electronic supplementary information (ESI) available: Full details of vapor pressure measurements, vapor deposition of functionalized diamondoids, apparatus (including PVD) and all characterization for all functionalized diamondoids self-assemblies investigated from vapor phases and solution dip-coating. See DOI: 10.1039/c4nr04442h |
This journal is © The Royal Society of Chemistry 2015 |