RICERCA ITALIANA     

Ceramic nanocomposites from preceramic precursors and carbon nanotubes

Università degli Studi di Trieste

Abstract

This research project deals with the preparation and the structural, physical and mechanical characterizazion of novel ceramic nanocomposites made of a silicon oxycarbide matrix and carbon nanotubes (CNTs).
Silicon oxycarbide glasses (SiCO) belong to the family of Polymer Derived Ceramics (PDCs). PDCs have shown excellent properties: they resist crystallization up to extremely high temperatures (up to 1800°C) and they are creep and oxidation resistant.
The PDCs are processed by a novel, low-cost polymer-based process which consists of two essential steps. First, the organic precursor is polymerized into a rigid plastic form, for example fiber or film. These forms are then pyrolyzed under controlled conditions at about 1000°C to convert them into the ceramic.
The introduction of CNTs into the PDC matrix will allow to improve the excellent properties of these materials, in particular their mechanical, thermal and electrical properties. Indeed, CNTs may be the strongest materials known and they also have very high thermal and electrical conductivity.
The project will be developed in the following steps:
1) The research unit of Bologna (henceforth, UOBO) will study the optimal procedure for the dispersion of the CNTs in preceramic precursors obtained by (i) the sol-gel method (research unit of Trento, UOTN) or (ii) from silicone resins (UOBO);
2) Characterization of the structural evolution during pyrolysis by means of: DTA/TG, Dilatometry, FT-IR, XRD, SEM (UOTN e UOBO); TG coupled with Mass Spectrometry (TG/MS) and surface area measurements (BET) (UOTN);
3) Study for the production of specific shapes: fibers (UOTN), foams and films (UOBO) and bulk (UOTN and UOBO);
4) Structural (XRD, SEM) and mechanical characterization of the nanocomposite components: (UOTN, UOBO);
5) Fine structural characterization via micro Raman spectroscopy, HRTEM, Micro-ESCA (research unit of Trieste, UOTS);
6) Mapping of the residual stresses by means of micro-Raman (piezo-spectroscopyc effect) (UOTS);
7) Characterization of the electrical and thermal properties (UOTN, UOBO, UOTS).
The present project combines two groups active in the field of PDCs, UOTN and UOBO, and one group expert in the fine structural characterization of ceramic materials via Raman spectroscopy. UOTN is leader in the field of SiCO glasses from gels and has a specific expertise in fibers and bulks processing. Sol-gel process allows an excellent control of the composition and architecture of the preceramic polymer as well as easy addition of extra elements such as B, Al, Ti etc. UOBO is leader in the development of SiCO foams and joints from commercial silicone resins.
UOTN and UOBO, will use their complementary expertise for the production, for the first time, of the novel CNTs/SiCO nanocomposites and UOTS will contribute to the structural characterization with two fundamental techniques: Raman spectroscopy and high resolution transmission electron microscopy.
The innovative character of these nanocomposites might then lead also to patent applications.

In its essential guidelines, this project has already been presented to the Ministry for University and Research in the call for proposal PRIN (COFIN) 2004. We have chosen to re-submit it for two reasons:
i) in 2004 both reviewers ranked the project in the best group (class A). The project was not sponsored only because of lack of governmental funds;
ii)UOTS is acquiring a newly developed instruments, to date unique in Italy, which is particularly suitable for the characterization of composites with CNT's. This instrument (which is an integration of Atomic Force Microscopy, AFM, Scanning Near-field Optical Microscopy, SNOM, and Raman spectroscopy) will be presented in section 2.2 and will be discussed at length in the specific project of UOTS. The uniqueness of this instrument fully complies with the requirement of research of relevant national interest explicitly cited in the call for proposal PRIN (COFIN) 2005. <<<

Principal Investigator

Valter SERGO Università degli Studi di TRIESTE

Research Objectives

The goal of the research project is the development, for the first time, of a new class of ceramic nanocomposites from preceramic precursors (PDCs) and carbon nanotubes (CNTs). In particular we will develop CNTs/SiCO nanocomposites.
We also propose to develop specific components (such as fibers, films, foams and bulk samples) which might increase even more the expected properties of the PDCs/CNTs nanocomposites. These components will also be used to measure the physical/mechanical properties of these novel materials.
Achieving the following intermediate objectives will pursue the goal:
1. Study of the optimal conditions for the insertion of CNTs in preceramic precursors (either sol-gel solutions and siloxane polymers). We will study the role of chemical additives (surfactants) on the CNTs dispersion in the liquid preceramic. We will evaluate the possibility of functionalize the CNTs surface with suitable bonding agents to promote either their dispersion and the maximum amount of CNTs that can be introduced in the PDCs. For this objective we will benefit from the long experience of UOPD in the field of functionalizing fullerenes to help their insertion in sol-gel matrices. We will also study the role of the type of silicon polymers and the different silicon alkoxides on the CNTs dispersion. To characterize the CNTs dispersion in the preceramic polymers, SEM and HRTEM analysis will be performed before and after pyrolysis.
2. Study of the influence of CNTs on the pyrolysis process of the preceramic precursors. Indeed, it is reasonable to expect that the presence of CNTs will modify the pyrolysis process of the polymeric matrix. Therefore, the conversion process will be followed by means of DTA/TG, TG/MS, FT-IR, SEM, BET e Raman spectroscopy, and the microstructural evolution will be compared with that of the pure preceramic precursor (without CNTs). The objective of this specific study will be to optimize the pyrolyis parameters (heating rate, maximum temperature and holding time, atmosphere) in order to obtain a homogeneous and dense CNTs/SiCO nanocomposite.
3. Structural and nanostructural characterization of SiCO/CNTs nanocomposites. The specific objective is to understand how the CNTs are dispersed in the ceramic matrix and to analyze the CNT/SiCO interface. Moreover, we will look for the presence, in the ceramic matrix, of nanocrystalline phases and residual porosity. In this study we will use the following techniques: HRTEM, SEM, XRD e BET. For the characterization of the carbon phases (CNTs and free carbon forme d during pyrolysis) the UOTS will use mainly the Raman spectroscopy. UOTS will also study, with photoelectron microscopy, the chemical bonds present in the system, (Si-C, Si-O and C-C). This technique will be applied for the first time for the characterization of PCDs.
4. The synthesis conditions will be optimized for the development of specific components: fibers, films, foams and bulk samples. In order to fabricate fibers or coatings, UOTN will use his expertise in this field, especially to control the viscosity of the liquid precursors (sol-gel or polymer) containing the CNTs. For the production of foams, the group of Padova, which is leader in this field, will design a specific process, for example, by adding to the precursor solution, an expanding agent (e.g. Freon). Finally, bulk samples will be processed either by casting sol-gel solutions (UOTN) or by pressing partially crosslinked silicon resins (UOBO). To achieve the production of bulk samples, a key point will be the optimization of the pyrolysis conditions to avoid crack formation.
5. The production of bulk samples will then allow performing a physical/mechanical characterization on these novel nanocomposites. In particular, fibers and bulk samples will be used to obtain the mechanical properties of these novel materials (tensile strength, elastic modulus, fracture toughness etc.) The interaction crack/CNT will be evaluated via SEM/HRTEM observations. The ultimate goal of the specific study will be to correlate the mechanical properties with the structure and nanostructure of the PDC/CNTs nanocomposites. Bulk samples will also be investigated to identify the possible presence of a residual stress field and its formation during pyrolysis. The mechanical strength of foam samples will be measured and compared with that of unloaded samples. Finally, the electrical and thermal conductivity will be studied.
6. Study of the high temperature stability both in inert and oxidizing atmosphere. This is an important property for these novel composites because PDCs are candidates for high temperature applications. The specific goal is to understand the kinetics of the oxidation of the SiCO matrix and if and how the CNTs will modify it. This study will be carried out either on powders and on bulk samples by measuring the weight change or the thickness of the silica layer formed upon oxidation. Finally, to study the stability in inert atmosphere we will follow the changes of the structure and of the physical/mechanical properties after annealing at high temperature in the range 1200-1600°C.
Finally we mention tha fact that a secondary aim of this project, possibly even more important for the national community active in the field of nanocomposites, will be the benchmarking of the integrated AFM/SNOM/Raman system performed by UOTS. This benchmarking necessarily requires materials with nanometric discontinuities like CNT's and the nanocomposites object of the present project are ideally suited, since CNT's have very intense Raman spectra, whereas the surrounding amorphous matrix presents much weaker, sometimes vanishing Raman signals. <<<

Timescale

24 months

National and international background

Polymer Derived Ceramics (PDCs) are a new a new class of nano-materials which display very interesting properties [1]. Actually, even if they are mainly amorphous they display the same creep [2] and oxidation resistance [3] as crystalline SiC and thermal stability up to 2000°C [4]. From a chemical point of view, PDCs belong to two families: silicon carbonitrides (SiCN) and silicon oxycarbides (SiCO). From a structural point of view, both materials are formed by an amorphous network built up by covalent Si-C, Si-O and Si-N bonds. Usually, an excess of C (compared to the stoichiometric amount necessary to saturate all the valences of Si atoms), is present and forms a graphite-like phase. In the scientific literature references can be found which attribute to the presence of this C phase and its distribution in the amorphous matrix many of the exceptional properties of these materials, such as the high chemical durability in strongly aggressive media or the resistance to crystallization [5]. These materials are processed through an innovative method based on the polymer technology. The starting material is a metallorganic polymer that is first shaped into a plastic form, then crosslinked into a thermosetting polymer and finally it is converted into the ceramic with a pyrolysis process in controlled atmosphere. The first commercial products made with this technology are the SiC-based Nicalon [6] and the SiTiCO-based Tyranno fibers [7]. Then, many other ceramic systems have been successfully produced and we like to remind here the boron-containing silicon carbonitrides for their exceptionally high thermal stability up to 2000°C [8]. This process has many advantages compared to the traditional powder processing route, like: (i) is complete below 1000 °C, yet the products can be used at much higher temperature (1600 °C for SiCN); (ii) through an accurate control of the chemical composition of the preceramic precursors it is easy to introduce in PDCs either Si, C and SiC nanoclusters [9] or a second phase (metals, oxides, carbides, nitrides, …) [10]. Accordingly, ceramic materials stable to high temperature and having specific functionalities (optical, electrical, magnetic, mechanical) can be easily produced [11]; (iii) difficult to shape components such as: fibers [6-7, 12], films [13], foams and microporous components [14] and matrices for composites [15] can be easily processed. As we already mentioned, silicon oxycarbides (SiCO) are an important family of PDCs. Even if they are not as stable as the carbonitrides, yet are far more resistant to crystallization than vitreous silica, their viscosity is up to two order of magnitude higher [16] and display superior chemical durability [17]. Moreover, SiCO are a simpler system compared to carbonitrides and boroncarbonitrides and can be used as a model easier to study [18, 19].
Silicon oxycarbides can be prepared starting from cheap commercial silicon resins, following the process already described [20, 21], or from sol-gel precursors [22, 23]. According to the sol-gel process, the starting molecular precursors are silicon alkoxides having the general formula: R-Si(OEt)3, where R is H, CH3, CH2=CH, C6H5 etc. Hydrolysis condensation reactions of these alkoxides lead to a gel, which is the preceramic precursor. When the sol-gel method is used, the shaping of the components is performed in the sol stage, before gelling, according to different technologies: drawing for fibers [24], dip or spin coating for films [9], and casting for bulk components [25]. Even if the two precursors (silicon resins and sol-gel materials) lead to similar SiCO ceramics they display complementary features: silicon polymers are less expensive, normally display a higher ceramic yield and allow to use all the shaping polymer technologies (extrusion, injection molding, foaming…) [10]. Sol-gel method is far more suited when (i) a detailed control of the composition and architecture of the preceramic precursor is desired by selecting the proper mixture of silicon alkoxides [26]; (ii) to form ceramic matrices via infiltration [15]; (iii) if multicomponent systems, containing Si, B, Al, Ti, Zr, etc. are preferred [27, 28].
There are still many fundamental questions waiting for answers before all the capabilities of PDCs will be exploited. Accordingly, the nanostructural characterization of PDCs is not a trivial problem and requires a multitechnique approach. For example, the study of the carbon-based phase, which is usually present in PDCs, seems to be a fundamental step for the understanding the exceptional properties of these materials. The most appropriate techniques for the characterization of C in different forms is Raman spectroscopy, which is more and more applied in this field [29]. Transformation from the polymeric to the ceramic state is accompanied by hydrocarbon evolution and volume shrinkage. Even if the shrinkage is comparable with the one observed in the traditional sintering, gas evolution can cause cracks formation and the development of a residual stress field. This occurs when the distance for diffusion of the gaseous species is to high and limits the practical application of the polymer pyrolysis method to components that have a maximum thickness of 1-2 mm in the shortest dimension. Accordingly, fibers, films, porous or microcomponents can be easily prepared and bulk components are not, at least at the moment. Characterization of residual stress field in PDCs is certainly a required step to overcome this important technological limitation and allow increasing the dimensions of the components that can be processed. Even in this case, Raman spectroscopy, through the piezo-spectroscopic effect, could be profitably applied by correlating the shift of Raman carbon bands with the local stress.
As far as the mechanical properties are concerned, PDCs show excellent values of strength, elastic modulus and high temperature creep resistance but the fracture toughness is still low (2-3 MPa m1/2) [30].
CNTs [31] introduction in the PDC matrix could broaden even more the excellent properties of these new materials, in particular the mechanical, electrical and thermal properties. Indeed, CNTs may be the strongest materials known: single walled nano tubes, SWNTs, display strength up to 50 GPa and elastic modulus up to 1000 GPa [32], and they have exceptionally high values of electrical and thermal conductivity [33]. Carbon nanotubes can be produced as single walled or multi walled structures using techniques such as CVD, arch discharge and are commercially available in a wide range of purity, price and amounts. They can be considered as hybrids between molecules and materials. Due to their shape factor (>10000) and periodic structures, they can be modeled essentially as monodimensional crystals.
In the scientific literature papers dealing with ceramics reinforced with CNTs are limited, mainly due to the difficulties in processing these composites with the traditional powder sintering technique [34]. For this reason recent papers report the processing of CNTs-ceramic matrix composites using novel processing routes [35-37].
The possibility of dispersing the CNTs in the liquid preceramic precursor and then obtaining the nanocomposite by pyrolysis at moderate temperatures seem to be an extremely appealing method to overcome the problems related with powders sintering. In spite of these attractive features we did not find in the literature any papers reporting the processing of ceramic nanocomposites from preceramic polymers and CNTs and this field seem still unexplored and ideally suited for an exciting research work. Indeed it is appropriate to briefly introduce here the new integrated AFM/SNOM/Raman set up which is being acquired by UOTS and which is a fundamental breakthrough in the characterization of nanocomposites with CNT's. While a more lengthy discussion is given in the specific project of UOTS, here we will present only the possibility that is affords within the frame of the present proposal. AFM is a well established technique for the characterization of the topography of surfaces. In fig. 1 it is possible to see an AFM image of a single CNT recorded with integrated AFM/SNOM/Raman set up.



Fig. 1: AFM image of a single CNT. The circle in darker color represents the dimension of the probe used to obtain a Raman spectrum, which is about 200 nm.

AS it can be seen from the image it is possible to acquire Raman spectra with an unprecedented lateral resolution (about 200 nm). Till the advent of this integrated system, the best possible resolution with Raman spectroscopy was about 1-2 micron with a conventional micro-focused Raman spectroscope. The improvement by one order of magnitude makes now possible to study one single nanotube with respect to i) the orientation (chirality) of CNT and ii) the stress sustained by the CNT. This is due to the fact that the Raman spectrum of CNT's is extremely intense, is sensitive to chirality and is sensitive to stress (under the same condition -laser power and integration time, the Raman spectrum of CNT is one order of magnitude more intense than tetragonal zirconia, the ceramic material with the most intenseknown Raman spectrum). Indeed, the strength of the Raman signal of CNT's makes the nanocomposites object of the present study an ideal benchmark to assess the ultimate potential of this integrated system. The integrated AFM/SNOM/Raman, presently the only such instrument available in Italy, has been examined very recently on "Nature" [38]. <<<