SiC-Based Composites Sintered with High Pressure Method

Silicon carbide-based materials usually have high hardness (2500 – 2800 HV) and thus have superior wear resistance. Nevertheless, the tribological performance of SiC is determined by many factors, such as the grain size of mated materials or the reactions in the presence of oxygen and humidity in the surrounding atmosphere. For example, in unlubricated sliding, wear resistance of SiC ceramics can be greater in air than in inert atmosphere owing to thin soft oxide films reducing friction and local surface pressure. (Gahr et al., 2001; Guicciardi et al., 2007). The friction and wear properties of SiC materials (both in dry and lubricating conditions) have been studied extensively because they are used in applications like bearings, cylinder liners and mechanical seals (Murthy et al., 2004). Silicon carbide-based ceramics have high melting point (~2500 °C), high thermal conductivity (43 – 145 W/m·K – depending on a temperature and phase composition), low thermal expansion (~4,5×10-6·K-1), and high temperature capability. Silicon carbide is a semiconductor which can be doped n-type by nitrogen or phosphorus and p-type by aluminium, boron, gallium or beryllium. Due to the combination of its thermal and electrical properties, SiC is applied in a resistance heating, flame igniters and electronic components. Relatively pure SiC has also an excellent corrosion resistance in the presence of hot acids and bases (Richerson, 2004). Silicon carbide powder compacts are difficult to densify without additives because of the covalent nature of the Si–C bonds and the associated low self-diffusion coefficient. Therefore, Reaction Sintering (RS) in the presence of liquid silicon as well as Hot Isotactic Pressing (HIP) are frequently used to obtain a high quality, full dense SiC ceramics. Typical room temperature flexural strength of SiC-based materials is about 350-550 MPa. High-strength RS-SiC (over 1000 MPa in a 3-point bending test) was developed by controlling the residual Si size under 100 nm. (Magnani et al., 2000; Suyama et al., 2003). Silicon carbide ceramics have the ability to increase in strength with increase of temperature. It was reported that flexural strength of some kind of commercial SiC ceramic increase is from 413 MPa at the room temperature to around 580 MPa at 1800 °C (Richerson, 2004). For hot-pressed silicon carbide with addition of 0.15-1.0 wt% Al2O3, the high-temperature strength has been improved from 200 MPa to 700 MPa by decreasing the grain boundary concentration of both Al and O at 1500 °C (Kinoshita et al., 1997).

A favorable combination of properties makes SiC materials suitable for many engineering applications, including parts of machines and devices exposed to the abrasion, the high temperature, the corrosive environment, etc. A major disadvantage of SiC ceramic materials is their low fracture toughness, which usually does not exceed about 3.5 MPa·m 1/2 (Lee et al., 2007;Suyama et al., 2003). Low values of K Ic coefficient exclude these materials from numerous applications with dynamic loads, e.g. in machining processes. There are various ways to improve the fracture toughness of ceramic materials. One of them involves obtaining a composite material by the introduction of the additional phases in the form of nano-, micro-or sub-micro-sized particles to the base material. Some papers indicate that nanosized structures have great potential to essentially improve the mechanical performance of ceramic materials even at high temperatures (Awaji et al., 2002;Derby, 1998;Kim et al., 2006;Niihara et al., 1999). Depending on the type of introduced particles, composites can take advantage of different strengthening mechanisms, such as the crack deflection, crack bridging, crack branching, crack bowing, crack pinning, microcracking, thermal residual stress toughening, transformation toughening and synergism toughening. For example, metallic particles are capable of plastic deformation, thus absorption of energy and bridging of a growing crack, resulting in increased strengthening ( Fig. 1a) (Yeomans, 2008). On the other hand, hard ceramic particles, like borides or nitrides, can introduce a favorable stress state which can cause a toughening effect by crack deflection and crack bifurcation ( Fig. 1b) (Xu, 2005). An addition of metal borides such as ZrB 2 , TaB 2 , NbB 2 or TiB 2 , promote densification of SiC powder as well as improve hardness and other mechanical properties of the material as a whole (Tanaka et al., 2003). Fig. 1. Example of strengthening mechanisms which can occur in ceramic matrix composites with dispersed "soft" metallic or/and "hard" ceramic particles: a) crack bridging, b) crack deflection and crack bifurcation The wide group of materials containing the silicon carbide are SiC/Si 3 N 4 composites. In such materials predominant phase is silicon nitride, while SiC content does not usually exceed 30 vol.%. Silicon nitride has a lower hardness but a higher fracture toughness than silicon carbide. If SiC particles are uniformly dispersed in the Si 3 N 4 ceramics, high strength can be obtained from room temperature to elevated temperature. It was reported that the strength of 1000 MPa at 1400°C is obtained in nano-composites having ultra-fine SiC particles added into the Si 3 N 4 matrix. This improvement was mainly attributed to the suppression of a grain boundary sliding by intergranular SiC particles bonded directly with the Si 3 N 4 grain in the atomic scale without any impurity phases (Hirano & Niihara, 1995;Yamada & Kamiya, 1999). SiC/Si 3 N 4 composites have an ability to crack healing under high temperature and applied stress, to exhibit a significantly higher creep resistance and fracture toughness compared to the monolithic materials. (Ando et al., 2002;Lojanová et al., 2010;Sajgalík et al., 2000;Takahashi et al., 2010). The combination of the fair fracture toughness with high hardness, wear resistance and mechanical strength at elevated temperatures makes SiC/Si 3 N 4 ceramics a promising material for cutting tools (Eblagon et al., 2007). Despite many studies on materials based on silicon carbide and silicon nitride, there is a lack of knowledge about the SiC/Si 3 N 4 composites where the predominant phase is SiC. In the presented work, the materials contained from 0 to 100% of silicon carbide were investigated.

Description of experiment
The purpose of the presented experiment was to study the influence of High Pressure -High Temperature (HPHT) sintering on the phase composition, microstructure and selected properties of SiC/Si 3 N 4 composites as well as to study the effect of the addition of thirdphase particles selected from metals (Ti) or ceramics (TiB 2 , cBN -cubic Boron Nitride) to the SiC -Si 3 N 4 system. The main goal was to improve fracture toughness and wear resistance of the investigated materials.
The composites were manufactured and tested in two stages. The first stage consisted in sintering of materials having, in its initial composition, only SiC and/or Si 3 N 4 powder(s). Samples sintered from nano-, sub-micro-and micropowders with various silicon carbide to silicon nitride ratios were investigated at this stage. At the second stage the best SiC/Si 3 N 4 composite manufactured at the first stage was subjected to modification, consisting of: -use of various types of SiC and Si 3 N 4 powders, -addition of metallic phase in the form of Ti particles, -addition of boride (TiB 2 ) phase, -addition of superhard (cBN) phase. All materials were sintered with the HPHT method. The parameters of sintering: time and temperature were chosen individually for each composition. The obtained samples were subjected to a series of studies, which included: phase composition and crystallite size analysis by X-ray diffraction, measurements of density by hydrostatic method and Young's modulus by the ultrasonic method, measurement of hardness and and fracture toughness using Vickers indentation as well as studies of tribological properties using the Ball-On-Disk method.

HPHT method of sintering
Pressure is a versatile tool in solid state physics, materials engineering and geological sciences. Under the influence of high pressure and temperature there are a lot of changes in physical, chemical and structural properties of materials (Eremets, 1996). It gives a possibility to generate of new, non-existent in nature phases, or phases which occur only in inaccessible places, such as the earth core (Manghnani et al., 1980). The use of pressure as a parameter in the study of materials was pioneered principally by Professor P. W. Bridgman, who for forty years investigated most of the elements and many other materials using diverse techniques (Bridgman, 1964). There are many design solutions to ensure High Pressure -High Temperature (HPHT) conditions for obtaining and examination of materials. Depending on the design assumptions, it is possible to achieve very high pressures, up to several hundred gigapascals, as in the case of Diamond Anvils Cell (DAC). Such devices, due to their small size, are intended solely for laboratory investigations (XRD in-situ study, neutron diffraction etc.) (Piermarini, 2008). For the purposes of industrial and semi-industrial production of materials the most frequently the "Belt" or "Bridgman" type of equipment is used (Eremets, 1996;Hall, 1960;Khvostantsev et al., 2004). These apparatuses provide a relatively large working volume, the optimum pressure distribution and the possibility of achieving high temperatures. In the toroidal type of Bridgman apparatus the quasi-hydrostatic compression of the material is achieved as a result of plastic deformation of the so called "gasket" (Fig. 2). Fig. 2. Sintering process in a Bridgman-type HPHT system. Quasi-hydrostatic compression of the preliminary consolidated powders (sample -1) is achieved as a result of plastic deformation of the gasket material (2) between anvils (3); electrical heating is provided by a high-power transformer (4) and graphite resistive heater (5) Gaskets are made of special kinds of metamorphic rocks such as pyrophyllite, "lithographic stone" or catlinite (Filonenko & Zibrov, 2001;Prikhna, 2008). The toroidal chamber, depending on its volume (usually from od 0.3 do 1 cm 3 ), can generate pressures up to 12 GPa and temperature up to ~2500 °C. The presented system is used often for production of synthetic diamonds and for sintering of wide range of superhard composites based on polycrystalline diamond (PCD) or polycrystalline cubic boron nitride (PcBN). Under the influence of a simultaneous action of pressure and temperature the sintering process occurs much faster than in the case of free sintering. A typical duration of sintering process with HPHT method is about 0.5 -2 minutes (Fig. 3) while the free sintering requires several hours. Short duration of the process contributes to the grain growth limitation, which is essential in the case of sintering of nanopowdes. The materials obtained with HPHT method are characterized by almost a 100% level of densification, isotropy of properties and sometimes by a completely different phase composition in relation to the same free-sintered materials, due to the different thermodynamic conditions of the manufacturing process.  After drying, the mixtures were preliminarily compressed into a disc of r 15 mm diameter and 5 mm height under pressure of ~200 MPa. The green bodies with the addition of TiH 2 were additionally annealed in a vacuum at a temperature of 800 °C for 1h in order to remove the hydrogen and obtain pure metallic titanium. The materials were obtained at high pressure (6 GPa) in the wide range of temperatures ( 430 -2150 °C depending of composition) using a Bridgman-type toroidal apparatus (Fig. 2). The sintering temperatures were established experimentally for each material to obtain crack-free samples with the highest values of density and mechanical properties. The duration of the sintering process was 40s for nanopowders and 60 s for the others. The sintered compacts were subsequently ground to remove remains of graphite after the technological process of sintering and to obtain the required quality and surface parallelism for Young's modulus determination. The samples provided for microscopic investigations and for mechanical tests were additionally polished using diamond slurries.

Investigation methods
Densities of the sintered samples were measured by the hydrostatic method. The uncertainty of the measurements was below 0.02 g/cm 3 , which gave a relative error value of below 0.5 % (excluding measurements of small pieces of broken samples, where the error was up to 0.1 g/cm 3 , due to their insufficient volume and mass). Young's modulus of the samples obtained by HPHT sintering was measured by means of transmission velocity of ultrasonic waves through the sample, using a Panametrics Epoch III ultrasonic flaw detector. Calculations were carried out according to (Eq. 1): where: E -Young's modulus, C L -velocity of the longitudinal wave, C T -velocity of the transversal wave, ρ -density of the material. The velocities of transverse and longitudinal waves were determined as a ratio of sample thickness and relevant transition time. (2) www.intechopen.com where: K Ic -critical stress intensity factor, ϕ -constrain factor, H -Vickers hardness, E -Young's modulus, a -half of indent diagonal, c -length of crack.
Microstructural observations were carried out on the densified materials using a JEOL JXA-50A Scanning electron Microscope equipped with back scattering electron (BSE) imaging.
In the Ball-On-Disc tests, the coefficient of friction and the specific wear rate of the sintered samples in contact with Si 3 N 4 ball were determined using a CETR UMT-2MT (USA) universal mechanical tester. In the Ball-On-Disc method, sliding contact is brought about by pushing a ball specimen onto a rotating disc specimen under a constant load (Fig. 4). The loading mechanism applied a controlled load F n to the ball holder and the friction force was measured continuously during the test using an extensometer. For each test, a new ball was used or the ball was rotated so that a new surface was in contact with the disc. The ball and disc samples were washed in ethyl alcohol and dried. The size of the disc-shaped samples was ~13.5 in diameter and ~3.8 mm in height with the surfaces flatness and parallelism within 0.02 mm. The roughness of the tested surface was not more than 0.1 µm R a . The following test conditions were established: ball diameter -2 mm, applied load -4 N, sliding speed -0.1 m/s, diameter of the sliding circle -2 ÷ 5 mm, sliding distance -100 m, calculated duration of the test -1000 s. The tests were carried out without lubricant at room temperature. Each test was repeated at least three times. Coefficient of friction was calculated from (Eq. 3): where: µ -coefficient of friction, F f -measured friction force, F n -applied normal force. After completing the test, according to ISO 20808:2004 E standard, the cross-sectional profile of the wear track at four places at intervals of 90° was measured using a contact stylus profilometer PRO500 (CETR, USA). Then the average cross-sectional area of the wear track was calculated. The volume of material removed was calculated as a product of crosssectional area of the wear track and their circumference. Specific wear rate was calculated from (Eq. 4): where: W s -specific wear rate, V -volume of removed material, L -sliding distance.

Materials sintered from nano-, sub-micro-, and micro-SiC/Si 3 N 4 powders
A macroscopic view, phase composition and crystallite size of materials sintered from nanopowders under the pressure of 6 GPa at temperatures ranging from 430 to 1880 ° C in the period of time 40s are shown in Table 2 The initial crystallite size of sintered SiC nanopowder was about 4.1 nm and did not change until the sintering temperature of 1170 °C. At the sintering temperature of 1450°C crystallites reached an average size of 10.6 nm while at the maximum applied temperature (1880 °C) their size was 110 nm. This material showed no phase transformation. The sinters obtained at 1880 °C as well as the initial powder have the cubic structure of β-SiC phase. Sintered Si 3 N 4 powder remained amorphous until the temperature of 890 °C. At the temperature of 1170 °C a new phase crystallized. The X-ray diffraction pattern of this phase corresponds to the o'-sialon with orthorhombic structure and Si 1.8 Al 0.2 O 1.2 N 1.8 stoichiometry.
As evidenced by the chemical formula, this compound has a low content of aluminum with reference to silicon while the quantities of nitrogen and oxygen are comparable. It can be assumed that o'-sialon is a transition phase between the amorphous silicon nitride and a completely crystalline β-Si 3 N 4 phase. O'-sialon was formed probably due to embedding a certain amount of oxygen and impurities adsorbed on the surface of powder to an atomic lattice of crystallizing silicon nitride. The crystallite size of this phase was estimated to around be 77 nm. The samples obtained at temperatures of 1450 and 1880 °C contain only β-Si 3 N 4 phase with the crystallite size 77.4 and 143 nm respectively. 50 SiC/50 Si 3 N 4 -vol.% nanocomposite was sintered at the same temperatures as silicon carbide powder without additions. The phase composition of sintered composites did not differ qualitatively from the sum of their components sintered separately. There was no formation of new phases in the reaction between silicon carbide and silicon nitride. Sintering of composites, especially at higher temperatures, leads to lower grain growth than it is in single-phase powders sintered separately. This indicates a favorable effect of inhibiting grain growth in the composite.
The comparison of physical-mechanical properties of nano-, sub-micro-and micro-SiC/Si 3 N 4 materials sintered at different temperatures is presented in Fig. 5 and Table 3. Generally, nanocomposites are characterized by the lowest physical-mechanical properties of the three granulometric types of the investigated materials. Densities and Young's modulus values of the best nanostructured samples do not exceed 2.55 g/cm 3 and 135 GPa respectively. In most cases, nanostructured SiC/Si 3 N 4 samples are characterized by a lot of cracks (see Table 2).
Cracking of such ceramics occurs as a result of the presence in their structure of residual micro-and macro-stresses which overcome the strength of the produced material. The fine powder is characterized by a very large specific surface and high gas content in the sample due to the absorption process of the material particles. During heating, as a result of the increase in temperature, the volume of gases increases, which causes cracking or even permanent fragmentation of the sample. In order to prevent the cracking phenomena in the samples, various conditions of the sintering process were tested. Depending on composition, materials characterized by the highest level of densification and the best mechanical properties were obtained at different temperatures: 890 °C for pure Si 3 N 4 and 1880 °C for 50 Si 3 N 4 /50 SiC -vol.% composite. Unfortunately, some of these samples had cracks as well. Different kinds of internal cracks, delamination and other defects of microstructure occurred in most of the nanostructured samples. These defects cause a scattering of caustic waves propagated through the material and, in consequence, the impossibility of Young's modulus measurements using ultrasonic probes (marked as **nm in Table 3). Composites without cracks were obtained only from micro-and sub-micro-structured powders. Only for these composites hardness and fracture toughness were measured.  Table 3. Physical-mechanical properties of selected nano-, sub-micro-and micro-SiC/Si 3 N 4 materials sintered at optimal temperatures; *optimum temperature for selected properties, e.g. 1690 (K Ic ) -the best fracture toughness; **nm -non measurable with ultrasonic method www.intechopen.com The composites obtained from submicron powders are characterized by the best properties. Density and Young's modulus of the best 70 SiC/30 Si 3 N 4 -vol.% compacts sintered at 1880 °C were 3.18 g/cm 3 and 377 GPa respectively (over 99% and 90% of the theoretical values). This material is also characterized by the highest hardness (HV1 ~3000) and relatively good fracture toughness (4.9 MPa·m 1/2 ). The same material sintered at a lower temperature (1690 °C) has slightly lower values of density (3.14 g/cm 3 ), Young's modulus (363 GPa) and hardness (HV1 2626) but higher fracture toughness (5.6 MPa·m 1/2 ) - Table 3. HPHT sintered sub-micro-70 SiC/30 Si 3 N 4 -vol.% composites have a better combination of mechanical properties than comparable commercial materials (Table 4). The composites obtained from the micro-sized powders have their properties intermediate between the nano-and sub-micro-structured materials. Even though the best micro-SiC/Si 3 N 4 samples are crack free and have fairly good density (>94% of theoretical values) and indentation fracture toughness (4.6 -4.8 MPa·m 1/2 ), their Young's modulus and hardness are much lower than for the sub-micro-structured samples. The insufficient mechanical properties of microstructured Si 3 N 4 -SiC materials can be attributed not only to grain size but also to specific properties of the initial powder resulting from their production method (e.g. shape of the grains, impurities, oxidation etc.). The choice of suitable initial powders for the given method of sintering is the critical factor in achieving good quality of ceramics.

Si 3 N 4 based ceramics SiC based ceramics
For all the investigated samples, independently of their grain size, a strong influence of indentation load on hardness values can be observed. Increasing the indentation load causes a decrease in hardness values (Table 3). Among the composites sintered without additional phases, the highest degree of densification and best mechanical properties were demonstrated by composite obtained from submicron powders 70 SiC(sub-micro)/30 Si 3 N 4 (sub-micro, Starck) -vol%. (Fig. 6 and Table 5). This composite was selected for modification by the addition of the third phase particles. The modification of the 70 SiC/30 Si 3 N 4 composite by the addition of Ti was not successful. The samples with the addition of 8 vol.% Ti introduced in the form of TiH 2 , sintered at low temperatures, were characterized by a very low Young's modulus, whilst all the samples sintered at temperatures above ~1200 °C were cracked. A decrease in density was observed with increasing sintering temperature, whilst Young's modulus showed an upward trend (Fig. 7). The composites with the addition of TiB 2 were characterized by a high degree of densification, a high Young's modulus and improved K Ic as compared to the unmodified composite. No improvement in hardness was observed (Table 5). In the case of 70 SiC/30 Si 3 N 4 material with the addition of 8 vol.% TiB 2 , there is some increase in density and Young's modulus with increasing temperature (Fig. 8). Composite with the addition of 30 vol.% TiB 2 , shows an increase in density with sintering temperature up to a maximum value, and then its stabilization. A further increase of the sintering temperature results in cracking of the samples (Fig. 9). The composites modified by the addition of 8 vol.% cBN micropowder have better properties than the composites with TiB 2 but a tendency to cracking of this material is noticeable. The use of nano-cBN particles as a modifier causes the deterioration of the properties and cracking of samples (Fig. 10). The composites modified by the addition of 30 vol.% cBN micropowder, showed the best mechanical properties ( Fig. 11 and Table 5).  high content of tungsten carbide and zirconium dioxide from the vessel and grinding media used to prepare the mixtures (white areas visible in the microstructures - Fig. 12). The highest quantity of WC was admixed to composite containing 30% of cBN super-abrasive powder. WC has density of 15.7 g/cm 3 . It explains too high value (118%) of relative density of 70 SiC/30 Si 3 N 4 + 30vol.% cBN composite. An example of a crack which developed in a Vickers indentation test in the composite modified by addition 30% of cBN is presented in Fig. 13. In this material the mixed mode of crack propagation can be observed. Some parts of fracture are of an intra-crystalline character (indicated as 1 in Fig. 13) while some are inter-crystalline (indicated as 2 in Fig. 13). In both cases the change of direction of crack propagation is visible. A similar effect can be observed in the composites modified by the TiB 2 phase. This indicates that the crack deflection mechanism influences on toughening of composites modified by ceramic particles.  Additionally, for comparison, the commercial Si 3 N 4 based cutting tool material (ISCAR, IS9-grade) was also tested. The mean curves of friction coefficient for investigated materials in a sliding contact with the Si 3 N 4 ball are presented in Fig. 14. A comparison of the specific wear rate determined by the wear tracks measurement is presented in Fig. 15 (micro) and commercial material are classified. Their specific wear rates are 8.7 × 10 -6 and 7.2 × 10 -6 mm 3 /N·m respectively. Significantly better are composites modified by addition of TiB 2 particels. The specific wear rate of 70 SiC(sub-micro)/30 Si 3 N 4 (submicro, Starck) + 8 vol.% TiB 2 (micro) and 70 SiC(sub-micro)/30 Si 3 N 4 (sub-micro, Starck)+ 30 vol.% TiB 2 (micro) samples equal 3.7 × 10 -6 and 2.3 × 10 -6 mm 3 /N·m respectively. The highest wear resistant is exhibited by 70 SiC(sub-micro)/30 Si 3 N 4 (sub-micro, Starck) + 30 vol.% cBN(micro) composite with its specific wear rate value equals only 1.2×10 -6 mm 3 /N·m.

Conclusions
The performed research proves that the HPHT sintering is a method of the future for compacting SiC/Si 3 N 4 nanopowders, due to the short time of the process amounting to 40 seconds that permits the grain growth limitations. The obtained compacts were characterized by the crystallites sizes of 4 to 143 nm, depending on the sintering parameters. SiC and SiC/Si 3 N 4 samples sintered from nanopowders are characterized by the presence of cracks. Cracking of such ceramics occurs as a result of residual micro-and macro-stresses in their structure which overcome the strength of the produced material. The fine powder is characterized by a very large specific surface and high gas content in the sample due to the absorption process of the material particles. During heating, as a result of the increase in temperature, the volume of gases increases, which causes cracking or even permanent fragmentation of the sample. The research regarding the mechanical properties of SiC/Si 3 N 4 composites indicates that materials obtained from submicron powders display the best properties. Density and Young's modulus of the best 70 SiC/30 Si 3 N 4 -vol.% compacts, sintered at 1880 °C, were 3.18 g/cm 3 (over 99% the theoretical values) and 377 GPa respectively. This material is also characterized by the highest hardness (HV1 ~3000) and relatively good fracture toughness (4.9 MPa·m 1/2 ). The same material sintered at a lower temperature (1690 °C) has slightly lower values of density (3.14 g/cm 3 ), Young's modulus (363 GPa) and hardness (HV1 2626) but higher fracture toughness (5.6 MPa·m 1/2 ). HPHT sintered sub-micro-70 SiC/30 Si 3 N 4 -vol.% composites have a better combination of mechanical properties than comparable commercial materials. The research concerning mmodification of sub-micro-70 SiC/30 Si 3 N 4 -vol.% composites proves that the addition of the third phase in the form of TiB 2 or cBN particles contribute to their further improvement. Composites modified by the addition of 30vol.% cBN micropowder are characterized by the best combination of Young's modulus, hardness, fracture toughness, coefficient of friction and the specific wear rate. Such properties predispose 70 SiC/30 Si 3 N 4 + 30vol.% cBN composites to various advanced engineering applications including their use for wear parts and cutting tools.

Acknowledgments
This study was carried out within the framework of the project funded by the Polish Ministry of Science and Higher Education (Project number: DPN/N111/BIALORUS/2009).
The author would like to thank Prof. M. Bućko from AGH University of Science and Technology in Krakow for XRD analysis. The author would also like to thank his colleagues from The Institute of Advanced Manufacturing Technology in Krakow for Vickers's indentation tests and for SEM studies. Finally, the author would like to thank his supervisor Prof. L. Jaworska for her optimism and valuable advice throughout this research.