Direct current (dc) plasma-assisted sintering of metal parts is a promising and relatively new research and development field in powder metallurgy (PM). In the present entry, it is intended to introduce the reader to the main applications of the dc plasma sintering process in PM. To achieve this goal, the present entry is divided in a brief introduction and sections in which the bases of the dc plasma abnormal glow discharge regime and its influence in the sintering process are carefully treated. In this case, a clear language is purposely used to didactically introduce the reader to this “fascinating glow world”, the dc plasma-assisted sintering of metal parts, aiming to put in evidence the main points on physicochemical aspects of the plasma environment, basic knowledge of the plasma heating, and surface-related phenomena during dc plasma sintering of parts. All these aspects are treated considering the main techniques of the dc plasma-assisted sintering process applied to PM. Finally, some results on DC plasma heating, sintering and surface modification are presented.
Want more information on Direct Current Sintering? Feel free to contact us.
New technologies tend to require more and more advanced materials and manufacturing processes. Additionally, due to environmental problems that the world is facing, especially in developed countries, the regulation concerning the environmental impact of these materials and processes are becoming increasingly restrictive. In this context, plasma-assisted processes are very competitive since they usually are low-environmental-impact processes and they can produce high-performance materials in a very efficient way. Such aspects, allied to the possibility of producing very reductive atmospheres, motivated the development of the plasma-assisted sintering process, being that in most of research carried out in this domain, plasma is generated using direct current (dc) power supply.
In the dc plasma-assisted sintering process performed in an abnormal regime glow discharge, pressed metal parts to be sintered are subjected to a highly reactive plasma environment [1–4]. It is due to the presence of a great number of ionized and excited gas species, created by collision processes, that leads the species present in a gas mixture, usually constituted of reductive (like H2) and neutral (like Ar) gases, to be ionized, excited, and/or dissociated. Due to the interaction of plasma and the pressed part, temperatures sufficiently high to sinter metals can be reached by the energy transfer from the plasma species to the part surface by collisions of accelerated species (this heating mode is termed “cathode heating mode”). It is also possible to heat pressed parts by an indirect way, in which the plasma energy is transferred to heating components, acting as cathode of the discharge, that heat the parts to be sintered mainly by radiation (here termed “anode and/or floating potential heating mode”). Depending on the design of the electrical discharge (electrode arrangement), different electrical configurations and heating modes are possible. In each case different potentialities of the plasma-assisted process can be explored, giving rise to different dc plasma sintering techniques [5, 6]. In addition, the very active physicochemical plasma environment, besides heating, can lead the metallic parts to present surface characteristics that are directly related to the plasma species bombardment and, so, exclusive for the plasma sintering process. The plasma environment can promote oxide-reduction reaction, cleaning, sputtering, deposition, and redeposition phenomena, all of them can be conveniently explored to tailor the sintered part surface. The abovementioned aspects suggest that during sintering, several phenomena related to the plasma will have influence in the part sintering processes, so to understand the dc plasma-assisted sintering process as a whole it is necessary to introduce some basis of plasma. In the next sections bases of dc plasma physics for the sintering purpose are presented, considering the different dc plasma sintering techniques, followed by the presentation of some results on plasma heating and surface modification. So the chapter ends by a discussion on the advantages of the plasma-assisted over the conventional sintering process and the final remarks.
AdvertisementPlasmas commonly used in dc plasma sintering process characteristically present a degree of ionization of about 10−5 [1]. For the sintering purpose, an abnormal glow discharge can be easily obtained by applying a potential difference between two electrodes (cathode and anode) placed inside a chamber at low pressure, containing the gas mixture, which is usually constituted of 80 vol.% Ar + 20 vol.% H2 for sintering. In the initial state, electrons are accelerated by the electric field and collide with neutral species (atoms/molecules), in much higher number density, resulting in partial ionization/excitation of the gas. When the number density of the produced ionized and excited species is great enough, a self-sustained bright-aspect glow electrical discharge is attained; thus, the plasma is formed [1–4].
Among the different discharge regimes which can be established from the current-voltage characteristic of a dc plasma system, namely Corona, Townsend, Subnormal, Normal, Abnormal, Transition to arc, and Arc [1, 4, 7], the abnormal regime is of special interest for dc plasma sintering process. This is mainly due to the following aspects:
It is the discharge regime in which the cathode can be totally covered by the glow, condition necessary to perform homogeneous heating; and
It presents the basic feature for which the electrical current monotonically increases with the applied voltage. This allows the application of high voltages resulting in increased ionization/excitation of the gas and makes possible to control the plasma reactivity and the cathode (or sintering) temperature.
The abnormal regime is sustained for current densities typically higher than 2 mA/cm2 [1] and for pressures generally in the range of 10−2 to 102 Torr [8–10].
Figure 1 schematically presents the normal, abnormal, transition to arc, and arc regime regions in a hypothetical current-voltage characteristic of a dc discharge [1, 4, 7], being the abnormal and arc regimes typically applied for materials processing [1, 4]. It is worth to be mentioned that for certain operation conditions a specific critical voltage (Vc) for the considered dc plasma system can be attained, at least locally. In this case, the working abnormal regime discharge can abruptly change to arc (passing through the transition from abnormal regime to the arc regime), and the sintering process is interrupted by the power supply security system (arc management system). There are two main aspects that can contribute to the critical voltage of a dc plasma system to be achieved. The first one is the presence of organic constituent in the electrode surface, directly related to the dc plasma device cleanness, which is detrimental for dc plasma processes and also increases the risk of arc formation. The second one is the geometrical and dimensional arrangement related to the design of the electrodes and their respective insulating components.
Arc formation represents the main limitation for the use of dc plasmas in industrial applications. If the arc regime is established, high current density and high temperature (usually over 3000°C) are observed, and it can lead to surface damages in the sintering part or in the power supply system. It is to be noted that the arc formation risk in plasma processing was partially overcome in the half of the last century with the introduction of the pulsed dc plasma power supplies, and nowadays industrial plasma power supplies are additionally equipped with advanced arc managing systems to minimize such problem. In this case, for each switched-off time of the pulse (see Figure 2a, b), the system current is decreased, tending to zero, so the risk for arc formation is reduced. Despite these advances, it remains an important challenge regarding the know-how and R&D of dc plasma sintering devices.
By considering that the arc formation risk increases as the plasma power is increased, another strategy used in the industry to overcome this important problem is the use of hot-wall plasma chamber. In this case the power needed to reach the processing temperature is mainly provided by an auxiliary heating system, reducing the necessary plasma power to achieve the desired sintering temperature, decreasing the risk of arc formation.
To go further on the understanding of the plasma-assisted sintering, it is necessary to stress in more details how the potential distribution along the discharge is changed after breakdown (when the glow discharge or the plasma itself is obtained).
As the plasma is obtained by the gas ionization, mainly due to the collisions between high-energy electrons and neutral gas species, the potential distribution (indicated by red lines in Figure 3a, b) between the two electrodes is changed from that represented in Figure 3(a) to that in Figure 3(b). In Figure 3(a), the scheme of a gas at low pressure and the respective electrodes of a typical electrical system used in dc glow discharge chambers are shown. It is to be noted that ions and electrons of the gas, initially present at very low number density (around 103 cm−3), are accelerated by the electrical field imposed by the power supply. For a specific potential and sufficient low pressure, the dielectric gas breakdown is achieved, phenomena governed by Paschen’s law [7], and a glow discharge (thus the plasma) is established, changing the potential distribution along the discharge to that presented in Figure 3(b) due to the generation of a significant number of charged species. In abnormal glow regime, the glow region is approximately equipotential (negligible electrical field), presenting a positive potential, known as plasma potential (Vpl), on the order of +10 V [1], and the electrical field, which before breakdown was approximately linear between the electrodes, becomes restricted next to both the cathode and anode sheaths. The ions produced by ionization collisions (with high-energy electrons) that randomly reach the glow region-sheath interfaces are accelerated to the respective electrodes and can make additional collisions with neutral species, resulting, for example, in the known charge-exchange collisions in the sheath [1]. This is an important kind of collision, since fast (high kinetic energy) neutrals driven to the electrode surfaces are produced by this mean, explaining why not only fast ions bombard such surfaces but also fast neutrals.
Figure 3(b) corresponds to the usual kind of glow discharge used in dc plasma sintering process usually termed as the cathode configuration [5, 6], where the part to be sintered acts as the cathode of the discharge. There are two distinct regions with considerable electrical fields along the discharge, the cathode fall where the potential varies from the plasma potential down to the negative potential imposed by the power supply (Vsup), and the anode fall where the potential varies from the plasma potential down to zero at the grounded electrode, where the part could also be placed to undergo the sintering process.
At this point it is important to introduce the reader for the different ways or techniques for which a part can be sintered under dc plasma environmental.
Some different possibilities or techniques (discharge configuration) to carry out the plasma sintering process can be listed as follows:
Case 1: Sintering of parts in the cathode configuration in cold-wall chambers (without auxiliary resistive heating);
Case 2: Sintering of parts in the hollow cathode discharge (HCD) configuration in cold-wall chambers (without auxiliary resistive heating);
Case 3: Sintering of parts in the anode (or floating potential) configuration in cold-wall chambers (without auxiliary resistive heating); and
Case 4: Sintering of parts in any of the aforementioned configurations in hot-wall chambers (with auxiliary resistive heating).
For the first case, here termed as the cathode configuration in cold-wall chambers, the heating of the part at temperatures high enough for the sintering purpose (around 1100–1250°C, for iron parts) is achieved exclusively by the plasma fast species bombardment. For this case, if the gas mixture is composed of 80% Ar + 20% H2, pressure usually varies in the range of 10–30 Torr (1333–3999 Pa), considering pulse voltages of 600–700 V [11–16]. This condition, i.e., for parts acting as the cathode of the plasma device, is illustrated in Figure 3(b).
The second case is the hollow cathode discharge [8], as shown in Figure 4 that is a special kind of dc abnormal glow discharge, for which the same principles of the cathode configuration (see Figure 3b) are also valid. But in this configuration, by considering that both the cathode walls present a same potential, the potential distribution along the discharge is changed, being comprised by two similar electrical field regions next to each cathode surface, and the glow region that is virtually field-free. In this case, the grounded anode reference is outside from the hollow cathode region (not shown in Figure 4).
For the hollow cathode configuration, considering the same gas mixture, sintering temperatures can be attained by using lower gas pressures or voltage than those used in cathode configuration. In hollow cathode configuration, for a gas mixture composed of 80% Ar + 20% H2, the pressure range usually varies from 1 (133 Pa) to 9 (1200 Pa) Torr, considering pulse voltages on the order of 450 V [5, 6, 17–23]. Here, the heating of the part to the sintering temperature is carried out by the fast plasma species bombardment, and the part to be sintered acts as one cathode of the formed hollow cathode discharge. This situation is illustrated in the bottom of Figure 4, in which cylindrical parts to be sintered act as the inner cathode of a typical annular hollow cathode discharge, as presented in references [17–23]. It is to be noted that, for this configuration, plasma is much more ionized than that obtained in cathode configuration. This is due to the electron pendulum motion between the cathode falls that repulse electrons, keeping most of them arrested inside the discharge, thus increasing the collision probability and the ionization rate. As a reference, in a hollow cathode discharge, the current densities can be two orders of magnitude higher than that of a conventional discharge [7].
For the third case, parts are placed in the anode or in the floating potential (electrically insulated) so that no significant heating will be produced by plasma fast species bombardment. In this case, the discharge cathode is heated, and the part is mainly heated by radiation of the “hot cathode.” In addition, the very chemical-active plasma environment acts in the part surface, but the physical effects related to the fast species bombardment are strongly reduced.
Finally, all configurations presented here can be applied in a hot-wall chamber so that the heating effect can be partially supplied by an auxiliary-resistive heating system (the fourth case). For example, in sintering of parts in the cathode configuration in hot-wall chambers, the part heating can be performed by the fast plasma species bombardment as well as by the heat transferred from the auxiliary-resistive heating system. In this case, the main role of the plasma is creating a highly reactive atmosphere, as chemically as physically, by means of excited-species-enriched environment. The use of an auxiliary-resistive heating system makes possible to decrease the high risk of plasma instabilities when high power input is necessary to perform sintering. In addition, it tends to reduce the texturing effects caused by sputtering, an effect which is typical of surfaces sintered in cathode configuration, as presented in [5]. In the anode/floating potential configuration [5, 6], the heating is due to the heat transferred by radiation from the cathode walls heated by plasma species bombardment and by using an auxiliary-resistive heating system. In this case, the relatively intense fast plasma species (ions and neutrals) bombardment verified for parts acting as cathode does not occur, and the parts sintered in anode do not present significant texturing effect caused by sputtering.
Additionally, the anode configuration (with or without hot wall) also makes possible to perform simultaneous sputtering-deposition treatment with the sintering process. For this situation, atoms sputtered from the (hot) cathode wall can be deposited on the anode (and/or floating part) surface, being that surface alloying is also expected to occur, depending on the composition of each electrode as well as the plasma parameters. This configuration has also been used to perform the debinding of powder injection molding (PIM) parts [5, 6, 24–28]. It is worth to be mentioned that the anode configuration comprises the very well-established hybrid sintering furnace (or hybrid dc plasma reactor) designed for plasma-assisted debinding and the sintering of powder injection molded parts [24–28]. This system has already been used in industrial production of PIM parts. If it is the interest, the reader can access the furnace scheme presented in reference [5].
Independently from the chosen configuration, the dc plasma sintering procedure is frequently divided into four steps. For the case of iron parts, the first step is the cleaning of the parts usually under a H2 + Ar (or pure H2) glow discharge at 723 K (450°C) for 30 min, using 133 Pa (1–3 Torr) pressure. The three other steps are the heating stage of the parts to be sintered at moderate heating rates up to the sintering temperature, the sintering stage, and, finally, the cooling stage under the used gas mixture flow. Such procedure is valid for all above mentioned dc plasma sintering techniques. As a general rule, cathodes are negatively biased at pulse voltage of at least 450 V, but they can be set up to about 1000 V, generally using square form pulsed power supply with frequency ranging from 1 kHz up to few hundreds of kHz. The choice of the pulse voltage and duty cycle depends on the sintering technique and on the discharge parameters like the gas mixture and pressure, besides others. In addition, for the case of hot-wall chamber, the desired power to be transferred to the plasma can be chosen by setting the auxiliary heating system parameters (keeping in mind the sintering temperature to be attained).
AdvertisementFigure 5 shows typical scheme of a sintering contact between two particles (scheme adapted from Thümmler and Oberacker [29]) but here subjected to the dc plasma environment. In that case [29], the main sintering mechanisms occurring in metals under the total absence of plasma environment, excepting the plastic flow, are presented and carefully discussed, namely, superficial diffusion (way 1), evaporation and recondensation (way 2), volume diffusion (way 3), and grain-boundary diffusion (way 4), as presented in Figure 5.
Under the plasma environment, for particles next to the surface, the sputtering [1, 2, 4, 8–10, 30–36] caused by the fast plasma species bombarding the cathode surface becomes important, since it leads the density of sputtered metal atoms to be increased in the plasma phase, thus acting directly on the enhancement of the evaporation and recondensation mechanism (way 2). Moreover, the superficial diffusion mechanism (way 1) tends to be strongly incremented by the fast plasma bombardment, as well as the volume diffusion mechanism (way 3) at the surface, since the vacancy density next to the surface tends to be increased, as one of the possible events of the plasma-surface (interface) interaction (see Figure 6). Finally, as a result of the use of high-purity hydrogen in the gas mixture, associated with the hydrogen dissociation in the plasma environment, the oxide-reduction effect is higher than in conventional processes, and the grain-boundary diffusion (way 4) tends to be also increased. It would be due to a supposed higher cleanness (based on the oxygen-free boundaries) attained in grain-boundary sites (particle contact). If the reader is interested, diffusion in solids is very well treated in [37].
Figure 6 shows the typical events of the plasma-surface interface interaction. The main events occurring more significantly in the cathode (part) surface, where the species energy is the highest, can be listed as follows [1, 5, 6]:
Composition changes and chemical reactions as a result from the use of reactive gas species in the plasma (for all cases of dc plasma sintering techniques considered before) as well as of atoms sputtered from another cathode (it is supposed to be viable for the cases 2 and 3, only) being deposited in the part surface and diffusing into the substrate bulk;
Ion implantation, as a result of the use of high pulse voltages, more probable in the hollow cathode discharge, due to the presence of fast species of higher energy for this particular configuration;
Increase of the punctual defect density (like vacancies, interstitial, and/or substitutional atoms) in the first atomic layers of the substrate;
Reflection of impingent plasma species;
Emission of secondary electrons, which play important role in the discharge maintenance, since they are accelerated by the potential fall of the cathode sheath into the glow region, acquiring high energy;
Sputtering, consisting of surface (metal and/or nonmetal) atoms torn off from its original surface as a result of the high-energy plasma species bombardment; and
Heating of the cathode, as a result of the momentum transfer as the high kinetic energy species bombard its surface.
It is worth to be mentioned that in dc plasma sintering process one of the main roles of the plasma is providing a highly reductive atmosphere compared to conventional sintering furnaces. In this way, the smaller the oxygen species density in plasma, the higher is its oxide-reduction potential. Regarding the residual air pressure inside the vacuum chamber, it is to be noted that the oxygen partial pressure (PO2) must be controlled and kept to low values, usually much lower than 0.27 Pa (0.21 × 10−2 Torr, the usual residual pressure of primary vacuum systems). In this case, certifying that stanch gas lines are present, by reducing leakage points all over the plasma installation, is imperative for a good sintering process. In addition, since all the dc plasma sintering techniques considered here are carried out under gas flow, gas-washing procedure of the vacuum chamber, by intercalating the evacuation/high-purity gas filling of it, makes possible to additionally decrease the PO2 of the referred system.
At this point, the use of the parameter gas flow regarding the plasma environment should be stressed to the good understanding of the reader. A correct gas mixture flow to the plasma sintering treatment is important since the gas flow is responsible by changing the gas species in the vacuum chamber. As different oxygen sources can be present in the plasma sintering environment like oxygen-based species (e.g., H2O) adsorbed in the anode walls (internal to the vacuum chamber), oxides (present as oxide film on the powder particles) in the part to be sintered, and in organic compounds (like stearates used to lubricate the metal powder particles before the compaction step of the pressed part), the use of an adequate flow can prevent a possible chemical-potential change of the atmosphere from reductive to oxidant due to impurities being incorporated to the atmosphere. In previous work [23], the influence of the gas mixture flow on the processing parameters of hollow cathode discharge iron sintering was studied, and it was shown that by using a flow of 2 × 10−6 N m3 s−1, the stainless steel external cathode was completely oxidized, but by using a flow of 5 × 10−6 N m3 s−1, a clean and bright-aspect surface was achieved after the iron part sintering in the central cathode. It is worth to be mentioned that these values are only valid for the experimental apparatus and samples used in [23].
Regarding the gas utilized in plasma sintering processes, besides hydrogen (H2), which is a strong oxide reductor usually present in sintering gas mixtures, Ar is mainly used for the heating purpose. This is because its relatively high atomic mass makes the energy transfer for elastic collisions to the treating material more effective, resulting in heating of the metal part to be sintered. The hydrogen, since it is a strong reductive gas, is important to convert oxides into metal, considering that oxide layers are present in metal particles.
Finally, another important step to be considered in the dc plasma sintering process is the plasma cleaning stage before sintering. The cleaning stage is usually carried out using the same Ar + H2 sintering gas mixture, or eventually in a pure H2 predischarge at low pressure (normally at 3 Torr), before the isothermal sintering stage. It is aimed, in the cleaning stage, eliminating the undesired influence of microarcs, which can be originated from the plasma species bombarding organic molecules present in electrode surfaces, and can be responsible by leading the discharge regime change from the abnormal to arc. This step has also an important role on the degassing of the plasma chamber wall, reducing contamination by such species in the sintering step.
Contact us to discuss your requirements of Spark Plasma Sintering Machine Price. Our experienced sales team can help you identify the options that best suit your needs.
Now that the reader is familiar with plasma, to illustrate some important effects of plasma on the sintering process, in the next section, some results on plasma heating and surface modification during plasma sintering will be presented.
AdvertisementFigure 7(a, b, c) shows different heating curves for cylindrical iron cathode (10 mm diameter × 25.4 mm height) presented in the bottom of Figure 4 for different conditions by varying the discharge configuration (with and without hollow cathode effect) and plasma parameters. Figure 7(a) shows the temperatures attained by varying the switched-on time (equivalent to the duty cycle) at the hollow cathode discharge (HCD) configuration for different pressures. Figure 7(b) presents the temperatures attained by varying the cathode pulse voltage (Vp) and the switched-on time for both the cathode and HCD (for an intercathode distance a = 6 mm) configurations, and the same is presented in Figure 7(c) but for intercathode distance a = 9 mm. The indicated temperatures in Figure 7(a, b and c) were measured in the abovementioned cathode (usually termed as the central cathode of the hollow cathode configuration) by inserting a thermocouple inside it. For the cathode configuration, the cylindrical external (hollow) cathode was simply removed, thus leading to a heating condition where only one cathode (the central one) was present in the discharge. For more details, the reader can access the references [17–23]. In all cases, temperature increment can be expected as the switched-on time of the power supply, and thus the mean power transferred to the plasma is increased.
Considering other aspects, Figure 7(a) results also show the effect of the gas pressure on the cathode heating. Results indicate that the higher the pressure the higher is the ionization of the annular glow discharge, thus, the temperature achieved in the cathode. In the example shown using dashed red lines, for a specific switched-on time of 40 μs, considering square wave period of 200 μs (duty cycle of 20%), temperatures around 550, 750, and 1000°C are obtained for 0.6, 1.0, and 3.0 Torr pressures, respectively. At 0.6 Torr pressure, an annular glow discharge of instable operation is attained (indicated by the dashed black line for switched-on time ranging from 140 to 190 μs), being not adequate for the sintering purpose. This result agrees well with the expected for the product ap (a = inter-cathode distance and p = pressure), coming from more basic plasma studies, for which a stable hollow cathode discharge occurs for products ap ranging from 0.375 to 3.75 cm Torr [31]. Note that for a = 6 mm and p = 0.6 Torr, a product ap equal to 0.36 cm Torr is attained, which falls out of the proposed range for a stable condition. At 1 Torr, a duty cycle of 90% (maximum enabled by the power supply) has been used aiming to attain temperature high enough to perform iron sintering (~1100°C). And, for 3 Torr, the maximum temperature measurable by the K-type thermocouple was attained before the maximum duty cycle enabled by the power supply (~1250°C at 80 μs).
Figure 7(b and c) aims to put in evidence the role of the applied pulse voltage (Vp) on the cathode heating effect as well as the occurrence or not of the hollow cathode effect. It is evidenced that, as expected, the higher the voltage the higher is the temperature achieved in the cathode if considered the same configuration. On the other hand, considering the occurrence or not of hollow cathode effect, at 1 Torr pressure and a = 6 mm, for example, by using a 560 V pulse voltage, for a specific switched-on time of 80 μs, and considering a square wave period of 200 μs (duty cycle of 40%, see Figure 7b), the cathode temperature is increased from about 420°C, at the cathode configuration heating (CCH) to 1200°C, at the hollow cathode discharge (HCD). Finally, at 0.6 Torr pressure and a = 9 mm, for example, by using a 720 V pulse voltage, for a specific switched-on time of 140 μs, and considering a square wave period of 200 μs (duty cycle of 70%, see Figure 7c), the cathode temperature is increased from about 500°C, at the cathode configuration heating (CCH) to 1100°C, at the hollow cathode discharge (HCD). Being the interest, the reader can access additional details about the HCD in references [17–23].
In brief, from all results presented in Figure 7(a, b and c) as a general orientation regarding the cathode plasma heating, the use of higher pressures, pulse voltages, and duty cycles leads to higher glow discharge ionization with consequent higher current and dissipated power and thus the cathode heating to be improved, making possible to achieve temperatures high enough to sinter several metals. If the cathode configuration is considered only, in the absence of the hollow cathode effect, the use of pressures around 10–30 Torr is necessary to attain sintering temperatures for iron components, as confirmed in [11–16], whereas pressures around 3–10 Torr are enough in the HCD, as shown in Figure 7(a, b and c).
AdvertisementFigure 8(a–d) shows, for comparison purpose, SEM micrographs of the top (subjected to HCD) and bottom (non-subjected to HCD) of a cylindrical green (pressed) iron (Ancorsteel 1000C iron powder) sample (Figure 8a and c, respectively) sintered in the central cathode of a hollow cathode discharge (HCD), using external stainless steel cathode ((Figure 8b and d, respectively). In this case, sintering was performed at 1150°C, during 2 hours, using inter-cathode distance of 6 mm, with a flow of 5 × 10−6 N m3 s−1, in 80% Ar + 20% H2 gas mixture, and a pressure of 3 Torr. It is to be noted that in Figure 8(b and d), showing the surface condition after sintering, SEM micrographs were taken in the same positions of Figure 8(a and c), indicating the surface condition of the pressed iron sample, before sintering. The top of the sample was exposed to the plasma (as shown in Figure 9) aiming to put in evidence how effective the mass transport is affected by the discharge (plasma environment). Additionally to Figure 8 (a–d) micrographs, Figure 10 shows the Cr and Ni concentration profiles obtained in the originally pure iron-pressed sample surface, as a result of the deposition of metal atoms sputtered from the external cathode made of stainless steel on the pure iron sample top. All results shown in Figures 8(a–d) and 10 confirm the role of the sputtering, typically enhanced in HCD, in intensifying the mass transport in plasma phase and sample surface (as shown in Figure 5 near the particle surfaces), thus improving the sintering mechanisms of the superficial diffusion (way 1) and the evaporation and recondensation (way 2). In this case, the role of the sputtering has supposedly increased the metal atoms density in plasma phase, which tends to recondense (or to be deposited) next to the surfaces subjected to the intense fast plasma species bombardment, preferentially in concave areas. Note that by confronting the iron sample bottom surface (non-subjected to the plasma species bombardment) in a same position in its green (before sintering) and sintered states (Figure 8c and d, respectively), apparently no additional surface densification was verified. This fact confirms that the sample bottom was sintered through the known mechanisms expected for the conventional sintering (as shown in [29]). Nevertheless, contrarily, for the top surface (subjected to the plasma species bombardment), important surface densification was achieved, which is confirmed by comparing the iron sample top surface (before plasma species bombardment) in a same position in its green and sintered states (Figure 8a and b, respectively). In this case, the sample top was sintered with increased superficial diffusion (way 1), and the evaporation and recondensation (way 2) improved by sputtering-depositing mechanism occurring in plasma (as shown in Figure 5) that conclude this discussion.
AdvertisementConsidering all the presented phenomena involved in dc plasma sintering, it is expected that this process can present important technological advantage over “conventional sintering.” As presented in [5, 6], some advantages of the dc plasma-assisted sintering process in respect to conventional sintering technique, as a consequence of the plasma species bombardment, besides other physicochemical phenomena, could be listed as follows:
Possibility of surface diffusion increment;
Possibility of surface texturing obtainment (see Figure 8b results) and eventual surface densification of the sintered part;
Easy obtainment of highly reductive environment, enabling sintering of metals that tend to form very stable oxides (like Ti and stainless steels);
Possibility of surface alloying of the sintered part carried out simultaneously to the part sintering, due to the sputtering-depositing effects (i.e., the enrichment of the part surface with alloying elements); and
Finally, possibility of carrying out thermochemical treatments (such as nitriding, carburizing, and/or carbonitriding) that can be simultaneously performed in the sintering thermal cycle (as presented in reference [38]), or just after the sintering, in a single loading processing, by readjusting temperature and gas mixture, introducing reactive gases like nitrogen (N2) and/or carbon (e.g., by means of the use of CH4).
In this work, dc plasma-assisted sintering of metal parts is shown as a promising and relatively new research and development field in PM. The main aim of this work was to introduce the reader to the main applications of the dc plasma sintering process in the referred field. To achieve this goal, the present chapter was divided in a brief introduction and sections in which the bases of the abnormal glow discharge regime (dc plasma) were carefully treated, showing the main particularities of this new and “fascinating glow world” of dc plasma applied to the sintering of metal parts. Finally, the main techniques of dc plasma sintering, thermodynamic fundamentals of the plasma environment, aspects of the plasma heating, and plasma-surface-related phenomena of sintered parts were successfully treated, making possible to the reader to achieve a good understanding on the great potentialities of the dc plasma sintering process applied to the powder metallurgy.
AdvertisementThe authors would like to acknowledge the Brazilian agency CNPq for the financial support of this work.
This work is licensed under a Creative Commons Attribution 4.0 International License. The images or other third party material in this article are included in the article's Creative Commons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder in order to reproduce the material. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/
Electric current activated/assisted sintering (ECAS) techniques, such as electrical discharge sintering (EDS) or resistive sintering (RS), have been intensively investigated for longer than 50 years. In this work, a novel system including an electrically insulated graphite die for Spark Plasma Sintering (SPS) is described, which allows the sintering of any refractory ceramic material in less than 1 minute starting from room temperature with heating rates higher than 2000°C/min and an energy consumption up to 100 times lower than with SPS. The system alternates or combines direct resistive sintering (DRS) and indirect resistive sintering (IRS). Electrical insulation of the die has been achieved through the insertion of a film made of alumina fibers between the graphite die and the graphite punches, which are protected from the alumina fiber film by a graphite foil. This system localized the electric current directly through the sample (conductive materials) as in DRS and EDS, or through the thin graphite foil (non-conductive materials) as in IRS, and is the first system capable of being used under EDS or RS conditions independently combining current concentration/localization phenomena.
The sintering of ceramic materials from powders has been the subject of intense research over the last century. The use of high temperatures and long holding times in traditional techniques, such as pressureless sintering or hot pressing in conventional furnaces, has been a limiting factor for the industrial application of ceramic materials. There has therefore been a continuous quest for sintering techniques with higher energy efficiency by the scientific community. A promising approach is the use of electric current activated/assisted sintering (ECAS) techniques, in which the application of an electric current allows the sintering of ceramic materials in shorter periods of time compared to more traditional sintering techniques. There are many ECAS techniques to assist materials processing, such as powder wire discharge1, mechanical alloying2, powder consolidation3,4, chemical synthesis5 and solidification6, carried in different pieces of sintering apparatus with more than 600 patents registered7. ECAS processes could be divided into two main groups as proposed by Orru et al.8; (i) electrical discharge sintering (EDS); (ii) resistive sintering (RS). EDS is carried out for conductive materials as a high voltage electric discharge pulse, generated by a capacitor bank4,9, is forced to go through the volume of the sample located within an electrically insulating die8, usually made of borosilicate (BS) glass or pyrex. Softening of BS glass at high temperatures limits the use of EDS to ~800°C, which is too low for the sintering of refractory ceramic materials. Therefore, ceramists have focused on RS, which involves the application of low voltage (1–10 V) and high electric current through an electrically conducting die with the specimen under a mechanical pressure. Since its development in the early 1990s, spark plasma sintering (SPS) has been the most widely studied ECAS technique throughout the ceramic community. The industrial implementation of SPS has grown tremendously over the last few years as a consequence of the possibility of high sintering temperatures (up to ~2400°C), high heating rates (up to 600°C/min) and the application of mechanical pressures of up to ~100 MPa, which enable the sintering of ceramics in short periods of time compared with pressureless sintering or hot pressing (10–30 minutes instead of a few hours)10. In addition, SPS is suitable for the production of industrial scale samples with a maximum diameter of 300 mm.
In terms of energy efficiency, the power consumption of laboratory scale furnaces is around 5–10 kW for pressureless sintering, hot pressing (HP) and SPS. Recently, a transition at high temperature from electrically insulating to electrically conductive behavior has allowed the sintering of ceramics in few seconds (3–5 s)11 under low power dissipation conditions (~0.1 W/mm3) once the transition temperature is reached. This phenomenon is called flash sintering and is observed in ionic conductor and semiconductor ceramics11,12. During the flash event itself, flash sintering is more efficient than SPS but a conventional furnace is needed to reach the temperature required to induce the electrically conductive condition, thus compromising the global energy efficiency of this process. Therefore, the higher heating rate of SPS (up to 600°C/min) compared to conventional methods such as HP (50–100°C/min) and pressureless sintering (5–10°C/min) makes it the most energy efficient technique at present owing to the relatively short time over which energy is consumed. However, there remains room for improvement in the energy inefficiency during SPS processing since (i) most of the energy used is consumed in heating the graphite die rather than the specimen, (ii) the big cross section of the die produces a low resistive path by which the electric current can avoid the specimen itself. Therefore, current localization through or near to the specimen could improve the energy efficiency and shorten the process.
For this purpose, a flexible Al2O3 fiber-based film was used to electrically insulate the graphite punches from the graphite die and concentrate the electric current through the graphite punch column like in traditional EDS. Furthermore, a thin graphite foil was placed between the Al2O3 fiber-based film and the graphite punch column, as seen in . Unlike solid ceramic insulation, an Al2O3 fiber-based film is not sensitive to thermal shock and is also thermally insulating, which minimizes thermal loss through conduction to the graphite die. In addition, the resistance of the graphite foil is ~20 times higher than the resistance of the same length of the graphite punch for a diameter of 15 mm and graphite foil of thickness 0.2 mm assuming the same resistivity for both. This high resistance of the foil ensures that the current goes through the sample for conductive materials. With electrically insulating materials, the current goes through the graphite foil, which then acts as a highly-resistive heating element, localizing the electrical heat dissipation next to the specimen. Moreover, an Al2O3 fiber-based film could be placed between the specimen and graphite punches to force the current to go through the graphite foil independently of electrical nature of the specimen, as observed in . In addition, a piece of graphite foil is placed between the sample and the Al2O3 fiber-based film to avoid a chemical reaction during high temperature sintering as well as to improve thermal transport during ultra-fast processes in case of sintering materials with low thermal conductivity, such as Al2O3. This working mode is called forced resistive sintering (FRS) here.
The direct current mode was demonstrated using ZrB2, MoSi2 and ZrB2/20 vol. % MoSi2, all of which are electrically conductive at room temperature. These were densified by DRS for 60 seconds at different current levels as shown in . The bulk electrical resistivities of ZrB2 and MoSi2 are 10−7 13 and 2 × 10−5 Ωm14, respectively. Assuming a resistivity of 10−5 Ωm for the graphite foil, the foil was therefore approximately 2000 and 10 times more resistive than bulk ZrB2 and MoSi2 respectively, which ensured that most of the current went through the specimen in these experiments. shows that the relative density increased with current level, except for ZrB2, which exhibited a density reduction from 91.6% to 87.9% when the current increased from 1100 A to 1170 A. A detailed comparison of the process parameters between SPS and DRS for 91.6% and 87.9% dense ZrB2 is shown in . A sintering temperature of 1920°C, heating rate of 100°C/min, hold time of 5 min and a pressure of 80 MPa were selected based on a previous study15 and the limitations of the SPS current. Regarding the 1100 A sintered ZrB2, the maximum current for sintering was ~30% lower in DRS than SPS whilst the voltage is higher ( ), indicating a higher power dissipation ( ) all of which was concentrated on the ZrB2 specimen itself, which resulted in a DRS energy consumption of 7% compared to SPS in this case. Note that the temperature indicated in is a significant underestimate of the specimen temperature for DRS because it is measured on the unheated die rather than the specimen.
The relative density of the ZrB2 was 95.3% after SPS sintering. This relatively low value is mainly related to ZrO2 formation during sintering and residual porosity, which is related to B2O3 volatilization during sintering, as the density of ZrO2 (from 5.68 for cubic phase and 6.09 g/cm3 for tetragonal phase) is lower than for ZrB2 (6.12 g/cm3). This is difficult to avoid because ZrB2 powders contain ~1 wt. % of O impurity which transforms to ~3 vol. % of ZrO2 supposing that it completely reacts during sintering as follows:
SiC and MoSi2 are used as sintering additives to minimize ZrO2 formation during sintering and fabricate a pore free ZrB2-based composite15,16. Although the mean density of the best DRS specimen was lower (91.6%), the density in the centre of the specimen was 95.6%, similar to that achieved with SPS. The lower mean density was caused by the lower density outer ring which can be observed in . This could be removed by increasing the current ( ) but in this case the mean density was reduced to 87.9% as a consequence of partial melting during sintering in one of the specimen faces ( ), which was squeezed out leaving residual porosity. It should also be noted that the DRS was conducted with a lower pressure (14 MPa) than the SPS (80 MPa) in this work. Much higher densities and shorter sintering processes could presumably be achieved with the new technique by using higher pressures, which could potentially be of great help to fabricate nanostructured materials as the combination of fast heating and high pressure are key parameters to fabricate nanostructured materials17,18.
The rings formed during DRS ( ) are thought to result from the thermal gradients in DRS as the temperature is reduced near the edge by conduction to the unheated surroundings. In the centre of the sample, the ZrB2 grains are completely surrounded by thin regions of C particles ( ). The grain boundary C becomes more equiaxed and distinct far from the centre ( ) and the ZrB2 grains become smaller. Near the edge of the sample, discrete equiaxed C particles are visible and they no longer decorate the ZrB2 grain boundaries ( ). Pores are observed instead of C particles in the ZrB2 fabricated by DRS at 1100 A and the ZrB2 fabricated by SPS as seen in , respectively. However, DRSed ZrB2 pores contain small particles of ZrB2 and ZrO2, checked by energy dispersive X-Ray spectroscopy (EDX), while empty pores are observed in SPSed ZrB2. Furthermore, the grain size of DRSed ZrB2 is visibly smaller than SPSed ZrB2. Ultra-fast heating higher than 2000°C/min avoids intermediate temperature coarsening, leading to faster sintering and lower grain growth compared to SPS in spite of using lower pressure during sintering. Therefore, electric current concentration could be suitable for the sintering of nanostructured materials.
B2O3 volatilization during sintering leaves residual porosity, as observed in . C phase is observed at grain boundaries in and their corresponding C maps in . In addition, O maps are observed in , which show high O brightness in some boundaries between C and ZrB2. shows a microstructure after polishing and removing the melting track from the sample shown in , which is clear evidence of a ZrB2-C eutectic formation. Therefore, maximum temperature is characterized by melting point of ZrB2-C eutectic, which is 2390°C for ~33 mol% C-ZrB219 and indicate that gradient temperature in DRS mode is ~800°C. Furthermore, eutectic formation limits the maximum temperature for ZrB2 strength testing with graphite fixtures to ~2300°C20 and ZrB2 sintering using a graphite mold. Whether the C precipitates at grain boundaries or is vapor deposited is characterized by the occurrence of C sublimation.
Maximum temperature for ZrB2 sintered at 1170 A is ~2400°C that has a vapor pressure for C sublimation of 0.1 Pa21, which is lower than 3 Pa of pressure maintained during experiments. Therefore, all carbon observed in is formed through the diffusion of C from the graphite film and punches over the ZrB2 lattice and precipitation at ZrB2 grain boundaries. Furthermore, illustrates a material produced under out of equilibrium conditions. Nonetheless, it points out the potential of DRS for producing C-MeB2 composites (Me = Ta, Zr, Hf), which could be of great technological interest for the scientific community as for example the bonding and impregnation between C and MeB2 is a current challenge to minimize ablation during hypersonic re-entries22. However, the current/pressure profile needs to be optimized for this purpose.
C deposition in ZrB2 pores is likely to be characterized by a voltage/resistivity reduction in . ZrB2-C eutectic formation also accounts for relative density drop observed in . Moreover, the observation of grain boundaries revealed the existence of Zr-oxycarbides ( ) as a consequence of the reaction between C and ZrO2 confirmed by EDX ( )23. Zr-oxycarbides particles were observed just in few grain boundaries as ZrO2 is a residual phase (1 wt.% of O impurities). Recently, it has been reported an onset of 1600° C for the generation of Zr-oxycarbides with different C/O ratio in ZrB2-based composites. In the cases of MoSi2 and ZrB2/20 vol. % MoSi2, the density and microstructure was more homogeneous throughout the sample ( ), indicating that homogeneity could be controlled under DRS. The SPS temperatures to fully sinter MoSi2, ZrB2/20 vol. % MoSi2 and ZrB2 are 1300°C24, 1750°C16 and 1950°C15, respectively.
In the results above, the specimens conduct electricity but the graphite foil is an electrically conductive layer, which could conduct electricity when using an electrically insulating ceramic such as Al2O3, producing a RS process. Therefore, heat dissipation could be localized through the volume of the sample (DRS) or through the graphite foil (IRS) depending on electric nature of the sample. This is the first system to date suitable for EDS and RS, either DRS or IRS. Al2O3 was melted using an electric current as low as 350 A in ~52 seconds. This level of current is not sufficient to heat the conventional 15 mm graphite SPS die to 600°C in 2 minutes, as observed in . The melting of the Al2O3 is characterized by a power spike in as the voltage increases to keep the current constant as the fluid Al2O3 is extruded between the graphite punch and the graphite foil, reducing the conductive path and increasing the resistance. This example illustrates the magnitude of the energy savings of this new RS setting of the graphite die and the potential of the localized current sintering (LCS) technique. Moreover, temperature measured with the external pyrometer was only ~1100°C by the time the Al2O3 melted, which indicates the temperature gradient as a consequence of electric and thermal insulation could be as high as ~1000°C.
When a lower current of 340 A was used, melting was avoided and a 96.2% dense Al2O3 specimen was obtained after supplying this current for 60 s, as observed in .
The FRS setting of was also used with alumina. In this case, 98.1% dense Al2O3 was produced using a current of 410 A, as shown in . For a truly insulating specimen, little difference in response would be expected between the FRS setting and the IRS setting and the difference observed here indicates that at the high temperature during sintering, a significant current was passing through the alumina in the IRS setting.
Although the FRS setting was successful in sintering the specimen, the prevention of direct energy dissipation within the specimen reduced its energy efficiency compared to IRS. The FRS energy consumption was 50% higher than by IRS in case of Al2O3.
shows the microstructure of the dense Al2O3 fabricated by FRS. The microstructure consists of large Al2O3 grains (~3 μm) with clusters of nano-grains. This is a new microstructure for Al2O3, which was evidently obtained by an out-of-equilibrium process. This demonstrates the possibilities for making new microstructures with potentially new properties by these rapid heating methods.
To explore the relative merits of the different techniques, electrically conductive MoSi2 was also densified by FRS. Dense MoSi2 could be fabricated by FRS using 410 A, which is half of the current required by DRS. shows the microstructure of dense MoSi2 fabricated by DRS, in which dislocations are visible in MoSi2 suggesting the presence of plastic deformation during densification by DRS.
It is clear that with suitable optimization, dense specimens can be made with either the DRS/IRS or the FRS setting. There are three main considerations in optimizing these. The first is the uniformity and extent of heating. It was evident from the inhomogeneity of the structures in that the DRS mode for conductive specimens can lead to severe temperature gradients because most of the heating is within the specimen and the edges are therefore in contact with relatively cold material. The success of the FRS setting in producing fully dense alumina and MoSi2 is at least partly because much more heating is from the graphite foil around the exterior of the specimen, conducting the heat inwards, producing a more uniform temperature profile. According to the material, the thickness of the graphite foil and therefore its resistance, and the properties of the insulating layer can potentially be balanced to give uniform temperatures and therefore improved sintering. The geometry of resistive elements could also play a role in tailoring the electric field in order to induce electric field assisted sintering (FAST) effects25. Similar considerations may apply to the loss of heat from the specimen to the cold graphite punches, which could be reduced with the aid of thermal insulation, including the alumina film used in this work which has a thermal conductivity of 0.2 W/(mK) while for graphite it is 81 W/(mK) according to manufacturers data.
A second optimization is for the energy efficiency. All three settings are obviously more efficient than SPS because only the specimen and its immediate surroundings are heated and because the processing times are much shorter. In the first attempt at sintering ZrB2 above, for instance, the improvement was by more than an order of magnitude but systematic balancing of the heating and optimization of the processing time could lead to even more dramatic reductions in energy consumption.
The third optimization is related to more accurate temperature readings. Temperature readings in the outer graphite die wall are from ~800 to 1000°C lower than real sample temperatures because of; (i) the rapidity of the process and (ii) the electrical and thermal insulation provided by the Al2O3-based film. Therefore, using a pyrometer to read the temperature on the top graphite punch near the sample, which is available in bigger and more modern SPS furnaces than the one used herein, would result in a more realistic temperature value for DRS and IRS processes. However, this limitation is still present in FRS processes although reproducibility is more accurate than in pyrometer-controlled sintering processes as DRS, IRS and FRS fully depend on current profile.
The fourth optimization, and a further benefit of these methods, is in minimizing the capacity and therefore the capital expense of the equipment used for a particular size of specimen. Recently, a concern has been raised about size limitations of samples sintered by SPS due to the need of higher currents to maintain high heating rates (100°C/min) and high temperatures for industrial size components26. In fact, direct current SPS technology is limited to the supply of 48 kA for a maximum sample size of 300 mm. The new LCS technique can be matched to the maximum current of the machine and achieve large specimens, more efficiently with smaller and less expensive power sources. The production of large samples would also benefit from the absolute reduction of energy used and the tailoring of thermal homogeneity in the other two areas of optimization.
In summary, the design of a novel electrically heated hot pressing system in which the die is electrically insulated from the rest of the system is described here. A current is passed through the rams and can either bypass the specimen through a graphite foil or pass through the specimen, according to the conductivity of the specimen and whether it is electrically insulated from the rams or not. The electrically insulated die is structurally equivalent to SPS die but the system is electrically equivalent to a micron sized die, which dissipates heat more locally and more quickly than a conventional SPS graphite die. This localized current sintering combines and improves on the advantages of existing techniques and enables the sintering of advanced materials and refractory ceramics by EDS and RS, either DRS/IRS or FRS, in less than 60 seconds from room temperature. The minimum contact pressure (14 MPa) was supplied in this study to demonstrate the success of this technique under low pressures but further improvements in densification could be produced by the use of higher pressure. Constant current profile was used to fabricate dense ceramics in 60 seconds. However, the combination of higher pressures up to 100 MPa in addition to optimization of heating/current profile during sintering could potentially lead to fabrication of nanostructured materials. Further advantages are energy savings of 1–2 orders of magnitude compared with SPS and the potential to produce novel microstructures owing to the high heating rate in LCS that is potentially a technique to fabricate low-cost, out of equilibrium materials, which is one of the challenges for the international ceramic community over the coming years27.
For more Hot Press Sinteringinformation, please contact us. We will provide professional answers.