Plasma Chemistry and Plasma Processing, Vol. 19, No.

1, 1999

Thermal Plasma Technology: Where Do We Stand

and Where Are We Going?1
E. Pfender2
Received January 7, 1998; revised March 7, 1998

In this overview, an attempt is made to assess the present and future research and
development in thermal plasma processing of materials restricted to (1) thermal
plasma coating technologies, (2) thermal plasma synthesis of fine powders, (3)
thermal plasma waste destruction, and (4) thermal plasma spheroidization and
densification. Since thermal plasma processing is, in general, governed by a large
number of parameters, implementation of controls becomes mandatory. The lack
of sufficient controls combined with economic drawbacks in some cases has been
the main obstacle for the growth of thermal plasma technology. Present R&D
efforts, however, address these problems.

KEY WORDS: Thermal plasma technology assessment; research and develop-

ment; overview.

1. INTRODUCTION
Thermal plasma technology has passed through a gradual transition
stage from primarily space-related activities in the sixties to a more and
more materials-oriented focus in the eighties and nineties. Space-related
needs provided a strong impetus for basic thermal plasma research and
developments as, for example, the development of plasma torches covering
power levels ranging from 1 kW to more than 10 MW. Research specifically
geared toward an understanding of plasma/particulate interaction and the
chemistry in thermal plasmas, however, did not commence until the early
eighties, although some successful applications were already in existence at
that time as, for example, in the area of arc welding, arc cutting, and in
some specific areas of plasma synthesis of fine powders. At this point it
should be emphasized that the successful development of arc circuit breakers
and of arc lamps contributed immensely to our basic understanding of
electric arcs.
1
This paper has been presented as a keynote lecture at the 13th International Symposium on
Plasma Chemistry, Beijing, PRC, August 18-22, 1997.
2
Department of Mechanical Engineering and ERC for Plasma-Aided Manufacturing, Univer-
sity of Minnesota, Minneapolis, Minnesota 55455.
1
0272-4324/99/0300-0001S16.00/0 C 1999 Plenum Publishing Corporation
2 Pfender

In addition to these more conventional applications, thermal plasma

technology covers today a wide spectrum of applications as well as new
developments which may be classified as (1) thermal plasma coating tech-
niques, including plasma spraying, wire arc spraying, and plasma chemical
vapor deposition (TPCVD); (2) thermal plasma synthesis of fine powders,
in particular, powders in the nanometer size range; (3) thermal plasma waste
destruction, in particular, of toxic waste materials; (4) thermal plasma
densification of powders; (5) thermal plasma metallurgy, including melting
and remelting applications in large furnaces; and (6) thermal plasma extrac-
tive metallurgy.
In this overview, an attempt will be made to assess the present situation
in this field and to speculate on the role which thermal plasma technology
might play in the much broader context of material science and engineering
as we approach the next century.
Because of space limitations, only the first four of the previously men-
tioned thermal plasma technologies will be covered in this overview.

2. THERMAL PLASMA COATING TECHNOLOGIES

This section will include plasma spraying, wire arc spraying, and ther-
mal plasma chemical vapor deposition (TPCVD).

2.1. Plasma Spraying

Over the past 35 years, plasma spraying has become a well-established
and widely used technology with applications ranging from corrosion-, tem-
perature-, and abrasion-resistant coatings to the production of monolithic
and near-net shapes of metallic and ceramic parts. Powders of "glassy"
metals can be plasma sprayed without changing their amorphous character-
istics and, as demonstrated in recent years, superconductive materials can
be deposited by the plasma spray process.
Besides the most common atmospheric pressure plasma spray process
in ambient air (APS), other plasma spray processes have been developed,
including spraying at low pressures (LPPS), at supersonic velocities, under
controlled ambient conditions (for example, in argon), and even under
water. The design of plasma spray torches for the various plasma spray
processes has been essentially the same, based on producing a plasma jet by
a dc arc operated between a stick-type cathode and a nozzle-shaped anode
Thermal Plasma Technology: Where Do We Stand and Where Are We Going? 3

Fig. 1. Schematic of the plasma spray process with dc plasma torch.

as shown schematically in Fig. 1. Recent torch developments, however, have

been exploring central injection of the powder particles into the plasma,
shrouding of the plasma jet, and a combination of shrouding with antivor-
tex flow in order to improve deposition efficiency and quality of the coat-
ings. A new approach, considering plasma spraying using inductively
coupled high-frequency (RF) plasmas, is attracting increasing attention as
an efficient thermal spray process. Among recent innovations, automation
and robotics applied to the plasma spray process are probably the most
important new developments in this rapidly growing field.
In spite of these impressive developments, some of the underlying fun-
damentals of the plasma spray process are still poorly understood. This
applies to the characteristics of the plasma jet as well as to the interaction
of powder particles with the plasma and also to the formation of the coating
on a substrate.
For the case of dc plasma spraying (APS), the fluid dynamics of the
plasma jet leads to the development of turbulence associated with strong
entrainment of ambient gas into the plasma jet(1,2) as shown schematically
in Fig. 1. This behavior of a plasma jet as sketched in Fig. 2 has been
confirmed by shadowgraphs and CARS spectroscopy,(3) by conditional sam-
pling experiments,(4) and by probe sampling of the plasma jet.(5)
Superimposed on these fluid dynamic effects are a surging and whip-
ping motion commonly seen in time-resolved photographs of plasma jets
(Fig. 3). This motion is caused by axial and circumferential motion of the
anode arc root within the anode nozzle giving rise to arc instabilities.(6,7)
Correlations between these arc instabilities and arc voltage acoustic, and
light emission fluctuations(2) confirm that the observed plasma jet fluctu-
ations (Fig. 3) are to a large extent due to arc instabilities.
Recent studies of a commercial spray torch, considering both vortex
and straight flow of the plasma gas, revealed fluctuations in the frequency
4 Pfender

Fig. 2. Main regions of a transitional plasma torch.

range from 2 to 6 kHz and a strong dependence of the arc behavior on the
plasma gas composition (Ar/H2, Ar/He, Ar/N2 mixtures).(8) These fluctu-
ations affect both torch performance (anode lifetime) and coating quality.
Attempts have been made to measure the motion of the arc root in the
anode nozzle by using magnetic probes(9) and also to calculate the anode
arc root position in the anode nozzle(10) using Steenbeck's minimum
principle.(11,12) Experimental results are in reasonable agreement with ana-
lytical predictions.
The previously discussed fluctuations of the plasma jet may lead, in
extreme cases, to the situations shown in Fig. 4 where the injected powder
particles miss a substantial fraction of the hot plasma jet.(13) This may lead
to a severe degradation of the quality of the coatings due to unmelts and
only partially melted powder particles in the coatings.
Attempts to model the plasma spray process require, as a first step, a
comprehensive model of the plasma jet. Unfortunately, modeling of plasma
jets faces several obstacles as indicated by experimental observations. Diag-
nostics of plasma jets using emission spectroscopy laser scattering (Rayleigh
and collective Thomson scattering), laser Doppler anemometry, enthalpy
probes, and mass spectrometry (for references, see Ref. 14) indicates that
Thermal Plasma Technology: Where Do We Stand and Where Are We Going? 5

Fig. 3. Short time exposures (50 ns) of a de plasma jet.

Fig. 4. Schematic of the two-fluid model and boundaries for the computational domain.
6 Pfender

strong deviations from local thermodynamic equilibrium (LTE) may prevail

over almost the entire plasma jet volume. This fact, combined with the pre-
viously mentioned fluctuations of typical plasma spray jets, imposes severe
difficulties on modeling attempts. A realistic model must also take the large-
scale entrainment of ambient gas into account. Conventional turbulence
models cannot predict the intermittency and the unmixing phenomenon of
turbulent flows. These models neglect the "spottiness" or "fragmentariness"
of real turbulent flows which have been observed in many situations. Only
a multiphase model for turbulent flow can reproduce these effects.(15)
The model that is based on Spalding's approach(15,17) treats the plasma
jet as a two-fluid mixture consisting of hot, out-moving fragments and cold,
in-moving fragments as shown schematically in Fig. 4.
The governing equations include the transport equations for mass,
momentum, and energy for two different fluid parcels (in-moving parcels
and out-moving parcels). Auxiliary relations that govern the physical
phenomena of the interfluid mass, momentum, and energy exchange are
used together with a description of the mechanisms that control the growth
or diminution of the fragment size. The results may be presented in con-
ditional- and unconditional-averaged forms and compared with experimen-
tal results from enthalpy-probe measurements.(18,19)
Since the quality of a plasma-sprayed coating depends heavily on the
properties of particulates, including velocity, temperature, degree of melting,
and their statistical distributions, considerable efforts have been made to
develop predictive capabilities for the particle behavior in plasmas. An
initial simplified approach in which particle effects have been decoupled
from the plasma flow(20) has now been extended to iterative techniques(21)
to include the effects of the particle cloud on the plasma flow, and to the
stochastic particle spray model incorporated into the LAVA code.(22) The
LAVA code allows one to predict particle trajectories and thermal histories
including melting which is calculated simultaneously with the motion of
the gas in a fully self-consistent manner, including turbulent dispersion of
particles. Combined with transient, multicomponent, and non-LTE capa-
bilities, the stochastic particle spray model designates LAVA as a unique
comprehensive computational model.

2.1.1. Summary of Present R&D Efforts in Plasma Spraying

The focus of present research efforts seems to be on both diagnostics
and modeling with the goal of:
(a) improving and optimizing plasma torch performance. This includes
studies of plasma jet instabilities caused by the fluid dynamics and
the associated arc behavior in typical plasma spray torches. Closely
Thermal Plasma Technology: Where Do We Stand and Where Are We Going? 7

related to the arc behavior are problems of electrode erosion. A

new approach for modeling of turbulent plasma jets has been pro-
posed using a two-fluid description of the plasma and its surround-
ings. Comparisons with experimental data have been hampered by
fluctuations of plasma jets which, so far, have not been included in
corresponding models. Recently, attempts have been reported to
incorporate fluctuations in such models.(23)
(b) Another major objective of present research efforts is the control
of the particle spray pattern and the associated optimization of the
deposition efficiency. This aspect is of particular concern when
spraying of expensive powder materials is considered. Besides depo-
sition efficiency, the quality of coatings may also be strongly affec-
ted by the spray pattern.
(c) Some of the most recent research efforts have concentrated on coat-
ing formation on a substrate which includes splat formation, solidi-
fication of splats, splat microstructure, cohesion of neighboring
splats, adhesion of the coating to the substrate, and the control of
porosity and of unmelts in the coating.(24) It has been found that
the substrate temperature may be the governing parameter in terms
of coating adhesion.(24) Residual stresses which are inherent to this
coating process may be the determining factor for integrity or fail-
ure of a coating.

Since complete control of the plasma spray process is the ultimate goal
of all R&D efforts in this field, the establishment of comprehensive data-
bases and the selection of primary control parameters attracted particular
attention over the past years. This does not only apply to the APS, but also
to other plasma spray processes.
Recent developments are primarily concerned with:

(a) Sensor development. Such sensors must be robust to tolerate the

hostile plasma spray environment, but at the same time they have
to be affordable (cost-effective).
(b) Among various control strategies, feedback control of the plasma
jet behavior (enthalpy level and fluctuations of the jet) is considered
to be a viable option.
(c) Control of temperature and velocity of powder particles injected into
the plasma is a challenging problem and many of the leading
plasma spray laboratories all over the world are engaged in solving
this problem.
(d) In-situ control of the coating thickness is another challenging prob-
lem which has not yet been solved.
8 Pfender

Fig. 5. Schematic of the double-wire arc spray process.

(e) Substantial advances in the area of robotics have already been dem-
onstrated for both motion of the spray torch and the substrate during
the coating process.
Over the past years, applications of plasma spraying experienced a
slow, but consistent growth. As this technology further matures to the level
of complete, automated control, a more rapid growth of this technology
can be expected, because the lack of efficient controls has been the primary
obstacle for the growth of this field.

2.2. Wire Arc Spraying

Wire arc spraying is an inexpensive coating process, usually restricted

to spraying of metals and alloys. A wide array of substrate materials can
be coated with this process, including ceramics, metals, and plastics. The
applications range from wear-resistant coatings, to coatings for corrosion
protection of large-scale structures (for example, bridges), and to the resto-
ration of worn metallic parts. Although similarities with the plasma spray
process exist, there are fewer parameters governing this process.
Figure 5 shows a schematic of the dual wire arc spray arrangement.
The material to be deposited is introduced into the arc in the form of two
wires serving as consumable arc electrodes. A gas jet across the arc removes
Thermal Plasma Technology: Where Do We Stand and Where Are We Going? 9

Fig. 6. Micrographs of spray droplets captured in ice.

molten droplets from the wire tips, atomizes droplets, and drives them to
the substrate. The coating is formed by the impact, deformation, and rapid
solidification of individual molten droplets on the substrate, resulting in a
coating structure consisting of a series of overlapping lamellae. The
adhesion of the coating depends upon the interactions among individual
lamellae and between lamellae and the substrate. The bonding mechanisms
of arc-sprayed coatings are still poorly understood and process parameters
are still optimized by empirical methods. Excellent adhesion of coatings
is considered to be one of the most important prerequisites for industrial
application.(25,29)
If the interface adhesion is poor, coating detachment may occur
resulting in premature failure. The bond strength of a coating depends on
the extent of both physical and chemical interactions between the coating
and the substrate material and on the microstructure of the interfacial
region. Poor adhesion may be attributed to poor interfacial interlocking,
low degree of metallurgical bonding, and high internal stresses.
Adhesion strength depends to a large degree on particle velocities. In
conventional wire arc spraying, the velocities of particles are subject to cer-
tain limitations, so the coating produced with conventional primary gas
atomization has relatively high porosity and relatively low bond strength.
Secondary gas atomization spraying is a newly developed approach for
achieving more uniform particle size distributions, more focused spray pat-
terns, higher particle velocities, and improved coating properties.(30) As an
example, Fig. 6 shows the effect of secondary gas flow on the droplet sizes.
These SEM micrographs have been obtained from droplets sprayed onto a
sheet of ice for rapid solidification.
It has been shown that secondary gas-sprayed stainless steel coatings
on aluminum substrates reveal higher bond strength than those sprayed
10 Pfender

with only primary gas. The following three kinds of bonding mechanisms
are operative in spraying with secondary gas atomization: (1) physical bond-
ing, (2) mechanical bonding, (3) metallurgical bonding. The improved bond-
ing is believed to be due to the higher temperatures of the liquid metal
droplets at the instant of impact on the substrate. These higher temperatures
are the result of higher droplet velocities and of reduced entrainment of cold
air, leading to an interdiffusion layer between the substrate and the coating.
The presence of this interdiffusion layer as verified by elemental analysis of
the coating cross section is associated with the measurement of improved
bond strength.(30) Secondary gas flow can also reduce mixing of atomizing
gas with the surrounding air resulting in less oxide content and less chro-
mium loss (for stainless steel) in the coating.(31) For a further reduction of
the oxide content in the coating, carbon dioxide has been used as the atom-
izing gas, which also reduced the porosity in the coatings.(31)
One of the drawbacks of wire arc spraying is the generation of substan-
tial amounts of metal fumes which pose a human health hazard. Studies of
fume generation as a function of operational parameters for the related
process of arc welding have been reported in the literature.(32-34) The control
of fume generation at the source by modification of the process would lead
to a wider acceptance of wire arc spraying for industrial applications. In
recent experiments, using aluminum as the wire material, fume generation
has been quantified with computerized image processing. The results show
that metal evaporation at the wire tips is the primary source of fumes, and
oxidation seems to enhance metal evaporation. There is a pronounced asym-
metry of fume generation, because of the current concentration (high cur-
rent density) at the cathode.(35)

2.2.1. Summary of Present R&D Efforts in Wire Arc Spraying

Present research activities are geared toward improvement of coating
quality by:
(a) control of the arc,
(b) optimization of the flow, and
(c) control of the droplet formation.
Coating quality refers primarily to coating adhesion and cohesion, den-
sity (porosity), uniformity, oxidation, and thickness control. It should be
pointed out that the previously mentioned control functions are not inde-
pendent of each other. Arc and flow control, for example, are directly
coupled to the droplet formation.
Control of the arc has to be done in conjunction with the power source,
because the characteristics of the power source will affect the arc behavior.
Thermal Plasma Technology: Where Do We Stand and Where Are We Going? 11

Arc fluctuations which manifest themselves by arc voltage fluctuations

should be minimized and arc extinction must be avoided.(36)
Optimization of the flow includes the primary or atomizing gas flow
(supersonic), the secondary or shrouding flow, the type of gas being used,
and its temperature. The effects of turbulence on air entrainment and its
reduction by gas shrouding are active research objectives. In this context, it
should be mentioned that modeling of the wire arc spray process, including
flow effects, is another active research topic.
As previously mentioned, there are substantially fewer parameters
which affect the wire arc spray process compared to the plasma spray pro-
cess. Therefore, developments which are already in progress and which focus
on complete control systems of the wire arc spray process have a high poten-
tial for early success.

2.3. Thermal Plasma Chemical Vapor Deposition (TPCVD)

Over the past years, a new technology has been initiated which makes
use of thermal plasmas for chemical vapor deposition of an array of mater-
ials, and this process gained prominence by demonstrating the highest depo-
sition rates for diamond coatings. This emerging technology will fill the void
between the high quality, but extremely slow coating processes (sputtering,
physical vapor deposition) and the rapid, but difficult to control thermal
spray processes.
Because of its unique features, TPCVD may find numerous appli-
cations, and some of them are already considered for industrial fabrication
as, for example, diamond and dense ceramic or superconducting films.
There are indications that TPCVD may play an important role in the fabri-
cation of nanostructured films. As this emerging technology further
matures, there is no doubt that other applications will come into the picture.
In the process of TPCVD, a high-energy-density plasma produces high-
density vapor-phase precursors for the deposition of relatively thick films.
A typical arrangement for TPCVD is shown in Fig. 7. In this case, a dc
plasma torch generates a high-temperature, high-velocity plasma jet which
impinges on a cooled substrate. With temperatures close to the torch nozzle
exit exceeding 104 K, the precursor material, which is injected into the
plasma, is rapidly vaporized and, due to the high velocities of the plasma
jet (on the order of 100 m/s), accelerated toward the substrate. In front of
the cooled substrate, a boundary layer forms, characterized by steep gradi-
ents. Such boundary layers in chemically reacting gases attracted strong
interest in connection with space flight and re-entry simulation and, there-
fore, such boundary layers have been extensively analyzed.(37)
12 Pfender

Fig. 7. Schematic of diamond TPCVD with a dc plasma torch.

Besides dc plasma jets, other types of plasma reactors have also been
utilized which make use of high-frequency (RF), hybrid and microwave
plasmas. For increasing the available plasma volume, multiple torch
arrangements have been developed. A typical reactor with three torches
mounted in a way that the three jets coalesce to form an extended plasma
region (Fig. 8), and with reactant injection through a water-cooled probe
into the region where the three jets merge has been used for a parametric
study of the influence of the process parameters on the film
characteristics.(38,39) Typical power levels have been 36 kW total power, and
deposition rates of 40-60 /im/h are achieved at reactor pressures of 270 torr.
This reactor has also been used for the homoepitaxial growth of diamond
films with growth rates of 100-200 ^uM/h.(40,41)
As an example, Fig. 9 shows typical diamond films produced in this
reactor without (Fig. 9a) and with (Fig. 9b) secondary precursor injection.
Secondary precursor injection through a ring surrounding the plasma plume
(Fig. 8) provides more uniform precursor distribution over the substrate
surface, which has a pronounced effect on the surface morphology of the
deposited films. Figure 9b indicates an almost uniform film morphology
over the entire substrate diameter (30 mm).
Present research efforts dealing with this emerging technology are pri-
marily concerned with establishing the knowledge base, necessary for
further developments. Both experiments and modeling efforts consider the
chemistry in the boundary layer in front of a substrate in order to find the
Thermal Plasma Technology: Where Do We Stand and Where Are We Going? 13

Fig. 8. Schematic of the triple-torch plasma reactor for TPCVD.

most important growth species and to assist in understanding of the growth

process. In the case of diamond CVD, growth rate and crystal structure and
orientation as a function of the governing parameters are some of the most
important research topics.
As an example, Fig. 10 shows an arrangement for determining precur-
sor species in front of a substrate during diamond deposition in a RF
plasma.(42,43) This arrangement is very useful for manipulating the thickness
of the boundary layer in front of the substrate by varying the argon injection
flow rate through the injection probe indicated in Fig. 10. By increasing the
14 Pfender

Fig. 9. (a) Micrographs of diamond films deposited with the triple-torch plasma rector without
secondary precursors, (b) Micrographs of diamond films deposited with the triple-torch plasma
reactor with secondary precursors.

argon flow rate and keeping the other parameters the same, the boundary
layer thickness may be reduced and the effect of this reduction on the
boundary layer chemistry may be determined. Corresponding modeling
work shows that the importance of hydrocarbon and carbon species for
diamond shifts as the thickness of the boundary layer changes, as shown in
Fig. 11. Both the growth chemistry and the growth rate are strongly affected
by the thickness of the chemically reacting boundary layer in front of these
substrates.
For relatively thick boundary layers (=2 mm), the model predicts dia-
mond growth rates of =10jUm/h in accord with experimental data, with
the dominant growth species being methyl. For very thin boundary layers
(=0.1 mm) the model predicts diamond growth rates of several hundred jum/
h, with a transition to monatomic carbon as the dominant growth species
as shown in Fig. 11.(44,45)

2.3.1. Summary of Present R&D Efforts in TPCVD

Present research efforts focus on three aspects:
(a) Boundary and substrate surface chemistry. As previously men-
tioned, both experiments and modeling efforts have concentrated
on finding the primary growth species in front of the substrate in
Thermal Plasma Technology: Where Do We Stand and Where Are We Going? 15

Fig. 10. RF plasma reactor for TPCVD of diamond films with attached quadrupole mass
spectrometer.(42,43)

conjunction with modeling of the boundary layer and surface

chemistry.
(b) Control of film morphology and film quality are current research
topics in this field. Film quality is of particular concern for optical
and electronic applications of diamond films.
(c) Film growth rates and film adhesion. Although film deposition
using thermal plasmas as a tool for generating growth species result
in rather high growth rates, the quality of the deposited films may
or may not be sufficient for the intended application. Producing
high quality films at high growth rates remains still a challenge.
Film adhesion is one of the major problems in the field of diamond
deposition, especially for deposition on substrate materials which cannot
tolerate high temperatures (<500°C) and on a variety of metallic substrates
such as, for example, steel.(46) By using intermediate layers of materials to
16 Pfender

Fig. 11. Individual precursor's contribution to diamond growth.(44,45)

which diamond tends to adhere well (for example, Mo), or by using metallic
binder materials,(47,48) the adhesion problem may be avoided.
Although there is no clear-cut line between research and development,
this line is even more blurred in the case of TPCVD. Developments in this
area have been particularly visible in diamond film fabrication. Free-stand-
ing diamond wafers produced by TPCVD are already commercially avail-
able with diameters up to 10cm. Coating of complex three-dimensional
shapes is a difficult but not insurmountable problem and corresponding
efforts are already in progress.
As with most emerging technologies, economic considerations may be
the overriding concern for new developments. This has been the case, for
example, in the field of diamond film developments. The initially predicted
growth of the market in this field did not materialize.
As in plasma and wire arc spraying, complete control of the TPCVD
process is the ultimate goal of R&D efforts.

3. THERMAL PLASMA SYNTHESIS

Thermal plasmas which are a source of very high temperatures and
steep temperature gradients offer an attractive and chemically nonspecific
route for the synthesis of fine powders down to the nanometer size range.
Thermal Plasma Technology: Where Do We Stand and Where Are We Going? 17

The supersaturation of vapor species, which provides the driving force

for particle condensation, can be very large in thermal plasmas, leading to
the production of ultrafine particles by homogeneous nucleation. Ceramic
powders such as carbides, nitrides, oxides, and solid solutions have been
successfully synthesized in thermal plasma reactors. Thermal plasmas suit-
able for the synthesis of fine powders are primarily produced by means
of high-intensity arcs (ac or dc) and high-frequency discharges (RF and
microwave).
The high temperatures in thermal plasma reactors lead to short pro-
cessing times which translates into relatively small reactors with high
throughput. In spite of this and other advantages of thermal plasma reac-
tors, the high processing costs compared to competing processes must be
offset by superior properties of the products. In general, only high value-
added products are economically viable.
Recent advances in novel reactor and process designs have enhanced
the quality of powders produced in thermal plasmas. These novel designs
include the RF-dc hybrid reactor,(49) the reactive submerged arc (RSA),(50)
multiple plasma jets, and counterflow liquid-injection plasma reactors. They
aim at maximizing the heating, the mixing, and the residence time of mater-
ials in the plasma. Either the discharge itself or the plasma flame down-
stream of the discharge may be used for synthesizing the powders. In
thermal plasma synthesis, the reactants may be gases, liquids, or solids
before injection into the plasma. However the availability of gas-phase pre-
cursors for metals is severely limited. Therefore, the most commonly used
reactants for plasma synthesis have been solids. Several years ago, a liquid-
injection method was developed to overcome the problems associated with
solid injection and to capitalize on the benefits of gaseous reactants.(51)
In this case, a conventional dc plasma torch produces a high-tempera-
ture plasma jet, usually of an inert gas (Fig. 12). The atomized liquid-
precursor material is injected from the end opposing the plasma jet as a fine
mist in counterflow into the plasma, which gives rise to a recirculation vor-
tex. The recirculation provides extended dwell time of the precursor in a
relatively hot zone, assuring complete vaporization and dissociation of the
precursor materials. The rapid quench downstream of the hot zone induces
supersaturation of the dissociated precursor vapor, which in turn, leads to
the desired chemistry and to the formation of fine powder particles via
homogeneous nucleation. The powder is collected in the water-cooled col-
lection chamber and in a filter attached to the outlet of this chamber. The
same basic principle is used in other reactor configurations and with other
than liquid-precursor materials.
In the past, most of the oxide-synthesis work concentrated on binary
compounds. Large-scale production of TiO2 (pigment) has been pursued for
18 Pfender

Fig. 12. Schematic of the counterflow plasma reactor for injection of liquid precursors.

many years by Ti oxide in the United Kingdom using MW-sized arc gas
heaters.
With the advance of liquid-injection plasma synthesis within the last
eight years, ternary, quaternary, and higher component oxide solid solutions
including spinels of aluminates, ferrites, and chromites,(51-54) and the high-
temperature oxide superconductors,(51,55-58) have been synthesized in both
dc and RF plasmas. Figure 13 shows an example of a RF plasma reactor
which has been used for the synthesis of superconducting powders.(55)
Besides these more "exotic" oxide compounds, there has been little activity
in thermal-plasma synthesis of oxides.
Thermal Plasma Technology: Where Do We Stand and Where Are We Going? 19

Fig. 13. Schematic of a RF plasma reactor for plasma synthesis of fine powders.

In contrast to the synthesis of oxide powders, there are strong and

continuing efforts in thermal-plasma synthesis of nonoxide ceramic pow-
ders. The strongest interest seems to be in nitrides, followed by carbides and
finally borides. The most common reactants for thermal-plasma synthesis
of nonoxide ceramics are solids and metal halides.(59)
Recently an experiment has been described in which a plasma contain-
ing vapor-phase precursors was expanded through a nozzle with a hot cer-
amic wall,(60) as schematically shown in Fig. 14. This arrangement
approaches a configuration of one-dimensional flow with one-dimensional
temperature gradients in the direction of the flow in the nozzle, leading to
high uniformity of the quench rate. Furthermore, a nozzle provides much
higher quench rates than would be obtainable otherwise. Using this arrange-
ment, ultrafine SiC particles have been synthesized with SiCl4 and CH4 as
20 Pfender

Fig. 14. Schematic of plasma reactor for the synthesis on nanometer size particles.(60)

reactants. These results show that the particles produced in the process have
a narrow size distribution peaking at a few nanometers. Both SiC and Si
particles have been produced with median diameters of 8.5 nm and with
90% of the particles having diameters of less than 16 nm.

3.1. Summary of Present R&D Efforts in Thermal Plasma Synthesis

Present research efforts concentrate on non-oxide powders (nitrides,
carbides) and on new approaches for the synthesis of nanometer size par-
ticles. The future growth of this technology will hinge to a large extent on
the development and implementation of effective controls.
This includes control of the particle size and size distribution as well as
of the morphology of the produced powders. Novel reactor designs are of
great interest as, for example, multiple torch arrangements, hybrid reactors
combining dc torches with RF torches, and other innovative designs which
should not only improve the quality of the product but also enhance the
efficiency of the process.
Thermal Plasma Technology: Where Do We Stand and Where Are We Going? 21

Recent developments are primarily concerned with the scale-up of

plasma reactors and with automatic control of the process.

4. THERMAL PLASMA WASTE DESTRUCTION

Thermal plasma reactors offer unique advantages for the destruc-
tion of hazardous wastes: (1) the high energy density and temperatures
associated with thermal plasmas and the corresponding fast reaction times
offer the potential of large throughputs in a small reactor, (2) the high tem-
peratures can also be used to obtain very high quench rates allowing the
attainment of metastable states and nonequilibrium compositions, (3) the
high heat fluxes at the reactor boundaries lead to fast attainment of steady
state conditions allowing rapid start-up and shutdown times compared to
other thermal treatments such as incinerators, (4) use of electric energy
reduces gas flow needs and off-gas treatment requirements, and offers con-
trol over the chemistry, including the possibility of generating salable co-
products, and (5) all the characteristics combined allow easy integration into
a manufacturing process which generates hazardous wastes, thus permitting
the destruction of the wastes at the source. The major disadvantage of the
plasma process lies in the use of electricity as energy source, which influences
the process economics unfavorably. A further consideration is that plasma
processes have more process parameters to control compared to traditional
processes and require, therefore, a higher degree of automation in the
process control. And there seems still to be a lack of a solid engineering
base for certain types of large-scale installations.
There have been a number of plasma waste destruction process devel-
opments, and these developments can be classified as follows: (1) pyrolysis
of liquid hazardous wastes using reactors with nontransferred plasma
torches, (2) treatment of solid materials or slurries usually in a combination
of pyrolysis and compaction of residues using transferred arc plasma reac-
tors, and (3) the reclamation of waste products from manufacturing pro-
cesses such as baghouse dust from electric arc furnaces, where both non-
transferred and transferred arc reactors are being used. A special case is the
waste destruction process in which the process economy is improved by the
generation of a salable co-product. An example for this last category is a
process in which hazardous liquid chemicals are destroyed while diamonds
are deposited on a substrate in the reaction zone.
Figure 15 shows a schematic of a plasma pyrolysis reactor system devel-
oped by Westinghouse. This system resulted from a combination of initial
developments by Barton(61,62) with Westinghouse industrial plasma technol-
ogy. The reactor consists of a 1-MW nontransferred plasma torch, a refrac-
tory-lined hot reaction channel which exhausts the reaction products into a
22 Pfender

Fig. 15. Pyroplasma pyrolysis unit developed by Westinghouse.(63)

larger holding tank. From here they are vented to a scrubber. The West-
inghouse torch consists of two cylindrical water-cooled electrodes having
the same diameter, and the plasma gas (air) is injected through a narrow
gap between them. A magnetic field rotates the arc-electrode attachment,
reducing electrode wear. The liquid waste materials are injected into the jet
immediately downstream of the torch. Only liquid waste materials are
injected into the jet immediately downstream of the torch. Only liquid
wastes can be destroyed. A computer program has been developed based
on equilibrium chemistry which determines the optimal set of operating par-
ameters for any specific waste composition such as torch power, torch gas
flow rate, liquid waste flow rate, and possible need for additional reactants
to make the process chemistry more favorable. The process has been
operating with various wastes, and for PCB's being destroyed at a rate of
up to 12 liters/min with a torch power of 850 kW, destruction efficiencies
of eight-nines have been attained, with particulate and acid emission well
under the EPA guidelines.(63) The entire processing unit is self-contained
and mounted on a trailer and can thus be moved easily from one waste site
to the other. Hook-ups are required for power, water, and for the waste
water. The quality of the gaseous emissions and of the waste water is con-
stantly monitored.
Thermal Plasma Technology: Where Do We Stand and Where Are We Going? 23

Fig. 16, Schematic of the Retch plasma centrifugal furnace."

The treatment of contaminated solids with transferred arc plasma reac-

tors has probably seen the highest level of development activities. Figure 16
shows, as an example, a schematic of a commercial plasma reactor operated
by MGC Plasma in Muttenz, Switzerland. This reactor, based on a rotating
plasma furnace, has been developed by Retech.(64-68)
The reactor consists of a cylindrical tub of 2.4m diameter with a
1.2 MW plasma torch mounted above it operating with oxygen-enriched air
(see Fig. 16). The plasma torch is mounted in a way that allows the move-
ment of the tip which can thus direct the plasma flow toward any part of
the tub below it. The arc is struck between the torch and the tub, being
sustained at a typical length of 0.5 m with a voltage of 600 V at 2000 A. The
waste products are fed into the tub from an automated feeder. The system
is designed to accept complete drums with contaminated waste, which are
cut remotely in a space above the reactor before being fed into the plasma
unit. The heat from the plasma volatilizes the organic compounds and leads
to melting of the inorganic materials. Rotation of the tub at 50-70 rpm
allows the liquefied waste material to remain in the well while the gaseous
emissions escape through a hole in the center into an afterburner chamber
where air or oxygen is added and where an additional combustion process
assures the complete destruction of the toxic volatile components. A smaller
second torch (250 kW) can be used to avoid obstruction of the hole by
solidifying slag, if necessary. After the tub is filled with molten material, the
24 Pfender

rotation of the tub is reduced and the molten material is allowed to flow
through the center hole into a crucible where it forms a nonleachable slag.
The off-gas has to be treated with a scrubber, and additional treatment is
required for reduction of NO. Heat in the off-gas is recovered before it
enters the scrubber. The off-gas composition is continuously monitored.
Destruction efficiencies of better than 99.9968 have been achieved for
organic contaminants, and the slag has been demonstrated to be nonleach-
able with respect to heavy metals.
The Retech plasma centrifugal furnace is also the heart of a system
which has been developed at the DOE site at Butte, Montana, for the
destruction and vitrification of mixed wastes (nuclear and hazardous
organics).(69,70) The reactor is a slightly smaller version of the reactor used
in the installation in Switzerland, with a tub width of 1.8m and operating
at 600 kW.
The independent control over heat input and chemistry in a plasma
chemical reactor offers the opportunity of tailoring processes such that valu-
able co-products can be produced. Although no such process is commercial
yet, there are several in the laboratory or in advanced development stages.
An example is the generation of industrial diamonds during the destruction
of hazardous liquid waste materials, pursued at the University of
Minnesota.(71,72) A plasma reactor similar to that shown in Fig. 12 has been
used for these experiments. Destruction and removal efficiencies of over
99.99% have been achieved with benzene and CCl4 while simultaneously
diamonds were grown at linear growth rates of more than 0.1 mm/h. As an
example, Fig. 17 shows a micrograph of diamond synthesized from PCB's.

4.1. Present R&D Efforts in Thermal Plasma Waste Destruction

In comparison to other competing waste destruction processes, thermal
plasma waste destruction is inherently more expensive. Therefore, present
research activities in this area have been restricted to a few specific types
of toxic waste destruction problems where thermal plasmas offer unique
advantages. This fact is very obvious in the commercial field. Although the
potential advantages of plasma processes continue to spawn new develop-
ments, the economic issues are usually not as clear-cut.
The economic advantages of incineration technologies drive the use of
plasma technologies to the treatment of "problem wastes," i.e., of wastes
which pose difficulties to incinerators. Examples are wastes containing con-
taminated ceramic materials or refractory metals, or wastes where volatile
toxic elements require treatment of the off-gas. The higher temperatures of
the plasma process and the lower gas flow rates are of advantage in these
cases. An area which has still a considerable potential and which has seen
Thermal Plasma Technology: Where Do We Stand and Where Are We Going? 25

Fig. 17. Micrograph of diamond film deposited on Mo substrate using PCB's as precursor.

only a limited number of investigations is the use of plasma chemistry to

generate valuable co-products.

5. THERMAL PLASMA DENSIFICATION

This technology refers primarily to the densification of powders,
although densification by plasma sintering has also been reported. Powder
densification involves in-flight melting of the material in particulate form,
followed by gradual cooling and freezing before being collected at the bot-
tom of the densification chamber or in an appropriate collection device,
depending on the particle density and its size range. Using either direct
current (dc) or inductively coupled RF plasma torches, the process has been
successfully used for the densification and spheroidization of a large number
of materials, ranging from low-melting-point metals such as copper and
nickel to oxide ceramics such as alumina and yttria-stabilized zirconia, and
refractory metals such as molybdenum and tungsten.
Typical electron micrographs showing the morphology of the individ-
ual particles before and after densification in an inductively coupled RF
26 Pfender

Fig. 18. Micrograph of original (left side) and plasma densifled (right side) powders. From top
to bottom: A12O3, Mo, W.

plasma are shown in Fig. 18 for alumina, molybdenum, and tungsten pow-
ders. These have mostly a mean particle diameter in the range of 45-76 jUm.
Plasma sintering refers to the process of bonding particles and
densification of materials by applying heat from a plasma. Plasma sintering
of refractory materials, mainly ceramics, is a relatively new technology. At
the end of the 1960s Bennett et al.,(73,74) sintered some oxides in microwave
plasmas at 2450 MHz and gas pressures between 100-7000 Pa. They found
that alumina and some other ceramic materials can be sintered successfully
and rapidly in plasmas generated by microwave excitation. They observed
significantly finer grain sizes and greater densification rates than in conven-
tionally sintered materials at the same temperature. They also hypothesized
Thermal Plasma Technology: Where Do We Stand and Where Are We Going? 27

that the plasma might be enhancing sintering by cleaning the particle

surfaces.
The advantage of plasma sintering over conventional sintering is the
extremely high heat transfer rate to the sample, which may result in rapid
sintering (minutes) with minimal grain growth.(75-80) The latter is of particu-
lar importance for sintering of nanometer size particles. Unfortunately, at
this time plasma sintering is restricted to small-diameter samples with
rotational geometry. Another disadvantage of this process are the relatively
high energy losses. Both experimental and modeling work indicate that
recombination of charged species on the sample surface may be, under
certain conditions, the primary heat transfer mechanism.(80,81)

5.1. Recent R&D Efforts in Thermal Plasma Densification

There is relatively little research activity in plasma spheroidization and
densification, because powder densification is one of the simplest but most
successful applications of plasma technology in materials processing.
Present known research activities seem to concentrate on spheroidization
and densification of ceramic powders in RF plasmas(82)
Similar arguments hold for plasma sintering which has seen little
research activities over the past years. Very little is known about new devel-
opments in these areas.

6. CONCLUSIONS
The following conclusions are restricted to the areas which have been
covered by this overview, i.e., (1) thermal plasma coating technologies, (2)
thermal plasma synthesis of fine powders, (3) thermal plasma waste destruc-
tion, and (4) thermal plasma densification and spheroidization.
In general, the growth of thermal plasma technologies has been ham-
pered by the lack of controls. It is felt that synergistics between basic
research and development will be crucial for the implementation of controls.
This implementation will have a strong impact on the growth of thermal
plasma coating technologies.
Developments in plasma synthesis will also benefit from better controls,
and it is anticipated that the focus of research in this area will be increas-
ingly toward nanometer-structured materials.
Thermal plasma waste destruction is facing stiff competition from less
costly processes. An economic assessment, however, must include the entire
system, not only the plasma part. The possibility for interesting co-products
and the potential for integration of the waste destruction system in the over-
all manufacturing process may improve the economics.
28 Pfender

Thermal plasma densification and spheroidization of powders will

remain a viable technology.

ACKNOWLEDGMENTS
A number of colleagues and present and former graduate students con-
tributed to the material presented in this overview. Their contributions, as
well as the support from NSF and DOE, are gratefully acknowledged.

REFERENCES
1. R. Spores and E. Pfender, "Flow structure of a turbulent thermal plasma jet," Surf. Coat.
Technol. 37, 251 (1989).
2. R. Spores, "Analysis of the flow structure of a turbulent thermal plasma jet," Ph.D. Thesis,
University of Minnesota-Minneapolis (1989).
3. E. Pfender, J. Fincke, and R. Spores, "Entrainment of cold gas into thermal plasma jets,"
Plasma Chem. Plasma Process. 11, 529 (1991).
4. R. W. Schefer, V. Hartmann, and R. W. Dibble, AIAA J. 25, 1318 (1987).
5. M. Brossa and E. Pfender, "Probe measurements in thermal plasma jets," Plasma Chem.
Plasma Process. 8, 75 (1988).
6. S. A. Wutzke, "Conditions governing the symptomatic behavior of an electric arc in a
superimposed flow field," Ph.D. Thesis, University of Minnesota-Minneapolis (1967).
7. S. A. Wutzke, E. Pfender, and E. R. G. Eckert, "Study of electric-arc behavior with super-
imposed flow," AIAA J. 5, 707 (1967).
8. M. P. Planche, Z. Duan, O. Lagnoux, J. Heberlein, P. Fauchais, and E. Pfender, "Study
of arc fluctuations with different plasma spray torch configurations", Proceedings of the
13th International Symposium on Plasma Chemistry, edited by C. K. Wu (Beijing, China),
Vol. III, pp. 1460-1465 (1997).
9. M. P. Collares and E. Pfender, "Magnetic probe measurements in plasma spray torches,"
Proceedings of the 13th International Symposium on Plasma Chemistry, edited by C. K.
Wu (Beijing, China), Vol. III, pp. 1466-1470 (1997).
10. S. Paik, P. C. Huang, J. Heberlein, and E. Pfender, "Determination of the arc-root position
in a dc plasma torch," Plasma Chem. Plasma Process. 13, 379-397 (1993).
11. W. Finkelnburg and H. Maecker, "Electric arcs and thermal plasmas," Encyclopedia of
Physics, Vol. XXII, Springer, Berlin (1956).
12. Th. Peters, Z. Phys. 144, 612 (1956).
13. S. Malmberg, "Analysis of the plasma jet structure, particle motion, and coating quality
during dc plasma spraying," Ph.D. Thesis, University of Minnesota-Minneapolis (1994).
14. P. Fauchais, J. F. Coudert, and M. Vardelle, "Diagnostics in thermal plasma processing,"
Plasma Diagnostics, Vol. I (O. Anciello and D. L. Flamm, eds.), Academic Press, New
York (1989).
15. D. B. Spalding, "Two-fluid models of turbulence," CFDU Report CFD/85/4, CFDU,
Imperial College, London (1985).
16. D. B. Spalding, "A turbulence model for buoyant and combusting flows," CFDU Report
CFD/86/4, CFDU, Imperial College, London (1986).
17. D. B. Spalding, Int. J. Physiochem. Hydrodynam. 4, 323 (1983).
18. P. C. Huang, J. Heberlein, and E. Pfender, "A two-fluid model of turbulence for a thermal
plasma jet," Plasma Chem. Plasma Process. 15, 25-46 (1995).
Thermal Plasma Technology: Where Do We Stand and Where Are We Going? 29

19. P. C. Huang, "A turbulent swirling arc model and a two-fluid turbulence model for thermal
plasma sprays," Ph.D. Thesis, University of Minnesota-Minneapolis (1993).
20. D. Apelian, D. Wei, and M. Paliwal, Thin Solid Films 118, 395 (1984).
21. Y. P. Chyou and E. Pfender, "Behavior of particulates in thermal plasma flows," Plasma
Chem. Plasma Process. 9, 45-71 (1989).
22. C. H. Chang, "Numerical simulation of alumina spraying in argon-helium plasma jet,"
Proceedings of the International Thermal Spray Conference (Orlando, Florida), (June
1992), p. 793.
23. J. H. Park, Z. Duan, J. Heberlein, E. Pfender, Y. C. Lau, and H. P. Wang, "Modeling
of fluctuations experienced in N2 and N 2 /H2 plasma jets issuing into atmospheric air,"
Proceedings of the 13th International Symposium on Plasma Chemistry, edited by C. K.
Wu (Beijing, China), Vol. I, pp. 326-331 (1997).
24. P. Fauchais, A. C. Leger, M. Vardelle, and A. Vardelle, "Formation of plasma-sprayed
oxide coatings," Proceedings of the Julian Szekely Memorial Symposium on Materials
Processing and the TMS Fall Extraction & Processing Conference, edited by H. Y. Sohn,
J. W. Evans, and D. Apelian (Cambridge, Massachusetts), (October, 1997), pp. 571-592.
25. H.-D. Steffens, Z. Babiak, and M. Wewel, IEEE Trans. Plasma Sci. 18, 974 (1989).
26. M. L. Thorpe, Adv. Mater. Process. 134, 69 (1988).
27. D. R. Marantz and D. R. Marantz, Proceedings of the 3rd National Thermal Spray
Conference, p. 113(1990).
28. E. Sampson and L. Neider, Proceedings of the 5th National Thermal Spray Conference,
p. 271 (1993).
29. E. R. Sampson, Proceedings of the 5th National Thermal Spray Conference, p. 257 (1993).
30. X. Wang, J. Heberlein, E. Pfender, and W. Gerberich, "Effect of gas velocity and particle
velocity on coating adhesion in wire arc spraying," Proceedings of the 9th National Ther-
mal Spray Conference, Thermal Spray: Practical Solutions for Engineering Problems, edited
by C. C. Berndt, published by ASM International, (1996), pp. 807-811.
3 1 . X . Wang, J. Heberlein, E. Pfender, and W. Gerberich, "Effect of shrouded CO2 gas atomiz-
ation on coating properties in wire arc spray," Proceedings of the 8th National Thermal
Spray Conference (Houston, Texas), (1995), pp. 31-37.
32. R. F. Heile and D. C. Hill, Welding J. p. 201-205 (1975).
33. P. J. Wewitt and A. A. Hirst, Ann. Occup. Hyg. 37, 297 (1993).
34. M. Ushio, K. Nakata, M. Tanaka, H. Tong, and T. Mita, Trans. JWRI 23, 21 (1994).
35. T. Watanabe, X. Wang, J. Heberlein, and E. Pfender, "Fume generation mechanism in
wire arc spraying," ISPC-12 (Minneapolis, Minnesota, 1995), Vol. II, pp. 889-894.
36. T. Watanabe, X. Wang, J. Heberlein, E. Pfender, and W. Herwig, "Voltage and current
fluctuations in wire arc spraying as indications for coating properties," Proceedings of the
9th National Thermal Spray Conference, Thermal Spray: Practical Solutions for Engineer-
ing Problems, edited by C. C. Berndt, published by ASM International, (1996), pp. 577-
583.
37. E. R. G. Eckert and E. Pfender, "Advances in plasma heat transfer," Adv. Heat Transfer
4,229-313(1967).
38. Z. P. Lu, J. Heberlein, and E. Pfender, "Process study of thermal plasma chemical vapor
deposition of diamond, Part I: Substrate material, temperature, and methane concen-
tration," Plasma Chem. Plasma Proc. 12, 35-52 (1992).
39. Z. P. Lu, J. Heberlein, and E. Pfender, "Process study of thermal plasma chemical vapor
deposition of diamond, Part II: Pressure dependence and effect of substrate pretreatment,"
Plasma Chem. Plasma Proc. 12, 55-69 (1992).
40. K. A. Snail, C. M. Marks, Z. P. Lu, J. Heberlein, and E. Pfender, "High temperature,
high rate homoepitaxial synthesis of diamond in a thermal plasma reactor," Mater. Lett.
12, 301-305 (1991).
30 Pfender

41. Z. P. Lu, K. Snail, C. Marks, J. Heberlein, and E. Pfender, "High rate homoepitaxial
growth of diamond in thermal plasma," Proc. 2nd Int. Symp. Diamond Mater. (The Elec-
trochemical Society, Washington, DC, 1991), Vol. 91-8, pp. 99-106.
42. P. Greuel, H. Yoon, D. Ernie, and J. Roberts, Mater. Res. Soc. Symp. Proc. 334, 141
(1994).
43. P. Greuel, J. Roberts, and D. Ernie, Proceedings of the 12th International Symposium on
Plasma Chemistry, edited by J. Heberlein, D. Ernie, and J. Roberts (Minneapolis, Minne-
sota), Vol. IV, p. 2209 (1995).
44. B. W. Yu, "A model for chemical vapor deposition of diamond in a radio-frequency induc-
tion thermal plasma," Ph.D. Thesis, University of Minnesota-Minneapolis (1994).
45. B. W. Yu and S. L. Girshick, "Atomic carbon vapor as a diamond growth precursor in
thermal plasmas," J. Appl. Phys. 75, 3914-3923 (1994).
46. M. Asmann, C. F. M. Borges, J. Heberlein, and E. Pfender, "Thermal plasma chemical
vapor deposition of diamond on steel," Proceedings of the 13th International Symposium
on Plasma Chemistry, edited by C. K. Wu (Beijing, China), Vol. III, pp. 1206-1211 (1997).
47. C. Tsai, J. Nelson, W. Gerberich, J. Heberlein, and E. Pfender, "Metal reinforced thermal
plasma diamond coatings," J. Mater. Res. 7, 1967-1969 (1992).
48. C. Tsai, J. C. Nelson, W. W. Gerberich, J. Heberlein, and E. Pfender, "Diamond-metal
composite coatings on cemented carbide tools," Diamond Relat. Mater. 2, 617-620 (1993).
49. T. Yoshida, T. Tani, H. Nishimura, and K. Akashi, J. Appl. Phys. 2, 640 (1983).
50. A. Kumar and R. Roy, J. Mater. Res. 3, 1373 (1989).
51. P. Kong and E. Pfender, Proceedings of the 2nd Int. Conf. Ceram. Powder Processing Sci.
(Berchtesgaden, 1988).
52. P. Kong and E. Pfender, Combustion and Plasma Synthesis of High Temperature Materials
(Z. A. Munir and J. B. Holt, eds.), VCH Publishers, New York (1990), p. 420.
53. P. Kong, T. Or, L. Stachowicz, and E. Pfender, Better Ceramics through Chemistry IV,
Mater. Res. Soc. Symp. Proc., edited by B. J. J. Zelinski, C. J. Brinker, D. E. Clark, and
D. R. Ulrich (Pittsburgh, Pennsylvania), 180, 849 (1990).
54. T. Or, Z. Lu, L. Stachowicz, P. Kong, and E. Pfender, Plasma Processing and Synthesis
of Materials III, Mater. Res. Soc. Symp. Proc., edited by D. Apelian and J. Szekely (Pitts-
burgh, Pennsylvania), 190, 83 (1991).
55. H. Zhu, Y. C. Lau, and E. Pfender, "RF plasma synthesis of YBa2Cu3O7-x powders," J.
Supercond. 3, 171-175 (1990).
56. K. Terashima, T. Yoshida, and K. Akashi, Jpn. Symp. on Plasma Chemistry, p. 157 (1988).
57. T. Ono, M. Kagawa, Y. Syono, M. Ikebe, and Y. Muto, Plasma Chem. Plasma Process.
7, 201 (1987).
58. H. Zhu, Y. C. Lau, and E. Pfender, "Deposition of YBa2Cu3O7-x thick films by the spray-
ICP technique," Proc. 9th Int. Symp. Plasma Chem. (Pugnochiuso, Italy, 1989), Vol. 3,
pp. 1497.
59. P. C. Kong and E. Pfender, "Chapter 14: Plasma Processes," Carbide, Nitride and Boride
Materials Synthesis and Processing (A. W. Weimer, ed.), Chapman & Hall, London (1997).
60. N. Rao, S. Girshick, J. Heberlein, P. McMurry, S. Tones, D. Hausen, and B. Micheel,
Plasma Chem. Plasma Process. 15, 581 (1995).
61. T. G. Barton, "Problem waste disposal by plasma heating," Int. Recycling Congress
(Berlin), Vol. I, pp. 733-736 (1979).
62. J. V. R. Heberlein, "New developments in non-transferred plasma torch technology," Pro-
ceedings of the 2nd Japanese Symposium on Plasma Chemistry, (1989), p. 131.
63. J. V. R. Heberlein, W. J. Melilli, S. V. Dighe, and W. H. Reed, "Adaptation of non-
transferred plasma torches to new applications of plasma systems," Proceedings of the
Thermal Plasma Technology: Where Do We Stand and Where Are We Going? 31

Workshop on Industrial Plasma Applications, edited by M. I. Boulos (Pugnochiuso, Italy),

(1989), p. 1.
64. M. R. Funfschilling, W. Bernhard, R. C. Eschenbach, "Test results with the plasma cen-
trifugal furnace at Muttenz, Switzerland," Proceedings of the 1991 Incineration Conference
(Knoxville, Tennessee).
65. R. C. Eschenbach, "Use of plasma torches for melting special metals and for destroying
and stabilizing hazardous wastes," Proceedings of the Workshop on Industrial Plasma
Applications, edited by M. Boulos (Pugnochiuso, Italy), (1989), pp. 127-136.
66. R. C. Eschenbach, "Plasma centrifugal furnace for destroying hazardous wastes," Proceed-
ings of the 1st International EPRI Symposium (1990).
67. M. P. Schlienger, "Apparatus and method for high temperature disposal of hazardous
waste materials," U.S. Patent 4770109 (1988).
68. M. R. Funfschilling and R. C. Eschenbach, "A plasma centrifugal furnace for treating
hazardous waste, Muttenz, Switzerland," Proceedings of the Xllth Congress International
Union for Electroheat (Montreal, Canada) (1992).
69. J. W. Sears, R. C. Eschenbach, and R. A. Hill, "The plasma centrifugal furnace: A method
for stabilization and decomposition of toxic and radioactive wastes," Waste Management
10, 165 (1990).
70. R. Haun, R. Eschenbach, D. Battleson, C. Alsberg, and T. Jackson, "Site test results with
the PCF-6", Proceedings of the 1992 Incineration Conference.
71. E. Pfender, Q. Y. Han, T. W. Or, Z. P. Lu, and J. Heberlein, "Rapid synthesis of diamond
by counter-flow liquid injection into an atmospheric pressure plasma jet," Diamond Relat.
Mater. 1, 127-133 (1992).
72. Q. Y. Han, T. W. Or, Z. P. Lu, J. Heberlein, and E. Pfender, "dc thermal plasma depo-
sition of diamond films," Thermal Plasma Applications in Materials and Metallurgical Pro-
cessing, (N. El-Kaddah, ed.), TMS, Warrendale, Pennsylvania (1992), pp. 277-291.
73. C. E. G. Bennett, N. A. Kinnon, and L. S. Williams; "Sintering in gas discharges," Nature
(London) 217, 1287 (1968).
74. C. E. G. Bennett and N. A. Kinnon, "Glow discharge sintering of alumina," Kinetics of
Reactions in Ionic Systems (T. J. Gray and V. D. Frechette, eds.), Plenum Press, New York
(1969), p. 408.
75. D. L. Johnson and R. R. Rizzo, "Plasma sintering of a-alumina," Am. Ceram. Soc. Bull.
59, 467 (1980).
76. J. S. Kim and D. L. Johnson, "Plasma sintering of alumina," Am. Ceram. Soc. Bull. 62,
620 (1980).
77. D. L. Johnson, W. B. Sanderson, J. M. Knowlton, and E. L. Kemer, "Sintering of a-Al2O3
in gas plasmas," Adv. Ceram. 10, 656 (1985).
78. P. C. Kong, Y. C. Lau, and E. Pfender, "The effects of gas composition and gas pressure
on RF sintering of MgO," Proceedings of the MRS 1987 Spring Meeting (Anaheim,
California), 98, pp. 371-375 (1987).
79. E. Pfender and Y. C. Lee, "Heat transfer analysis of the plasma sintering process,"
Proceedings of the Materials Research Society (Elsevier Science Publishing Co., Inc.), 30,
p. 141 (1984).
80. N. P. Tandian, "Heat transfer in RF plasma sintering: A modeling and experimental
study," Ph.D. Thesis, University of Minnesota-Minneapolis (1994).
81. N. P. Tandian and E. Pfender, "Studies of the high frequency (RF) plasma sintering
process," Proceedings of the 13th International Symposium on Plasma Chemistry, edited
by C. K. Wu (Beijing, China), Vol. IV, pp. 1630-1635 (1997).
82. N. M. Dignard and M. Boulos, "Ceramic powder spheroidization under induction plasma
conditions," Proceedings of the 13th International Symposium on Plasma Chemistry,
edited by C. K. Wu (Beijing, China), Vol. III, p. 1031 (1997).