Sintered silicon carbide: A new ceramic vessel material for microwave chemistry in single-mode reactors

Article abstract of DOI:10.1002/chem.201001703

Silicon carbide (SiC) is a strongly microwave absorbing chemically inert ceramic material that can be utilized at extremely high temperatures due to its high melting point and very low thermal expansion coefficient. Microwave irradiation induces a flow of electrons in the semiconducting ceramic that heats the material very efficiently through resistance heating mechanisms. The use of SiC carbide reaction vessels in combination with a single-mode microwave reactor provides an almost complete shielding of the contents inside from the electromagnetic field. Therefore, such experiments do not involve electromagnetic field effects on the chemistry, since the semiconducting ceramic vial effectively prevents microwave irradiation from penetrating the reaction mixture. The involvement of electromagnetic field effects (specific/nonthermal microwave effects) on 21 selected chemical transformations was evaluated by comparing the results obtained in microwave-transparent Pyrex vials with experiments performed in SiC vials at the same reaction temperature. For most of the 21 reactions, the outcome in terms of conversion/purity/product yields using the two different vial types was virtually identical, indicating that the electromagnetic field had no direct influence on the reaction pathway. Due to the high chemical resistance of SiC, reactions involving corrosive reagents can be performed without degradation of the vessel material. Examples include high-temperature fluorine-chlorine exchange reactions using triethylamine trihydrofluoride, and the hydrolysis of nitriles with aqueous potassium hydroxide. The unique combination of high microwave absorptivity, thermal conductivity, and effusivity on the one hand, and excellent temperature, pressure and corrosion resistance on the other hand, makes this material ideal for the fabrication of reaction vessels for use in microwave reactors. Simulating conductive heat transfer in a microwave: Using reaction vials made out of strongly microwave-absorbing silicon carbide (SiC) in a microwave reactor simulates a conductively heated autoclave experiment due to efficient shielding of the electromagnetic field by the SiC vial. Advantages of SiC vials for microwave processing include their excellent corrosion resistance, thermal stability, and high thermal effusivity and conductivity.

Full text of DOI:10.1002/chem.201001703

DOI: 10.1002/chem.201001703
Sintered Silicon Carbide: A New Ceramic Vessel Material for Microwave
Chemistry in Single-Mode Reactors
Bernhard Gutmann, David Obermayer, Benedikt Reichart, Bojana Prekodravac,
Muhammad Irfan, Jennifer M. Kremsner, and C. Oliver Kappe*[a]
Abstract: Silicon carbide (SiC) is a
strongly microwave absorbing chemi-
cally inert ceramic material that can be
utilized at extremely high temperatures
due to its high melting point and very
low thermal expansion coefficient. Mi-
crowave irradiation induces a flow of
electrons in the semiconducting ceram-
ic that heats the material very efficient-
ly through resistance heating mecha-
nisms. The use of SiC carbide reaction
vessels in combination with a single-
mode microwave reactor provides an
almost complete shielding of the con-
tents inside from the electromagnetic
field. Therefore, such experiments do
not involve electromagnetic field ef-
fects on the chemistry, since the semi-
conducting ceramic vial effectively pre-
vents microwave irradiation from pene-
trating the reaction mixture. The in-
volvement of electromagnetic field ef-
fects (specific/nonthermal microwave
effects) on 21 selected chemical trans-
formations was evaluated by compar-
ing the results obtained in microwave-
transparent Pyrex vials with experi-
ments performed in SiC vials at the
same reaction temperature. For most
of the 21 reactions, the outcome in
terms of conversion/purity/product
yields using the two different vial types
was virtually identical, indicating that
the electromagnetic field had no direct
influence on the reaction pathway. Due
to the high chemical resistance of SiC,
reactions involving corrosive reagents
can be performed without degradation
of the vessel material. Examples in-
clude high-temperature fluorine–chlor-
ine exchange reactions using triethyla-
mine trihydrofluoride, and the hydroly-
sis of nitriles with aqueous potassium
hydroxide. The unique combination of
high microwave absorptivity, thermal
conductivity, and effusivity on the one
hand, and excellent temperature, pres-
sure and corrosion resistance on the
other hand, makes this material ideal
for the fabrication of reaction vessels
for use in microwave reactors.
Keywords: ceramics
·
dielectric
properties · microwave chemistry ·
silicon carbide · thermal effusivity
Introduction
crowave irradiation to be able to penetrate to the reaction
mixture, reaction vessels employed in microwave chemistry
are typically made out of low microwave absorbing or mi-
crowave transparent materials, such as borosilicate glass
(Pyrex), quartz, or suitable polymers like PTFE (Teflon).[2]
These materials exhibit many distinct and valuable advan-
tages when used in microwave chemistry, but also face some
limitations in high-end applications under extreme reaction
conditions as far as temperature/pressure resistance, or
chemical stability to aggressive media is concerned. Prob-
lems also exist when attempting to heat microwave transpar-
ent or low-absorbing reaction media in these types of reac-
tion vessels.
Microwave chemistry generally relies on the ability of the
reaction mixture to efficiently absorb microwave energy,
taking advantage of microwave dielectric heating phenom-
ena such as dipolar polarization or ionic conduction mecha-
nisms.[1] This produces efficient and rapid internal heating
(in-core volumetric heating) by direct interaction of electro-
magnetic irradiation with the molecules (solvents, reagents,
catalysts) that are present in the reaction mixture.[1] For mi-
[a] B. Gutmann, D. Obermayer, B. Reichart, B. Prekodravac, M. Irfan,
Dr. J. M. Kremsner, Prof. Dr. C. O. Kappe
Christian Doppler Laboratory for Microwave Chemistry (CDLMC)
and Institute of Chemistry, Karl-Franzens-University Graz
Heinrichstrasse 28, 8010 Graz (Austria)
Fax : (+43)0316-380-9840
E-mail: oliver.kappe@uni-graz.at
In recent years, the use of sintered silicon carbide (SiC)
has become increasingly popular for a variety of applica-
tions in microwave chemistry. SiC is a strongly microwave
absorbing chemically inert ceramic material that can be uti-
lized at extremely high temperatures due to its high melting
point (ꢀ27008C) and very low thermal expansion coeffi-
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201001703.
12182
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 12182 – 12194

FULL PAPER
cient.[3] Microwave irradiation induces a flow of electrons in
the semiconducting SiC that heats the material very effi-
ciently through resistance (ohmic) heating mechanisms.[3]
A
variety of SiC materials in the form of powders, granules,
and vessels (crucibles) have been used for some time in
high-temperature microwave processing applications taking
advantage of the extremely good microwave absorption
properties, chemical resistance, and high thermal conductivi-
ty.[4–9] Several studies have described the use of SiC as so-
called microwave susceptor or sensitizer in material scien-
ces,[4,5] aiding for example in the rapid microwave-assisted
synthesis of superconducting MgB2 at temperatures close to
8008C,[6] or in the microwave sintering of ZrO2 ceramics.[7]
In fact, because of its high melting point and chemical resist-
ance, SiC crucibles have been used for the microwave melt-
ing of aluminum, copper, and nuclear waste glass.[8] SiC has
also been exploited as a support material for metal catalysts
in the microwave-assisted thermochemical (800–10008C)
conversion of glycerol to syngas.[9]
More related to the field of synthetic chemistry, our group
has recently developed so-called passive heating elements
(PHEs) made out of sintered SiC that aid in the microwave
heating of weakly absorbing or transparent reaction mix-
tures.[10] The cylindrical shape of the PHEs allows the use of
these materials in reaction vials designed for both single-
mode[11,12] and multimode microwave reactors.[13] In addi-
tion, microtiter plates made out of SiC for parallel micro-
wave chemistry applications in multimode instruments were
introduced in the past few years.[14,15] For these applications,
the high thermal conductivity of SiC is of particular impor-
tance, avoiding the formation of temperature gradients
within the plate.[15]
Figure 1. Reaction vial made out of sintered silicon carbide (SiC) for per-
forming microwave chemistry under single-mode conditions. For compar-
ison purposes a standard 10 mL Pyrex vial with a snap cap, internal FO
probe sensor and a magnetic stir bar is also shown (see Experimental
Section for details).
this procedure, the high strength of the obtained SiC re-
mains almost unchanged up to about 16008C (ROCARꢁ S,
Table 1). As a suitable microwave platform we have selected
Table 1. Comparison of material data for sintered silicon carbide
(ROCARꢁ S) and borosilicate (Pyrex) glass.[a]
ROCARꢁ S Pyrex
melting point
m.p. [8C]
ꢀ2700
115
ꢀ800
1.2
thermal conductivity
l [W mÀ1 KÀ1
]
thermal coeff. of expansion a [KÀ1
]
3.0ꢂ10À6
0.6
3.3ꢂ10À6
0.7
specific heat capacity
density
Cp [JgÀ1 KÀ1
]
1 [gmLÀ1
]
3.15
15000
2
2.23
1400
–
thermal effusivity[b]
porosity (closed)
vickers hardness
e [JsÀ1/2 mÀ2 KÀ1
P [%]
]
HV 0.5
2300
–
Herein, we describe the development and utilization of
reaction vials designed for single-mode microwave chemistry
made entirely out of sintered SiC ceramic. Although the
concept of using a strongly microwave absorbing reaction
vessel for microwave chemistry may seem counterintuitive
or even irrational at first sight, the use of this technique pro-
vides several unique opportunities in microwave chemistry,
such as the investigation of specific/nonthermal microwave
effects, or the utilization of aggressive reaction media in a
high-temperature regime. The full details of this technology
are described in this publication.[16]
[a] Data from reference [20]. [b] The thermal effusivity e of a material is
defined as the square root of the product of its thermal conductivity and
volumetric heat capacity [e=(k1cp)0.5].
a Monowave 300 single-mode reactor (Anton Paar GmbH)
as this instrument allows simultaneous temperature mea-
surement and control by both an external infrared (IR) and
internal fiber-optic (FO) sensor.[17] Recent evidence has
demonstrated that in several instances the use of an external
IR sensor, recording the surface temperature of the vessel,
does not accurately reflect the genuine reaction temperature
inside the reaction vial and therefore may lead to erroneous
results.[18,19] As with the standard Pyrex vial, the SiC vessel
can be fitted with a thermoplastic snap cap and PTFE-
coated silicone septa to allow microwave processing up to
3008C and 30 bar pressure (Figure 1). For internal tempera-
ture measurements, punched seals to insert an immersing
tube were employed. For both vial types stirring is ensured
by magnetic stir bars.
Results and Discussion
Design and vessel construction: For the experiments de-
scribed in this work a 10 mL reaction vial made out of sin-
tered SiC of the exact same geometry as a standard micro-
wave-transparent Pyrex process vial was used (Figure 1).
The SiC vessel was fabricated by pressureless solid-phase
sintering of a green compact of silicon carbide in the pres-
ence of various sintering aids (such as carbon) at ꢀ20008C
in an inert atmosphere. As the sintering activity of SiC is
comparatively low, fine-grained powders with a grain size of
less than one micrometer were used for densification. Using
One of the important original design features of the SiC
vessel technology was the concept of shielding the reaction
mixture contained inside from the electromagnetic field in
order to investigate specific/nonthermal microwave effects
(see below).[16] Due to the high microwave absorbtivity of
Chem. Eur. J. 2010, 16, 12182 – 12194
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
12183

C. O. Kappe et al.
SiC (see Figure S1 in the Supporting Information) we have
assumed that the semiconducting ceramic will effectively
prevent microwave irradiation from penetrating to the reac-
tion mixture. Based on the results obtained with our previ-
ously used SiC vial prototype (1.8 mm wall thickness)[16] and
more recently performed electromagnetic field simula-
tions,[21] we have now moved to a 2.85 mm wall SiC vial.
This ensures that for most reaction mixtures, heating will
occur mainly by conventional heat-transfer mechanisms and
not by dielectric heating effects (see below).
The 10 mL SiC reaction vial (Figure 1) is mechanically ex-
tremely robust. While standard Pyrex microwave process
vials for single-mode microwave reactors have an approved
pressure rating of 20–30 bar and may effectively resist inter-
nal pressures of >50 bar before vessel failure,[2] the SiC tube
has a pressure rating of >200 bar and in fact destruction of
the vial could not be induced by standard pressure resist-
ance tests. Intense microwave heating of SiC vessels in a
multimode reactor leads to vial temperatures >6008C
within 2–3 min, confirming the excellent microwave absorb-
tivity of this material (Figure S2 in the Supporting Informa-
tion).
Temperature measurement: In a first set of experiments we
evaluated the heating characteristics of four common sol-
vents with vastly different loss tangents (tand)[22] in SiC
vials at constant microwave power using internal FO tem-
perature monitoring (Figure 2, top). The fact that nearly mi-
crowave transparent hexane (tand =0.02) is heated at the
same rate as the strongly absorbing EtOH (tand =0.941)
clearly indicates that the microwave field intensity inside
the SiC vial must be extremely low, and that heating occurs
in essence via conventional heat-transfer mechanisms and
not by dielectric heating effects (see below). Because of its
high thermal conductivity (ꢀ100 times higher than for
Pyrex glass, Table 1) the heat flow through the 2.85 mm wall
of the SiC reaction vessel is exceptionally fast. In addition,
due to its extremely high thermal effusivity (a measure for
the ability to exchange thermal energy with its surround-
ings) being ꢀ10 times higher than for Pyrex glass (Table 1),
the contents inside the SiC reaction vial are also heated in a
very efficient manner. A direct comparison of the heating
profiles attained in Pyrex and SiC vials for each of the four
individual solvents discussed above (Figure 2) demonstrates
that solvents are generally heated at an equal rate if not
faster in the SiC vial compared to the Pyrex vial, in particu-
lar in the high-temperature range (Figure S3).
Figure 2. Heating rates for 3 mL samples of hexane (tand=0.02),[2]
MeCN (tand=0.062),[2] EtOH (tand=0.941),[2] and bmimPF6 (tand=
0.185)[21a] at 130 W constant magnetron power in a 10 mL SiC vial
(2.85 mm wall thickness; top), and in a 10 mL Pyrex vial (2.85 mm wall
thickness; bottom). Single-mode microwave irradiation, magnetic stirring,
internal fiber-optic temperature measurement.
probe, or the external IR sensor as a lead sensor to control
the magnetron output power. Successful heating experi-
ments under temperature control in the Monowave 300 in-
strument were conducted either in the “as-fast-as-possible”
or a ramp mode.[17] In most instances, the temperatures re-
corded by the IR and FO probes were very similar, although
small differences were sometimes seen, mainly depending
on IR sensor calibration (Figure 3). Whereas temperature
control through the internal FO probe (IR as slave) is gen-
erally a beneficial choice for Pyrex vessel experiments in
Similar to standard microwave heating experiments in-
volving microwave transparent Pyrex vials, the use of higher
power levels in SiC vials, as expected, also leads to a more
rapid heating of the solvent as exemplified for N-methyl-2-
pyrrolidone (NMP; tand=0.275) in Figure S4 in the Sup-
porting Information.
Importantly, heating experiments using SiC vials in combi-
nation with single-mode microwave reactors cannot only be
performed in power control mode (Figure 2), but also in
temperature control mode by using either the internal FO
Figure 3. Temperature (internal FO and external IR), pressure (p) and
microwave power profiles (P) for the heating of 3 mL of water to 1208C
in a SiC vial performed with IR temperature control (FO probe as
slave). The corresponding profiles using FO temperature control are
shown in Figure S5 in the Supporting Information.
12184
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 12182 – 12194

Microwave Chemistry
FULL PAPER
which the selected target temperature is attained by dielec-
tric heating of the solvent, the IR sensor turned out to be
superior to the FO probe when SiC vessels were used.[23]
(Figure S7 in the Supporting Information). The effect is sim-
ilar to the use of passive heating elements (PHEs) made out
of the same ceramic material.[10] Weakly microwave absorb-
ing solvents,[22] for example toluene (tand=0.040) can be
heated much more rapidly to the desired target temperature
compared to an experiment in a Pyrex vessel (Figure S8 in
the Supporting Information). Importantly, the heating rate
using the SiC vial is largely independent on the tand of the
solvent or reaction mixture (Figure 2, top). As the tand for
most organic solvents (and water) generally decreases with
increasing temperature,[1,2,21] this effect is particularly valua-
ble for accessing a high-temperature regime in which genu-
ine dielectric heating phenomena are often difficult to sus-
tain.
An additional beneficial aspect of the use of SiC vials is
the fact that for low-absorbing reaction mixtures considera-
bly less magnetron power is used (Figure S8 in the Support-
ing Information). It is generally not advised to heat non- or
low-absorbing reaction mixtures for prolonged periods of
time in microwave reactors. This leads to the magnetron
continuously operating at the maximum power level trying
to reach the selected set temperature and ultimately can
result in overheating and damage of the magnetron or de-
structive coupling of microwave irradiation with sensitive in-
strument and/or vessel parts. Similar energy-saving effects
have been noticed using SiC passive heating elements.[10,11]
Because of the high thermal effusivity of the SiC material
(Table 1) the heat flow through the 2.85 mm wall of the SiC
vessel is exceptionally fast in both directions. This means
that the cooling efficiency by compressed air at the end of a
reaction is at least as high as when using a Pyrex vial. More
importantly, we have recently demonstrated that the use of
a SiC reaction vessel can aid in the prevention of exotherms,
as the heat generated during a microwave-assisted chemical
reaction can be efficiently exchanged with the comparatively
cool air in the microwave cavity via the SiC ceramic.[16,18]
Perhaps not surprisingly in this context, SiC ceramics have
also been employed in the construction of microreactors for
continuous-flow processing of strongly exothermic reactions,
since the thermal effusivity of SiC is even higher than that
of stainless steel.[24,25]
Heating and cooling characteristics: Figure 2 displays the
heating profiles of solvents with different tand values in SiC
and Pyrex vessels using constant microwave power. While
the heating profiles using the Pyrex vials followed the ex-
pected trend and were correlated to the tand value of the
solvent (Figure 2, bottom),[22] the heating of solvents in the
SiC vessel was apparently not related to their microwave ab-
sorbtivity, but rather dependent on other parameters such as
specific heat capacity, viscosity and heat-transfer coefficients
(Figure 2, top). Comparing the profiles obtained in the
2.85 mm SiC vial shown in Figure 2 (top) with our previous-
ly published data of essentially the same experiment, but
employing a 1.8 mm wall thickness vessel (see also Figure S6
in the Supporting Information),[16] the stronger shielding
using the thicker SiC vial is clearly evident, in particular for
the strongly microwave absorbing 1-butyl-3-methylimidazo-
lium hexafluorophosphate (bmimPF6). These experimental
data are in good agreement with electromagnetic field simu-
lations of the heating behavior of the same solvents compar-
ing 2 with 3 mm SiC vials.[21] It should be emphasized that
the microwave absorbtivity of organic solvents and of the
SiC ceramic is strongly dependent on the temperature,
which must be considered in these experiments and simula-
tions.[1,2] At room temperature the strongest microwave ab-
sorber of the five materials is EtOH, as it exhibits the high-
est tand value. At this frequency EtOH and MeCN are
heated primarily by means of a dipolar polarization mecha-
nism, hence their ability to absorb microwaves decreases
with increasing temperature as the reduced bulk viscosity
leads to reduced molecular friction.[1,21] The ionic liquid
bmimPF6 is heated by an ionic conduction mechanism, so its
ability to absorb microwaves increases with increasing tem-
perature.[1,21] At room temperature the loss tangent of SiC is
lower than both EtOH and bmimPF6, however, its tand in-
creases strongly with temperature and exhibits a peak at
2008C (Figure S1 in the Supporting Information). Both the
experimental data (heating profiles, Figure 2, top) and the
electromagnetic field simulations[21] suggest that for strongly
absorbing reaction mixtures at ambient temperature there is
still some contribution from dielectric heating phenomena,
while at temperatures >1008C most of the heating will
result from conductive heat-transfer mechanisms through
the hot vessel surface. The SiC vials can therefore be viewed
as “self-sealing” autoclaves, as with increasing temperature
less microwave energy will be able to penetrate inside, and
at the range of typical microwave-assisted transformations
(100–3008C) virtually all heating will occur by conventional
heat-transfer mechanisms.
Microwave field strength inside the SiC vials: A prime moti-
vation for the development of the SiC vial was the attempt
to shield the contents inside the vial from microwave irradi-
ation, and therefore to be able to separate thermal from
specific/nonthermal microwave effects (see below).[16] The
fact that the electric field strength and the power density
inside the SiC vial must be comparatively low can be de-
rived from the following set of experiments. First of all, the
heating profiles obtained for solvents with different micro-
wave absorbtivity at constant power (Figure 2, top) provide
clear evidence that the SiC vial provides efficient shielding
of its contents from the microwave field. The heating pro-
files for all four solvents almost run in parallel, in particular
at higher temperatures. As discussed already above, these
experiments clearly indicate that “microwave” heating in a
An apparent practical advantage of using SiC vials in mi-
crowave chemistry is that even completely microwave trans-
parent solvents or reaction mixtures can be efficiently
heated to high temperatures, which would be impossible uti-
lizing standard (nearly microwave transparent) Pyrex vials
Chem. Eur. J. 2010, 16, 12182 – 12194
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
12185

C. O. Kappe et al.
single-mode reactor using SiC vials essentially involves con-
ventional conductive heat-transfer mechanisms. These data
are in good qualitative agreement with the results obtained
from heating different solvents in parallel using SiC micro-
titer plates utilizing a multimode microwave instrument.[15]
In a complimentary set of experiments the four solvents
used in the experiment shown in Figure 2 were heated in
temperature control mode to a target temperature of 1808C.
Importantly, the used magnetron microwave power for all
four solvents—in sharp contrast to experiments in Pyrex
vessels—was virtually identical (Table 2 and Figure S9 in the
with electrodeless discharge lamps (EDLs) and filament
lamps. In the presence of a microwave field these devices
will illuminate and emit UV/Visual light.[26] Even by apply-
ing a 300 W magnetron output power in a single-mode in-
strument it was not possible to induce a gas discharge in the
SiC vial.[16] In contrast, in a standard Pyrex vial 1–5 W of mi-
crowave power was sufficient to trigger gas discharge, caus-
ing the emission of UV/Vis irradiation in these EDLs (Fig-
ure S10 in the Supporting Information).[16] Since filament
lamps do light at lower field strengths, similar experiments
were performed with these devices in order to detect the
presence of the electric field inside the SiC vessel. Indeed, a
weak lighting was observed with a 0.24 W filament lamp
(6 V, 40 mA, socketless LED-design with shortcut connec-
tors to prevent arcing) inside a SiC vessel (Figure S11 in the
Supporting Information). The calculated induced power of
0.04 W matches approximately with the observation of weak
lighting of the bulb at 15 W applied constant microwave
power in a single-mode instrument, indicating a very similar
field strength for both simulation and experiment.[21]
Table 2. Required average microwave power (P [in W]) comparing the
heating of solvents to 1808C for 20 min in SiC and Pyrex vials.[a]
SiC
Pyrex
tand P overall[b] P hold time[c] P overall[b] P hold time[c]
hexane
MeCN
EtOH
bmimPF6 0.185 20.3
empty 22.4
0.020 20.4
0.062 21.3
0.941 21.0
14.7
15.5
14.8
15.0
17.9
155.9
23.4
24.7
6.0
107.5
17.9
19.9
5.0
–
–
–
Based on the evidence presented herein, it can be con-
cluded that for most practical purposes in relation to micro-
wave chemistry, the use of SiC vials in a single-mode reactor
will provide an almost complete shielding of the contents
inside from the electromagnetic field. In essence, this tech-
nology allows to use a standard single-mode microwave re-
actor platform to perform a noncontact heating autoclave
experiment under very carefully controlled and monitored
conditions (heating and cooling ramp, final set temperature,
pressure, stirring speed). Importantly, this experiment—al-
though performed in a microwave reactor—does not involve
electromagnetic field effects on the chemistry, since the SiC
vial effectively prevents the penetration of microwave irra-
diation to the reaction mixture.
[a] Conditions: solvent (3 mL), 1808C set temperature (FO temperature
control), hold time 20 min, 10 mL SiC/Pyrex vials (2.85 mm wall thick-
ness), magnetic stirring. For a graphical representation of the required
energy see Figure S9 in the Supporting Information. [b] Overall average
power required during the entire heating process (20 min hold time and
2–3 min ramp time). [c] Average power required during the 20 min hold
time.
Supporting Information). Even for a completely empty vial,
the same magnetron power was used. These results strongly
suggest that the heating of the SiC vial in the temperature-
control mode is largely independent on its contents, which
again points to the strong shielding effect of the SiC vessel.
Additional support for the shielding concept of SiC vials
is derived from recently conducted electromagnetic field
simulations on heating effects comparing Pyrex and SiC
vials in a single-mode cavity by using the same solvent sys-
tems as used in our experimental work described herein.[21]
In short, the simulations suggest that the electric field and
power density distribution within a SiC tube is dependent
upon the dielectric properties of the solvent within it and on
the temperature. For low-absorbing solvents, such as
hexane, the power dissipated within the liquid due to dielec-
tric heating is negligible, as the tand value for hexane is
very low. For stronger absorbing solvents (MeCN, EtOH) a
significant contribution from dielectric heating effects can
be seen at room temperature in a 2 mm SiC vessel. Howev-
er, at higher temperatures (>1008C) these contributions are
significantly reduced (<10% dielectric heating) and most of
the heating will occur by conductive heat transfer from the
vessel. As expected, the simulations for 3 mm SiC vials dem-
onstrate an even stronger shielding effect.[21] Therefore, the
electromagnetic field simulations are in good overall agree-
ment with the experimental results described herein.
Investigating microwave effects: Regardless of the relatively
large body of published work in the field of microwave
chemistry, there is still considerable controversy on the
exact reasons why microwave irradiation is able to enhance
chemical processes. In particular, there is an ongoing debate
in the scientific community if the observed enhancements
are the result of purely thermal/kinetic effects as a conse-
quence of the rapid heating and high bulk reaction tempera-
tures that can be attained using microwave dielectric heat-
ing,[1,2] or are related to selective interactions of the electro-
magnetic field with specific substrate molecules, reagents, or
catalysts not connected to a macroscopic bulk temperature
effect (so-called “specific” or “nonthermal” microwave ef-
fects).[27,28] Indeed, there is experimental evidence that cer-
tain chemical transformations when carried out at the same
measured reaction temperature, by using either microwave
or conventional heating, lead to different results in terms of
product distribution (selectivity) and/or yield.[28]
The SiC vessel technology described herein therefore ap-
pears ideally suited to investigate the significance of specific
and nonthermal microwave effects, since by simply switching
from a microwave transparent Pyrex to a strongly micro-
As an additional method for estimating the shielding
effect of SiC vials, we performed a number of experiments
12186
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 12182 – 12194

Microwave Chemistry
FULL PAPER
wave absorbing SiC vial any effects of the electromagnetic
field on the reaction mixture (i.e., on the chemistry) can be
eliminated, while retaining the sealed vessel capabilities,
rapid heating/cooling and excellent process control features
inherent to modern microwave reactors.[2] To some degree,
this concept has been emulated previously by using passive
heating elements (PHEs) made out of SiC in conjunction
with standard Pyrex vials.[12] The addition of the SiC PHEs
can lead to a significant reduction (30–70%) in the required
microwave power as compared to experiments without heat-
ing element at the same temperature, and therefore in the
case of specific/nonthermal microwave effects should lead to
some measurable effects on conversion or yield.[12] Clearly,
the use of a vessel completely made out of SiC allows a
much more efficient investigation of the effects of electro-
magnetic fields on chemical transformations.
In the initial phase of our studies we have applied this
concept to a variety of microwave-assisted transformations
that
had
been
previously
investigated
in
our
group,[2,10–12,19,29,30] in which enhancements compared to stan-
dard oil bath reflux conditions had been published
(Schemes 1 and 2). The chosen reactions involved both
strongly and weakly microwave absorbing solvents in a tem-
perature range of 100–3008C using reaction times of 1–
70 min. Experiments were carried out on a 2–5 mL scale in
both low- and high-boiling solvents, and included reactions
that were either fully homogeneous, or involved heterogene-
ous starting materials, reagents, or catalysts. Gratifyingly, for
Scheme 1. Conversions/yields obtained for sealed vessel microwave experiments performed in Pyrex (P) or silicon carbide (S) reaction vials. Single-mode
microwave irradiation, magnetic stirring, internal fiber-optic temperature measurement. For further details, see the Supporting Information.
Chem. Eur. J. 2010, 16, 12182 – 12194
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
12187

C. O. Kappe et al.
all chemistry examples, it was possible to achieve similar in-
ternal FO temperature profiles switching from Pyrex to SiC
vials, with other process parameters like pressure and stir-
ring speed also being identical (Figure S12 in the Supporting
Information).
addition of 6-methylisocytosine to hexamethylenediisocya-
nate (HDI) leading to selective monourea formation
(Scheme 3).[31] This transformation is carried out in suspen-
For the eighteen chemistry examples presented in
Schemes 1 and 2, virtually identical results in terms of con-
version, purity profile, and/or isolated product yields were
Scheme 3. Nucleophilic addition of 6-methylisocytosine to hexamethyl-
AHCTUNGTREGeNNNU nediisocyanate.
sion using an excess of HDI (7.5 equiv) as a solvent. Impor-
tantly, the reaction mixture remains heterogeneous during
the addition process as the solid starting material is trans-
formed to the even less soluble urea product. Hulshof and
co-workers have recently studied this addition process and
have found that reactions performed under multimode mi-
crowave irradiation conditions were significantly faster com-
pared to experiments conducted by conventional heating,
applying the same reaction vessel, internal reaction temper-
ature, and stirring rate.[31] For example, at a reaction temper-
ature of 1008C, the isocyanate addition required ꢀ3 h to go
to full conversion using conventional heating, whereas the
microwave run was completed within ꢀ20 min.[31] The au-
thors have rationalized this apparent microwave effect by
selective heating of the solid/liquid interface, leading to an
enhanced solubility of 6-methylcytosine in the diffusion
layer and an increased reaction rate due to local overheat-
ing.[31]
A careful reinvestigation of this reaction revealed the fol-
lowing facts: in agreement with the studies by Hulshof and
co-workers[31] we found the isocyanate addition to be rather
sensitive to temperature (Figure S13 in the Supporting Infor-
mation) and to the quality of the HDI reagent. Therefore,
the same bottle of HDI was used for all comparison experi-
ments and extreme care was taken to exclude moisture
during reagent removal. In addition, we noticed from con-
ventional oil bath experiments using round-bottomed flasks
with magnetic stir bars that the stirring speed had a notable
effect on the conversion, with higher stirring rates leading to
increased conversions (Figure S14 in the Supporting Infor-
mation). By visually observing the progress of the heteroge-
neous addition process we noticed that within the range of
30–60% of conversion it was difficult to keep all solid mate-
rials properly suspended in the flask. At this stage, a solid
mass accumulated at the edges of the round-bottom flask,
thus not being effectively mixed with the HDI reagent. In-
terestingly, when the exact same reaction mixture was pro-
cessed at the identical nominal stirring speed (600 rpm) in a
10 mL Pyrex microwave process vial instead of a10 mL
round-bottomed flask, different reaction rates were ob-
served, with the flask experiment being significantly faster
than the vial experiment (Figure S15 in the Supporting In-
formation). This result clearly demonstrates that stirring/agi-
Scheme 2. Conversions/yields obtained for sealed vessel microwave ex-
periments performed in Pyrex (P) or silicon carbide (S) reaction vials.
Single-mode microwave irradiation, magnetic stirring, internal fiber-optic
temperature measurement. For further details, see the Supporting Infor-
mation.
obtained, comparing experiments that involve genuine mi-
crowave chemistry in Pyrex vials with “microwave heating”
in SiC vials. This confirms that for these specific cases only
bulk temperature effects are responsible for the observed
enhancements and that the electromagnetic field has no
direct influence on the reaction pathway (specific or non-
thermal microwave effects).[27,28] Since the time of our origi-
nal publication of this technology,[16] an additional four ex-
amples have been investigated in our laboratories, again not
exhibiting any difference between Pyrex and SiC processing
(Scheme S25 in the Supporting Information).[29]
After these successful proof-of-concept studies we wanted
to explore the use of the SiC vial concept for the investiga-
tion of more challenging chemical transformations for which
specific or nonthermal microwave effects had been explicitly
documented in the literature. Our first example involved the
12188
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 12182 – 12194

Microwave Chemistry
FULL PAPER
tation is a critical factor in this heterogeneous transforma-
tion (Scheme 3), and once again highlights the fact that the
commonly used 10 mL cylindrical microwave vials do not
provide a high stirring efficiency, in particular when hetero-
geneous mixtures or highly viscous reaction media are in-
volved.[32]
formed an experiment using the 10 mL Pyrex vial applying
conventional oil bath heating (Table 3, CONV Pyrex).
Again, the data obtained from this experiment closely
matched the microwave runs confirming the absence of any
specific/nonthermal microwave effect.
As the work of the Hulshof group[31] was originally per-
formed in a multimode microwave instrument (as opposed
to the single-mode reactor used herein), we have further-
more reproduced the chemistry shown in Scheme 3 in an
identical multimode system to exclude any influence of
varying electromagnetic field strength between single- and
multimode microwave instruments. However, also under
these conditions the conversions were similar to the conven-
tionally heated runs (Figure S18 in the Supporting Informa-
tion), and not the previously published microwave experi-
ments.[31]
With this background information in hand we next per-
formed comparison experiments for the addition of 6-meth-
ylisocytosine with HDI (Scheme 3) under microwave condi-
tions at 1008C using SiC and Pyrex vials (Ar atmosphere).
To ensure efficient agitation we employed two 10ꢂ3 mm stir
bars for both vial types. Control experiments have shown
that the use of two stir bars instead of only one significantly
enhances the agitation efficiency/reliability for this heteroge-
neous reaction mixture. This is of particular importance for
the SiC ceramic vials, in which stirring with Teflon-coated
magnetic stir bars may be compromised due to the different
surface morphology compared to glass. In our hands, when
performed at the same internal reaction temperature of
1008C (see Figure S16 in the Supporting Information for
temperature profiles), using the same heating and cooling
ramps, and applying the same stirring speed (600 rpm), the
conversion in both sets of addition experiments was identi-
cal within experimental error (Table 3 and Figure S17 in the
Supporting Information). These data suggest that no specif-
ic/nonthermal microwave effects are involved in this trans-
formation. To confirm this result we have additionally per-
As a second example, we have chosen the Leuckart re-
ductive amination of benzophenone with formamide in the
presence of formic acid as reducing agent (Scheme 4). This
Scheme 4. Leuckart reductive amination of benzophenone.
transformation was demonstrated by Loupy and co-workers
to exhibit a dramatic nonthermal microwave effect when
performed in a single-mode reactor at 2028C (IR tempera-
ture measurement): while the conventionally heated experi-
ment resulted in only 2% conversion within 30 min, the mi-
crowave heated run provided 99% conversion toward the
desired amide product during the same time period (97%
isolated yield).[33] This striking difference in conversion was
rationalized in terms of a nonthermal microwave effect, for
which the dipolar transition state in the Leuckart reductive
amination is stabilized by the electromagnetic field, thereby
leading to a lowering of activation energy and thus an en-
hancement of reactivity.[33,34]
Comparison studies between Pyrex and SiC vial single-
mode microwave experiments using the identical reaction
mixture composition but employing internal FO tempera-
ture monitoring quickly demonstrated that microwave ef-
fects are probably not involved in this chemistry. For both
types of experiments, high conversions could be realized
within 30–40 min (Figure 4). However, careful analysis of
the results indeed indicated that experiments performed ap-
plying dielectric heating in Pyrex vials provided somewhat
higher conversions compared to the identical experiments
using wall heating in SiC vials. We speculate that this phe-
nomenon may be related to an enhanced thermal decompo-
sition of the reducing agent formic acid on the hot SiC
vessel surface. It can be argued that the higher surface area
provided by the SiC ceramic (of higher porosity) will lead to
more rapid thermal decomposition of formic acid than in a
Table 3. Comparison of microwave heating in Pyrex and SiC vials for the
addition of 6-methylisocytosine to hexamethylenediisocyanate at 1008C
(Scheme 3).[a,b]
t
Heating
method
Conv.[c]
[%]
t
Heating
method
Conv.[c]
[%]
[min]
ACHTUNGTNER[NUNG min]
15
15
15
30
30
30
60
60
60
90
90
90
120
120
120
150
150
150
MW Pyrex
MW SiC
CONV Pyrex
MW Pyrex
MW SiC
CONV Pyrex
MW Pyrex
MW SiC
CONV Pyrex
MW Pyrex
MW SiC
CONV Pyrex
MW Pyrex
MW SiC
CONV Pyrex
MW Pyrex
MW SiC
3.6Æ1.5
3.9Æ0.9
180
180
180
210
210
210
240
240
240
270
270
270
300
300
300
MW Pyrex
MW SiC
CONV Pyrex
MW Pyrex
MW SiC
CONV Pyrex
MW Pyrex
MW SiC
CONV Pyrex
MW Pyrex
MW SiC
68.1Æ3.7
63.9Æ3.3
69.7Æ4.6
77.4Æ3.5
73.5Æ2.5
77.9Æ5.2
83.7Æ5.0
80.9Æ2.3
84.3Æ4.9
89.7Æ2.6
86.6Æ1.8
89.1Æ5.9
93.0Æ1.6
91.1Æ1.5
93.0Æ4.5
4.1Æ0.7
9.4Æ0.6
9.1Æ0.9
12.5Æ2.9
20.9Æ2.2
19.6Æ2.8
24.4Æ2.1
30.4Æ3.7
30.8Æ3.5
37.2Æ3.2
45.7Æ4.4
43.3Æ3.6
48.6Æ3.9
57.3Æ4.4
54.2Æ2.7
60.6Æ4.4
CONV Pyrex
MW Pyrex
MW SiC
CONV Pyrex
CONV Pyrex
[a] 6-Methylisocytosine (2.00 mmol) and hexamethylenediisocyanate
(15.0 mmol) were reacted neat at 1008C in 10 mL Pyrex or SiC vials em-
ploying either microwave (MW) or conventional oil bath heating
(CONV). Magnetic stirring, 600 rpm, two stir bars. For further details,
see the Experimental Section. [b] Internal reaction temperature mea-
sured by fiber-optic probe sensor. The heating profiles and a graphical
representation of conversion are reproduced in the Supporting Informa-
tion (Figure S16 and 17 in the Supporting Information). [c] Determined
by HPLC-UV analysis at 254 nm. All experiments were performed in
triplicate.
Chem. Eur. J. 2010, 16, 12182 – 12194
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
12189

C. O. Kappe et al.
enhanced thermal decomposition of the Co catalyst on the
hot SiC vessel surface of higher porosity.
Controlling thermal runaways and temperature overshoots:
The data presented so far have demonstrated that the use of
SiC vials in a single-mode microwave instrument essentially
mimics a conventionally heated autoclave experiment under
carefully controlled reaction conditions (internal tempera-
ture, pressure). Apart from the possibility to separate ther-
mal from specific/nonthermal microwave effects, an addi-
tional benefit of this technology involves controlling strongly
microwave absorbing and/or exothermic reaction mixtures,
in particular under solvent-free conditions. A case in point
is the synthesis of ionic liquids via alkylation of nitrogen-
containing heterocycles, which is notoriously difficult to con-
trol under microwave conditions.[18] Not only are these alky-
lations generally exothermic, but in addition the microwave
absorbtivity during the process changes significantly from
moderate (starting materials) to high (ionic liquids). To fur-
ther complicate matters, the tand of ionic liquids increases
with increasing temperature (for bmimPF6 the tand is 0.184
at 258C and 3.592 at 2008C),[21] therefore setting the stage
for thermal runaways under microwave conditions. Using
SiC reaction vials, the fact that the microwave absorbtivity
of the components is changing during the formation of the
ionic liquid is largely irrelevant, as the SiC vial efficiently
prevents microwave irradiation from penetrating to the re-
action mixture and heating will occur by conventional heat-
transfer mechanisms. Coupled with the ꢀ10 times higher ef-
fusivity of SiC compared to Pyrex (see Table 1) the genera-
tion of temperature overshoots or exotherms can often be
avoided, as the heat generated during ionic liquid formation
is efficiently exchanged with the comparatively cool air in
the microwave cavity through the SiC ceramic.[18]
Figure 4. Conversion over time for the Leuckart reductive amination of
benzophenone with formamide in the presence of formic acid
(Scheme 4). Experiments were performed at 2028C in triplicate using
Pyrex and SiC reaction vessels. Conversions were monitored by GC-FID
analysis. The heating profiles are reproduced in the Supporting Informa-
tion (Figure S19). For further details, see the Experimental Section.
Pyrex glass vessel. Evidence for this hypothesis was derived
by heating of pure formic acid to 2028C in SiC vials. The de-
composition of formic acid was notably faster than in Pyrex
vials as seen by the more rapid rise in pressure (Figure S20
in the Supporting Information).
An additional example in which the use of Pyrex and SiC
vials does not lead to identical results involves the Co-cata-
lyzed [2+2+2] cyclotrimerization of 1,2-dipropargylben-
zene with benzonitrile leading to a 9,10-dihydro-2-azaan-
thracene derivative (Scheme 5). Studies published by Dei-
Additional examples in which thermal runaways using mi-
crowave dielectric heating have been experienced are the al-
kyation of 1,2,4-triazole with 2,2’4’-trichloroacetophenone
(Scheme 1 m), and the alkylation of triphenylphosphine with
benzyl chloride (Scheme 1l). While in the presence of p-
xylene as solvent both alkylations can be reasonably con-
trolled using microwave dielectric heating principles (Sche-
me 1l and 1 m), under solvent-free conditions both transfor-
mations lead to strong exotherms, temperature overshoots
and inhomogeneities, and ultimately to thermal runaways.[19]
In particular in the case of triazole alkyation severe thermal
runaways were previously observed making this chemistry
unsuitable for microwave processing.[19] In contrast, the use
of SiC vials alleviates these problems to a significant extent
and even under solvent-free conditions, reproducible results
in a more controllable temperature regime can now be ob-
tained (Figures S21 and S22 in the Supporting Information).
It should be emphasized that the significant advantage of
performing these reactions in SiC vials inside a microwave
instrument—as opposed to a classical oil bath experiment—
lies in the noncontact-heating nature and the dynamic tem-
perature control of this method. Using a conventional oil
bath, exothermic reactions are often not noticed and remain
Scheme 5. Co-catalyzed [2+2+2] cyclotrimerization reactions. P=Pyrex
vial; S=SiC vial.
ters and co-workers in 2008 comparing microwave and con-
ventional heating at the same monitored reaction tempera-
ture have suggested the existence of
a nonthermal
microwave effect for this transformation.[35] A more recent
investigation using accurate internal reaction temperature
control and matching heating and cooling profiles for both
types of heating modes as close as possible has demonstrat-
ed that no microwave effects exist for this [2+2+2] cyclo-
trimerization reactions.[36] The Co-catalyzed [2+2+2] cyclo-
trimerization required 1608C to lead to complete conversion
within 20 min under microwave conditions and furnished a
78% isolated product yield. Performing an oil bath experi-
ment with a similar temperature profile as in the microwave
run also gave full conversion and a similar isolated product
yield (75%).[36] However, using the SiC vial under identical
conditions consistently provided conversion in the range of
only 45–60%. As in the case of the Leuckart reductive ami-
nation described above, we ascribe this phenomenon to an
12190
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 12182 – 12194

Microwave Chemistry
FULL PAPER
undetected unless the internal reaction temperature is care-
fully monitored. Even if a thermal overshoot is recognized
by internal temperature monitoring, there is no easy way of
reducing the thermal energy going into the system unless
the reaction vessel is physically removed from the bath
(contact heating). In contrast, a thermal overshoot in the
SiC microwave system is easily recognized by the internal
FO probe, which leads to an immediate adjustment of the
magnetron output power, or even to an active cooling step
by compressed air.
corrosion of microwave vials during usage represents a seri-
ous safety risk as the pressure rating of the heavy-walled
Pyrex vials (20-30 bar) cannot be maintained. Prolonged ex-
posure of Pyrex vials to TREAT-HF at high temperatures
and pressures or repeated vial usage must therefore be
strictly avoided. In addition, hydrogen fluoride mediated
vessel corrosion at these temperatures leads to the forma-
tion of gel-type materials, making an extractive workup of
the reaction mixture troublesome.
In contrast, the SiC vial proved to be completely resistant
to TREAT-HF even at 2508C for prolonged periods of time
(no measurable weight loss). Furthermore, workup of the re-
action mixture was simplified due to the absence of glass
corrosion products.
Similar investigations on the corrosion stability of SiC
versus Pyrex reaction vials under microwave conditions
were performed in the context of the high-temperature hy-
drolysis of aromatic nitriles using strong aqueous bases. The
hydrolysis of a nitrile function in alkaline medium to obtain
a carboxylic acid is a commonly used transformation in or-
ganic synthesis. The described conditions for hydrolysis, in
particular for aromatic nitriles, are in most cases rather
harsh, involving long reaction times and/or high tempera-
tures.[38] For our model studies, the hydrolysis of benzonitrile
and 2-methoxybenzonitrile using 20% aqueous KOH in a
temperature range of 100–1908C was evaluated (Scheme 7).
Corrosion stability and thermal resistance: Sintered SiC is a
universally corrosion-resistant material that exhibits excel-
lent heat stability and owing to its high hardness additional-
ly reduces wear and erosion.[3] This is the reason why SiC is
used for many key components in chemical plants and has
recently also been employed for the fabrication of gas-tight
microreactors operating at high temperatures (7008C) and
harsh environments.[24] The corrosion resistance of sintered
SiC is, like most ceramics, far better than of metals and even
aggressive media such as concentrated acids or bases, HF, or
chlorine gas have a negligible impact in terms of corro-
sion.[24] Based on these facts we have employed SiC vials for
conducting microwave-assisted, aliphatic fluorine–chlorine
exchange reactions using triethylamine trihydrofluoride
(TREAT-HF, Et3N·3HF) as a reagent.[36] TREAT-HF is a
commercially available, colorless, hygroscopic liquid with a
shelf life of at least one year and has been reported not to
corrode borosilicate glassware at ambient conditions.[37] Not
unlike ionic liquids, the distinctly polar properties of
TREAT-HF make it very suitable as a reagent (and/or sol-
vent) in microwave-heated transformations. In fluorine–
chlorine exchange reactions involving (trichloromethylthio)-
benzene at 2508C (Scheme 6)[36] we did, however, notice a
Scheme 7. Hydrolysis of aromatic nitriles using 20% aqueous KOH.
While for benzonitrile complete hydrolysis to benzoic acid
could be achieved after 1 h at 1508C (2 equiv KOH, 20%
aqueous solution), the sterically hindered 2-methoxybenzo-
nitrile required exposure at either 1708C for 2 h or at 1908C
for 1 h (5 equiv KOH, 20% aqueous solution). The results
in terms of HPLC conversion for experiments performed in
Pyrex and SiC vials were essentially the same; however, iso-
lated product yields in the SiC vial were significantly higher
(>85%), since hydrolysis experiments performed in Pyrex
vials resulted in large amounts of glass degradation products
(silicates), making product isolation troublesome.
A more detailed evaluation of the corrosion of standard
Pyrex microwave process vials with aqueous KOH demon-
strated that severe degradation is typically obtained at tem-
peratures>1508C, largely independent on the concentration
of KOH (20–60%) (Figure S24 and Table S1 in the Support-
ing Information). In several of these experiments vessel fail-
ures resulting in severe explosions were observed, in partic-
ular in the high-temperature/pressure range, or when the
filling volume was too low. In the latter case we assume that
evaporation of solvent (large headspace) leads to deposition
of solid KOH at the liquid boundary that will strongly inter-
Scheme 6. Fluorination reactions using triethylamine trihydrofluoride
(TREAT-HF).
significant corrosion of standard Pyrex microwave reaction
vials in the vapor space above the liquidꢃs surface. Although
generally considered noncorrosive,[37] TREAT-HF can re-
lease hydrogen fluoride at elevated temperatures and will
therefore attack borosilicate glass to some extent. In a con-
trol study measuring the weight loss of unused fresh Pyrex
microwave reaction vessels after exposure to TREAT-HF
under microwave conditions, we discovered that corrosion
occurs even at comparatively low temperatures (1008C) and
is highly dependent on reaction temperature and time.
While at 1008C the weight loss after 5 min is already meas-
urable but comparatively small, a very significant loss of
glass was encountered after a 30 min exposure at 2508C
(Figure S23 in the Supporting Information). Importantly, the
Chem. Eur. J. 2010, 16, 12182 – 12194
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
12191

C. O. Kappe et al.
act with the microwave field and can lead to a localized
melting of Pyrex glass. Experiments under microwave condi-
tions involving strong bases in combination with Pyrex vials
should therefore be avoided. In particular in the high-tem-
perature range, significant internal pressures will develop
(10–20 bar) that can lead to catastrophic vessel failures.
Control experiments with SiC vials demonstrated that the
ceramic material is completely corrosion resistant to KOH
as no weight loss of the vial was detected, even after expo-
sure to 20% KOH at 2108C for 1 h.
An additional feature of using a reaction vial made out of
a strongly microwave-absorbing and high-temperature re-
sistant material in microwave chemistry is that vessel fail-
ures due to the deposition of strongly absorbing materials
on the inside such as metals can be eliminated. In particular
in the field of homogeneous transition-metal catalysis using
microwave conditions[39] a common phenomenon is that the
metal catalyst (for example Pd) ultimately is being deposit-
ed on the inner vessel wall due to catalyst decomposition in-
itiated by the high reaction temperatures.[2,40] The inadver-
tently formed and strongly microwave absorbing solid zero-
valent metal deposits can cause melting of the Pyrex glass
or will lead to arcing phenomena.[40,41] Both effects ultimate-
ly may result in catastrophic vessel failures and to severe in-
strument damage. As the electromagnetic field strength
inside the SiC vial is dramatically reduced, the possibility
for arcing phenomena is minimized. In addition, the signifi-
cantly higher melting point of SiC compared to Pyrex glass
(see Table 1) will prevent melting of the reaction vial and
therefore should allow safe microwave processing even in
the presence of metals or other strongly microwave absorb-
ing materials. In all the experiments performed with SiC
vials in our group so far vessel failure or vial damage has
never been observed.
absorbtivity, thermal conductivity, and effusivity of SiC
allows a very rapid heat exchange with its surroundings,
which is not only important for achieving fast heating rates
of the reaction mixtures, but can also prevent exotherms
due to the noncontact heating nature of this technology. The
method described herein therefore makes it possible to rap-
idly separate thermal from specific/nonthermal effects, since
by employing the SiC vial any effects of the electromagnetic
field on the reaction mixture can be eliminated. Future
work using this technology will be aimed at studying the use
of SiC vials in other areas ranging from organic and polymer
synthesis to nanomaterials research, and investigating the
scale-up potential.
Experimental Section
General remarks: 1H NMR spectra were recorded on a Bruker 300 MHz
instrument. 13C NMR spectra were recorded on the same instrument at
75 MHz. Chemical shifts (d) are expressed in ppm downfield from TMS
as internal standard. The letters s, d, t, q, and m are used to indicate sin-
glet, doublet, triplet, quadruplet, and multiplet, respectively. Low-resolu-
tion mass spectra were obtained on an Agilent 1100 LC/MS instrument
using atmospheric pressure chemical ionization (APCI) in positive or
negative mode. GC-FID analysis was performed on a Trace-GC (Ther-
moFisher) with a flame ionization detector using a HP5 column (30 m ꢂ
0.250 mm ꢂ 0.025 mm). After 1 min at 508C the temperature was in-
creased in 258CminÀ1 steps up to 3008C and kept at 3008C for 4 min.
The detector gas for the flame ionization is H2 and compressed air (5.0
quality). GC-MS spectra were recorded using a Thermo Focus GC cou-
pled with a Thermo DSQ II (EI, 70 eV). A HP5-MS column (30 mꢂ
0.250 mmꢂ0.025 mm) was used with Helium as carrier gas (1 mLminÀ1
constant flow). The injector temperature was set to 2808C. After 1 min at
508C the temperature was increased in 258CminÀ1 steps up to 3008C and
kept at 3008C for 4 min. Analytical HPLC (Shimadzu LC20) analysis was
carried out on a C18 reversed-phase (RP) analytical column (150ꢂ
4.6 mm, particle size 5 mm) at 258C using a mobile phase A (water/aceto-
nitrile 90:10 (v/v) + 0.1% TFA) and B (MeCN + 0.1% TFA) at a flow
rate of 1.0 mLminÀ1. The following gradient was applied: linear increase
from solution 30% B to 100% B in 8 min, hold at 100% solution B for
Conclusion
2 min. TLC analyses were performed on pre-coated (silica gel 60 HF254
)
plates. Flash chromatography separations were performed on a Biotage
SP1 instrument using petroleum ether/ethyl acetate mixtures as eluent.
Melting points were determined on a Stuartꢄ SMP3 melting point appa-
ratus. All anhydrous solvents (stored over molecular sieves), and chemi-
cals were obtained from standard commercial vendors and were used
without any further purification.
In summary, we have demonstrated that the use of reaction
vials made out of sintered silicon carbide (SiC) ceramic in
microwave chemistry provides an interesting alternative to
the more commonly used standard Pyrex glass vessels. The
high melting point and pressure stability of SiC combined
with its excellent chemical resistance allows the safe use of
corrosive reagents in a high-temperature/pressure regime
not suitable for processing in standard glass vials. The virtu-
ally indestructible reaction vials have been used up to
3008C and 30 bar pressure, and potentially could be used at
even higher temperatures and pressures due the specific
properties of this ceramic material.
Importantly, as shown by electromagnetic field simula-
tions and heating experiments of solvents with different mi-
crowave absorbtivity, performing a microwave chemistry ex-
periment in a SiC vessel mimics a conventionally heated au-
toclave experiment, while retaining the rapid heating (flash
heating) and excellent process control features inherent to
microwave chemistry. The combination of high microwave
Microwave irradiation experiments: Microwave irradiation experiments
were performed using a Monowave 300 single-mode microwave reactor
from Anton Paar GmbH (Graz, Austria).[17] The instrument uses a maxi-
mum of 850 W magnetron output power (2.45 GHz) and can be operated
at 3008C reaction temperature and 30 bar pressure. The reaction temper-
ature was monitored by an external infrared sensor (IR) housed in the
side-walls of the microwave cavity measuring the surface temperature of
the reaction vessel, and/or by an internal fiber-optic (FO) temperature
probe (ruby thermometer)[42] protected by a borosilicate immersion well
inserted directly to the reaction mixture. The magnetron output power
can either be controlled by the FO probe (IR as slave) or by the IR
sensor (FO as slave). Heating experiments can be performed in tempera-
ture control mode (“as-fast-as-possible”, ramp mode), or using a constant
power option. Pressure sensing was achieved by a hydraulic sensor inte-
grated in the swiveling cover of the instrument. The reusable Pyrex vials
(10 mL and 30 mL) were sealed with PEEK snap caps and standard
PTFE coated silicone septa. In case of FO temperature measurement,
punched seals to insert the immersing tube were employed. Seals and
12192
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 12182 – 12194

Microwave Chemistry
FULL PAPER
caps can be used for both Pyrex and the 10 mL SiC vials. Precision of in-
ternal temperature measurement was provided by efficient stirring at a
fixed rate of 600 rpm. Reaction cooling was performed by compressed air
automatically after the heating period has elapsed. The required pressure
of 6–8 bar was also used to pneumatically seal the vials tightly at the be-
ginning to withstand 30 bar, and to ensure smooth release of potentially
remaining pressure before opening the cover.
[4] a) S. C. Fong, C. Y. Wang, T. H. Chang, T. S. Chin, Appl. Phys. Lett.
2009, 94, 102104; b) G. Wei, W. Qin, R. Kim, G. Wang, P. Zhu, D.
Thang, K. Zheng, L. Wang, Appl. Phys. A 2009, 96, 521; c) X. Fang,
R. Hutcheon, D. A. Scola, J. Polym. Sci. Polym. Chem. Ed. 2000, 38,
2526.
[5] H. Y. Cho, A. Ajaz, D. Himali, P. A. Waske, R. P. Johnson, J. Org.
Chem. 2009, 74, 4137.
[6] C. Dong, J. Guo, G. C. Fu, L. H. Yang, H. Chen, Supercond. Sci.
Technol. 2004, 17, L55.
[7] J. Lasri, P. D. Ramesh, L. Schachter, J. Am. Ceram. Soc. 2000, 83,
1465.
[8] E. F. Sturcken, Ceram. Trans. 1991, 21, 433.
[9] Y. Hawangchu, D. Atong, V. Sricharoenchaikul, Int. J. Chem. React.
Eng. 2010, 8, A65.
[10] J. M. Kremsner, C. O. Kappe, J. Org. Chem. 2006, 71, 4651.
[11] T. Razzaq, C. O. Kappe, ChemSusChem 2008, 1, 123.
[12] T. Razzaq, J. M. Kremsner, C. O. Kappe, J. Org. Chem. 2008, 73,
6321.
[13] J. Geuens, J. M. Kremsner, B. A. Nebel, S. Schober, R. A. Dom-
misse, M. Mittelbach, S. Tavernier, C. O. Kappe, B. U. W. Maes,
Energy Fuels 2008, 22, 643.
Sintered silicon carbide (SiC) vials: For the experiments described herein
either a standard 10 mL Pyrex tube (average weight: 25.4 g) or a vessel
made out of sintered silicon carbide (ROCARꢁ S; average weight: 39.0 g)
of the exact same geometry was used (Figure 1). Both vials are commer-
cially available from Anton Paar GmbH,[17] have a wall thickness of
2.85 mm and are fabricated for use with the Monowave 300 instrument.
As the SiC vial is opaque, any residues remaining inside from a particular
microwave experiment are difficult to identify. A thorough cleaning
regime is recommend after each experiment (acetone, water, KOH/iso-
propanol, aqua regia) to ensure that traces of materials attached to the
inner surface will not interfere in subsequent reactions. For strongly cor-
rosive reaction mixtures, the use of the FO temperature probe housed in
a borosilicate immersion well should be avoided. As the surface morphol-
ogy of the SiC is different from Pyrex glass, for strongly heterogeneous
reaction mixtures the use of two 10ꢂ3 mm stir bars is recommended for
SiC vials.
[14] J. M. Kremsner, A. Stadler, C. O. Kappe, J. Comb. Chem. 2007, 9,
285.
Comparison of SiC and Pyrex vessel experiments: All chemical transfor-
mations studied in this work (Scheme 1–7) are known in the literature
and for the most part (Schemes 1, 2, 5, and 6) have been previously in-
vestigated and optimized in our laboratories using standard single-mode
microwave technology. For the comparison experiments involving SiC
and Pyrex vials, the FO probe was chosen as lead sensor for all Pyrex
vial experiments, whereas regulation with the IR sensor as lead sensor
generally proved to be superior in combination with SiC vials (especially
in combination with reaction mixtures with high thermal inertia, for ex-
ample, high filling volumes). The corresponding SiC and Pyrex experi-
ments were conducted at the same internal FO temperature and hold-
time (see Figure S12 in the Supporting Information). Care was taken to
ensure efficient stirring in all experiments. The procedure for conducting
oil bath heating experiments with 10 mL Pyrex microwave process vials
and internal FO temperature monitoring (Scheme 3) has been described
previously.[19] Conversions and purities of the crude, homogenized reac-
tion mixtures were determined (where possible) by HPLC (215 or
254 nm), GC-FID and/or GC-MS, or 1H NMR spectroscopy. The purities
of the isolated products were confirmed by HPLC at 215 nm/254 nm and
[15] a) M. Damm, C. O. Kappe, J. Comb. Chem. 2009, 11, 460; b) M.
Damm, C. O. Kappe, Mol. Diversity 2009, 13, 529.
[16] For a preliminary report, see: D. Obermayer, B. Gutmann, C. O.
Kappe, Angew. Chem. 2009, 121, 8471; Angew. Chem. Int. Ed. 2009,
48, 8321.
[17] For further details on the Monowave 300 single-mode microwave re-
actor (Anton Paar GmbH, Graz, Austria), see http://www.anton-
paar.com/monowave300.
[18] D. Obermayer, C. O. Kappe, Org. Biomol. Chem. 2010, 8, 114.
[19] M. A. Herrero, J. M. Kremsner, C. O. Kappe, J. Org. Chem. 2008,
73, 36, and references therein.
[20] a) Borosilicate-glass technical data sheet: http://www.duran-group.
com; b) Information on silicon carbide (ROCAR S): http://www.
ceramtec.com; c) R. C. Weast, CRC Handbook of Chemistry and
Physics, 65th ed., CRC Press, New York, 1985.
[21] a) J. Robinson, S. Kingman, D. Irvine, P. Licence, A. Smith, G. Dimi-
trakis, D. Obermayer, C. O. Kappe, Phys. Chem. Chem. Phys. 2010,
12, 10793; see also: b) J. Robinson, S. Kingman, D. Irvine, P. Li-
cence, A. Smith, G. Dimitrakis, D. Obermayer, C. O. Kappe, Phys.
Chem. Chem. Phys. 2010, 12, 4750.
[22] The ability of a specific solvent to convert microwave energy into
heat at a given frequency and temperature is determined by the so-
called loss tangent (tand), expressed as the quotient, tand=e’’/e’, in
which e’’ is the dielectric loss, indicative of the efficiency with which
electromagnetic radiation is converted into heat, and e’ is the dielec-
tric constant, describing the ability of molecules to be polarized by
the electric field. In general, solvents used for microwave synthesis
can be classified as high (tand>0.5, for example: ethanol, DMSO,
methanol, formic acid), medium (tand=0.1–0.5, for example: acetic
acid, 1,2-dichlorobenzene, NMP, DMF, water), and low microwave
absorbing (tand<0.1, for example: chloroform, ethyl acetate, THF,
dichloromethane, toluene, hexane). See references [1] and [2] for
more details.
1
the products were identified by H NMR and by their melting points. The
exact methods of preparation including work-up, spectral and analytical
data for the obtained products can be found in the Supporting Informa-
tion.
Acknowledgements
This work was supported by a grant from the Christian Doppler Society
(CDG). We acknowledge Anton Paar GmbH for the provision of the
Monowave microwave reactor and technical support.
[23] When using SiC vials most of the microwave energy is absorbed by
the SiC vial and heat is subsequently transferred from the hot sur-
face to the reaction mixture via conduction/convection phenomena.
As the temperature control algorithm of the microwave instrument
in FO mode is optimized for Pyrex vials, overshoots and oscillations
of the temperature, especially for reaction mixtures with high ther-
mal inertia, for example high filling volumes, can sometimes be ob-
served (Figure S5 in the Supporting Information).
[24] F. Meschke, G. Riebler, V. Hessel, J. Schꢅrer, T. Baier, Chem. Eng.
Technol. 2005, 28, 465.
[25] J. H. Park, T. H. Yoon, H. J. Lee, D. P. Kim, J. Appl. Chem. 2006, 10,
160.
[1] a) D. R. Baghurst, D. M. P. Mingos, Chem. Soc. Rev. 1991, 20, 1;
b) C. Gabriel, S. Gabriel, E. H. Grant, B. S. Halstead, D. M. P.
Mingos, Chem. Soc. Rev. 1998, 27, 213.
[2] C. O. Kappe, D. Dallinger, S. S. Murphree, Practical Microwave Syn-
thesis for Organic Chemists—Strategies, Instruments, and Protocols,
Wiley-VCH, Weinheim, 2009.
[3] a) Properties of Silicon Carbide (Ed.: G. L. Harris), Institute of Elec-
trical Engineers, London, 1995; b) Silicon Carbide: Recent Major
Advances (Eds.: W. J. Choyke, H. Matsunami, G. Pensl, ), Springer,
Berlin, 2004; c) Advances in Silicon Carbide Processing and Applica-
tions (Eds.: S. E. Saddow, A. Agarwal), Artech House Inc., Nor-
wood, 2004.
Chem. Eur. J. 2010, 16, 12182 – 12194
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
12193

C. O. Kappe et al.
[26] P. Klꢆn, V Cꢇrkva in Microwaves in Organic Synthesis, 2nd ed. (Ed.:
A. Loupy), Wiley-VCH, Weinheim, 2006, Chapter 19, pp. 860–897.
[27] For a definition and examples for thermal, specific and nonthermal
microwave effects, see: a) C. O. Kappe, Angew. Chem. 2004, 116,
6408; Angew. Chem. Int. Ed. 2004, 43, 6250; b) See also refer-
ence [2].
[28] For leading reviews, see: a) L. Perreux, A. Loupy, Tetrahedron 2001,
57, 9199; b) L. Perreux, A. Loupy in Microwaves in Organic Synthe-
sis, 2nd ed. (Ed.: A. Loupy), Wiley-VCH: Weinheim, 2006, Chap-
ter 4, pp. 134–218; c) A. De La Hoz, A. Diaz-Ortiz, A. Moreno,
Chem. Soc. Rev. 2005, 34, 164; d) A. De La Hoz, A. Diaz-Ortiz, A.
Moreno in Microwaves in Organic Synthesis, 2nd ed. (Ed.: A.
Loupy), Wiley-VCH, Weinheim, 2006, Chapter 5, pp. 219–277.
[29] a) M. Damm, T. N. Glasnov, C. O. Kappe, Org. Process Res. Dev.
2010, 14, 215; b) M. Fuchs, W. Goessler, C. Pilger, C. O. Kappe, Adv.
Synth. Catal. 2010, 352, 323.
[36] D. Dallinger, M. Irfan, A. Suljanovic, C. O. Kappe, J. Org. Chem.
2010, 75, 5278.
[37] For a preliminary report, see: J. M. Kremsner, M. Rack, C. Pilger,
C. O. Kappe, Tetrahedron Lett. 2009, 50, 3665.
[38] a) S. Hara in Encyclopedia of Reagents for Organic Synthesis, Wiley,
New York, 2005; b) M. A. McClinton, Aldrichimica Acta 1995, 28,
31; c) G. Haufe, J. Prakt. Chem. 1996, 338, 99.
[39] G. Van Baelen, B. U. W. Maes, Tetrahedron 2008, 64, 5604.
[40] For general reviews on microwave-assisted transition-metal catalysis,
see: a) P. Appukkuttan, E. Van der Eycken, Eur. J. Org. Chem. 2008,
1133; b) P. Nilsson, K. Olofsson, M. Larhed, Top. Curr. Chem. 2006,
266, 103; c) N. E. Leadbeater, Chem. Commun. 2005, 2881; d) M.
Larhed, C. Moberg, A. Hallberg, Acc. Chem. Res. 2002, 35, 717.
[41] a) M. D. Bowman, N. E. Leadbeater, T. M. Barnard, Tetrahedron
Lett. 2008, 49, 195; b) J. D. Moseley, E. K. Woodman, Org. Process
Res. Dev. 2008, 12, 967; c) G. Shore, W.-J. Yoo, C.-J. Li, M. G.
Organ, Chem. Eur. J. 2010, 16, 126; d) B. H. Lipshutz, B. A. Frie-
man, C.-T. Lee, A. Lower, D. M. Nihan, B. R. Taft, Chem. Asian J.
2006, 1, 417.
[30] a) T. N. Glasnov, S. Findenig, C. O. Kappe, Chem. Eur. J. 2009, 15,
1001; b) T. Razzaq, T. N. Glasnov, C. O. Kappe, Eur. J. Org. Chem.
2009, 1321; c) S. S. Murphree, C. O. Kappe, J. Chem. Educ. 2009, 86,
227; d) A. Stadler, S. Pichler, G. Horeis, C. O. Kappe, Tetrahedron
2002, 58, 3177; e) D. Dallinger, C. O. Kappe, Nat. Prot. 2007, 2,
1713.
[42] For related arcing phenomena involving Mg turnings, see: B. H. P.
van de Kruijs, M. H. C. L. Dressen, J. Meuldijk, J. A. J. M. Veke-
mans, L. A. Hulshof, Org. Biomol. Chem. 2010, 8, 1688.
[31] a) M. H. C. L. Dressen, B. H. P. van de Kruijs, J. Meuldijk, J. A. J. M.
Vekemans, L. A. Hulshof, Org. Process Res. Dev. 2007, 11, 865;
b) M. H. C. L. Dressen, Ph.D. Thesis, Eindhoven University of Tech-
nology (The Netherlands), 2009; Chapter 4.
[32] For the effect of stirring efficiency on conversion in microwave-as-
sisted transformations, see: M. Irfan, M. Fuchs, T. N. Glasnov, C. O.
Kappe, Chem. Eur. J. 2009, 15, 11608.
[43] The measuring principle in the ruby thermometer is the decay time
of the photoluminescence of a ruby (corundum, Al2O3:Cr) crystal.
The crystal at the tip of the flexible fiber is excited by blinking light,
causing the luminescence. The decay time is a temperature depen-
dent physical constant of the material, therefore it can be used for
temperature measurement. Once adjusted this system is a calibra-
tion free tool for temperature measurement over the whole opera-
tion range and can also be used to adjust the IR sensor and the pres-
sure sensor of the system.
[33] A. Loupy, D. Monteux, A. Petit, J. M. Aizpurua, E. Dominguez, C.
Palomo, Tetrahedron Lett. 1996, 37, 8177.
[34] L. Perreux, A. Loupy in Microwaves in Organic Synthesis, 2nd ed.
(Ed.: A. Loupy), Wiley-VCH, Weinheim, 2006, Chapter 4, p. 211.
[35] Y. Zou, D. D. Young, A. Cruz-Montanez, A. Deiters, Org. Lett.
2008, 10, 4661.
Received: June 16, 2010
Published online: September 15, 2010
12194
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 12182 – 12194