Inductively heated oxides inside microreactors - Facile oxidations under flow conditions

Article abstract of DOI:10.1002/ejoc.201000628

Inductively heated iron oxides mixed with solid oxidants like CrO ₂ and NiO₂ efficiently oxidize organic substrates under flow conditions in fixed-bed reactors.

Full text of DOI:10.1002/ejoc.201000628

SHORT COMMUNICATION
DOI: 10.1002/ejoc.201000628
Inductively Heated Oxides Inside Microreactors – Facile Oxidations under
Flow Conditions
Jens Wegner,[a] Sascha Ceylan,[a] Carsten Friese,[b] and Andreas Kirschning*[a]
Keywords: Inductive heating / Microreactors / Oxidation / Metal oxides
Inductively heated iron oxides mixed with solid oxidants like CrO2 and NiO2 efficiently oxidize organic substrates under
flow conditions in fixed-bed reactors.
changing magnetic field.[2,3] Additionally, a very strong
rapidly alternating magnetic field will also induce eddy cur-
rents on any conductive material placed in the vicinity of
that field. These eddy currents are also able to heat the
metal by induction but to a lesser extent than magnetic
materials that mainly produce heat through the so called
Introduction
Recently, we disclosed the first application of ferromag-
netic materials like magnetic nanoparticles 1, coated with a
silica shell, as heating media in chemical reactions (Fig-
ure 1).[1]
hysteresis effect. With conductive materials, about 80% of
the heating effect occurs on the surface. Small or thin parts
generally heat more quickly than large thick parts, espe-
cially if the larger parts need to be heated all the way
through. Therefore, magnetic nanoparticles are ideally
suited for thermal activation by electromagnetic fields.
Results and Discussion
Our preliminary results clearly showed that inductive
Figure 1. Schematic drawing of magnetic nanoparticles 1 and tech-
heating of ferromagnetic materials inside flow reactors has
great future prospects particularly because the technical
setup is much simpler than corresponding microwave de-
vices so that it has the potential of becoming an additional
enabling technology for organic synthesis.[4]
nical setup.
We showed that inductive heating allows to perform
many principle endothermic reactions such as transesterifi-
cations, condensations, Pd-catalyzed cross coupling reac-
tions and Wittig olefinations under flow conditions. Al-
though magnetic nanoparticles have gained considerable
interest lately due to their magnetic properties, which allow
to remove them from a reaction mixture with a magnet,[2]
the important property of heating them in an electromag-
netic field has been, with the exception of our first com-
munication, unexploited in organic synthesis. The concept
of magnetic inductive hyperthermia is based on the fact that
loss of magnetic hysteresis creates heat (Néel relaxation),
when magnetic nanoparticles are exposed to a constantly
In this report we extend our studies on metal oxide medi-
ated oxidations under flow conditions. Several oxides such
as CrO2 (MagTrieveTM [5]
and NiO2 are widely employed
)
solid oxidants in organic chemistry. Commonly, the reduced
byproducts are also solids and should not or hardly release
metal impurities into solution. Therefore, these metal oxides
are well suited for being incorporated into fixed-bed flow
reactors. When mixed with inductively heatable materials
like nanoparticles 1[6] an ideal setup for continuous oxi-
dations in micro- or miniflow reactors should be achieved
for several reasons: (a) simple setup, (b) only superpara-
magnetic particles are the initial source for heating, which
occurs inside the reactor, (c) metal oxides are heated be-
cause of the close proximity to the heating source which
guarantees ideal heat exchange, (d) nanoparticles can be su-
perheated (above 500 °C!), and (e) metal leaching is kept to
a minimum.[7]
[a] Institut für Organische Chemie und Biomolekulares
Wirkstoffzentrum (BMWZ), Leibniz Universität Hannover,
Schneiderberg 1b, 30167 Hannover, Germany
Fax: +49-511-762-3011
E-mail: andreas.kirschning@oci.uni-hannover.de
[b] PIT – Plasma Induction Technology,
Lorettostraße 20, 40219 Düsseldorf, Germany
4372
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2010, 4372–4375

Facile Oxidations under Flow Conditions
Superparamagnetic materials like nanoparticles 1 can be induced by the inductive field on the efficiency of the pro-
heated in medium- or high-frequency fields. As the techni- cess is exemplified for Equation (2) in Scheme 1. At a flow
cal setup for the former is simpler than for the latter, we rate of 0.05 mL/min at 60 °C only 20% of transformation
carried out electromagnetic induction of heat in magnetic was encountered after one pass, whereas at 90 °C full con-
nanoparticles by applying a medium-frequency field (ca. version occurred. If the flow rate is increased to 0.1 mL/
25 kHz). The heating profile of these nanoparticles in a me- min, 70% conversion was reached at 80 °C.
dium-frequency field is reported in ref.[1]
In our initial studies we chose CrO2 and NiO2 (nickel
peroxide) as representative oxides. Commonly, superpara-
magnetic nanoparticles 1 were mixed with the oxidant in-
side a flow reactor made of glass and encased with an in-
ductor. The dimensions of the reactor are 14 cm length and
8.5 mm internal diameter. The inner volume (for the fluid)
of this fixed-bed reactor was determined to be about 4 mL.
In cases when a higher temperature than the boiling point
of the solvent was needed a tailor-made PEEK (polyether-
ether ketone) reactor with the same inner dimensions and
equipped with an inline back-pressure regulator was ap-
plied (see Figure 2, right). Principally, the processes can be
operated in a circular or a continuous mode.[8] However, we
optimized several reaction parameters to achieve full con-
version in one pass: (a) the concentration of reactants/prod-
uct for avoiding solubility problems, (b) the flow rates in
order to achieve full conversion in one pass, and (c) induc-
tion parameters for heating the superparamagnetic fixed
bed without degradation of reactants or products. This al-
lows to carry out oxidations in a continuous fashion until
the oxidant is exhausted.[9] It must be noted that tempera-
ture determination of individual nanoparticles is a very dif-
ficult task. As described in ref.[1], we used an IR-pyrometer
Scheme 1. Continuous (single pass; full conversion) CrO2-pro-
moted flow oxidation with inductive heating (isolated yields); yields
and conditions for batch reactions: 3: toluene, 90 °C, 7 h, 46 %; 5:
toluene, 90 °C, 22 h, 95%; 7: toluene, 90 °C, 7 h, 62 %; 9: toluene,
90 °C, 7 h, 95 %.
for measuring the surface temperature of the fixed bed in-
side the glass reactor.[10] We found that heating of the solu-
tions proceeds very rapidly once they have entered the reac-
tor. The desired temperature was commonly reached within
the first cm of the reactor.[11]
All these reactions were carried out in glass reactors that
cannot be operated at backpressure drops above 40 psi. In
order to circumvent these limitations, we employed our
PEEK reactor that is solvent-resistant and able to with-
stand backpressures well above 100 psi at higher tempera-
tures (up to 250 °C). These reactors can be operated when
substrates like benzyl alcohol 10 or benzyl ether 12 could
only rapidly (single pass under flow conditions) be oxidized
at higher temperature, typically well above the boiling point
of the solvent. Both reactions were performed in 0.15 
solutions at 100 psi back pressure (Scheme 2).
Noteworthy, when the reaction mixtures were inductively
heated in the presence of the oxidant MagTrieveTM above
150 °C in toluene, we observed some oxidation of the sol-
Figure 2. Flow reactor filled with magnetic nanoparticles and solid
oxidant (left) and PEEK reactor encased with an inductor (right).
First, we oxidized different alcohols such as the proparg- vent and formation of benzoic acid (ca. 0.04% with respect
ylic alcohol 6 and the benzyl alcohol 8 to the correspond- to toluene). For the oxidation of alcohol 10 to ketone 11
ing aldehyde 7 and ketone 9, respectively, using CrO2 as acetonitrile had to be used as solvent because of solubility
solid oxidant (Scheme 1). It was also possible to readily ox- problems.
idize the steroidal alcohol 4 or to directly transform weakly
At this point it needs to be stressed that despite the fact
activated C–H bonds as is shown for the oxidation of an- that CrO2 is paramagnetic and can be removed with a mag-
thracene (2) to anthraquinone (3). The reactions were car- net, it cannot be heated in an inductive field because it does
ried out in glass reactors with an inductive frequency of not have conductive properties.
25 kHz. The substrates were commonly dissolved in toluene
Compared to the same reactions conducted in a flask,
(0.15 ). The influence of the flow rate and temperature the workup is highly simplified. In principle it is possible to
Eur. J. Org. Chem. 2010, 4372–4375
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
4373

J. Wegner, S. Ceylan, C. Friese, A. Kirschning
SHORT COMMUNICATION
allyl alcohol 14 was oxidized to the corresponding aldehyde
15 under flow conditions. Importantly, we were also able to
transform benzylamines 16 and 18 into nitriles 17 and 19,
respectively.
In general we observed that, except for substrates 8 and
12, oxidations in the flow mode gave similar or higher yields
compared to batch experiments. The major practical advan-
tage of conducting these oxidations in the flow mode lies in
the better time/yield ratio. Whereas many reactions had to
be conducted for more than 22 h to achieve full conversion
in the flask, the same reaction took 2–7 h operation time
(which correlates with a residence time of 40–80 min) under
flow conditions [Scheme 1, Equation (2); Scheme 3, Equa-
Scheme 2. High-pressure, high-temperature oxidations (single pass)
in flow with PEEK reactor; yields and conditions for batch reac- tion (4)]. Additionally, workup for products collected from
tions: 11: CH3CN, 90 °C, 7 h, 57 %; 13: toluene, 90 °C, 7 h, 51 %.
flow syntheses was highly simplified.
remove the oxidants by a magnet or by filtration from the
batch reaction mixture, which requires a significant amount
of solvent for complete removal of all nanoparticles. In the
Conclusions
case of our fixed-bed flow system a clear solution left the
We disclose the first applications of inductive heating in-
reactor with very low metal contaminations[7] allowing for side flow reactors by employing mixtures of superparamag-
isolation of the product by direct removal of the solvent netic nanoparticles and solid oxides as fixed-bed materials.
under reduced pressure.
This fixed bed shows very good oxidizing properties and
In order to broaden this experimental concept, we ex- can be utilized to reach full conversion in a single pass. In
tended these studies to nickel peroxide as solid oxidant in- selected cases, the flow reactor was operated under high-
side our inductively heated flow reactors (Scheme 3). Thus, pressure/high-temperature conditions by using PEEK reac-
tors. In essence, besides conventional and microwave-initi-
ated heating, exposition of magnetic nanoparticles to an
electromagnetic field is a third and new principal way for
introducing thermal energy into a reactor.[4] This heating
concept can particularly well be applied in micro- and mini-
reactor setups and has potential to be employed in an in-
dustrial context.
Experimental Section
1
General Information: H NMR spectra were recorded at 400 MHz
with a Bruker AVS-400 spectrometer. 13C NMR spectra were re-
corded at 100 MHz with a Bruker AVS-400 spectrometer. Mass
spectra (EI) were obtained at 70 eV with a type VG Autospec appa-
ratus (Micromass) or with a type LCT (ESI) (Micromass). GC–MS
analyses were performed with an HP 5890 GC equipped with an
HP 5972 MS. Analytical thin-layer chromatography was performed
by using precoated silica gel 60 F254 plates (Merck, Darmstadt),
and the spots were visualized with UV light at 254 nm or by stain-
ing with H2SO4/4-methoxybenzaldehyde in ethanol. Flash column
chromatography was performed on Merck silica gel 60 (230–
400 mesh). The inductor was constructed by IFF GmbH (Isman-
ing). Pumps were obtained from CHELONA GmbH (Potsdam).
All reagents were purchased from Aldrich, Acros or ABCR. Super-
paramagnetic nanoparticles 1 were prepared according to ref.[12]
Toluene, acetonitrile and MagTrieveTM were used as received. NiO2
was prepared according to ref.[13] Leaching analytics of metal spe-
Scheme 3. Continuous [single pass; full conversion except for En-
cies (iron, nickel and chromium) were performed by using a SPEC-
try 4 (a,b)] nickel peroxide promoted flow oxidations with inductive
TRO ARCOS inductively coupled plasma-optical emission spec-
heating (isolated yields). (a) Substrates in toluene (0.15 ), induc-
trometer (ICP-OES, Spectro Analytical Instruments GmbH, Ger-
tive frequency 25 kHz, glass reactor. (b) PEEK reactor: 20 dis-
many) and calculated on an average of all conducted reactions. All
solved in toluene (0.125 ), 100 psi backpressure. (c) Yields and
products are known and were identified by comparison with the
data reported in the literature (see ref.[14]). For all oxidations 1 g of
oxidant was inserted into the reactor (CrO2 or NiO2). The precise
conditions for batch reactions: 15: toluene, 90 °C, 2 h, 72 %; 17:
toluene, 90 °C, 6 h, 72 %; 19: toluene, 90 °C, 4 h, 70 %; 21: toluene,
90 °C, 24 h, 42%.
4374
www.eurjoc.org
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2010, 4372–4375

Facile Oxidations under Flow Conditions
[7]
[8]
For all reactions described, leaching values were determined
for Fe, Cr and Ni, respectively, by using ICP-OES analysis:
Average values for Fe: 4.15 ppm; Cr: 1.75 ppm; Ni: 0.15 ppm.
Selected reviews on chemically functionalized flow-through
systems: a) G. Jas, A. Kirschning, Chem. Eur. J. 2003, 9, 5708–
5723; b) A. Kirschning, G. Jas, Top. Curr. Chem. 2004, 242,
209–239; c) I. R. Baxendale, S. V. Ley in New Avenues to Ef-
ficient Chemical Synthesis: Emerging Technologies (Eds.: P. H.
Seeberger, T. Blume), Springer, Berlin, Heidelberg, 2007, pp.
151–185; d) I. R. Baxendale, J. J. Hayward, S. Lanners, S. V.
Ley, C. D. Smith in Microreactors in Organic Synthesis and Ca-
talysis (Eds.: T. Wirth), Wiley-VCH, Weinheim, 2008, chapter
4.2, pp. 84–122; e) B. Ahmed-Omed, J. C. Brandt, T. Wirth,
Org. Biomol. Chem. 2007, 5, 733–740; f) B. P. Mason, K. E.
Price, J. L. Steinbacher, A. R. Bogdan, D. T. MacQuade, Chem.
Rev. 2007, 107, 2300–2318; g) T. Fukuyama, M. T. Rahman,
M. Sato, I. Ryu, Synlett 2008, 151–163; h) J. Yoshida, A. Na-
gaki, T. Yamada, Chem. Eur. J. 2008, 14, 7450–7459; i) X. Y.
Mak, P. Laurino, P. H. Seeberger, Beilstein J. Org. Chem. 2009,
5, 19; j) S. Ceylan, A. Kirschning in Recoverable and Recyclable
Catalysts (Ed.: M. Benaglia), John Wiley & Sons Ltd., Chich-
ester, 2009, chapter 13, pp. 379–411.
When two identical fixed-bed reactors are employed in a paral-
lel setup, the flow can be switched to the second reactor once
the oxidant inside the first flow reactor is exhausted. This reac-
tor can then be reactivated; see also: R. Lee, D. Donald, Tetra-
hedron Lett. 1997, 38, 3857–3860 for MagTrieveTM; K. Naka-
gawa, R. Konaka, T. Nakata, J. Org. Chem. 1962, 27, 1597–
1601 for NiO2. In an integrated flow system the LC-online de-
termination can be used to determine, when the oxidant is
exhausted, and the reactor has to be exchanged.
loading is difficult to estimate because mainly the surface of the
oxidant reacts. Commonly, a large excess of oxidant is employed as
reported in literature examples.[9] All reactions were performed on
a 0.5–1.0 mmolar scale.
Typical Procedure for CrO2 Oxidations: See Scheme 2, Equation
(1). A PEEK reactor was filled with a mixture of nanoparticles 1
and MagTrieveTM (2:1 wt.-%) and incased with the inductor. The
reactor (void volume: 4 mL) was connected to a pump and on the
opposite side to a back-pressure device (100 psi), which led to the
collection vial. The flow device was flushed with acetonitrile, and
the temperature was adjusted to 135 °C. Once the temperature of
the flow (flow rate 0.1 mL/min) had reached constancy, a solution
of alcohol 10 in acetonitrile (0.15 ) was pumped through the sys-
tem (residence time ca. 40 min). The reaction mixture was col-
lected, and, after removal of the solvent under reduced pressure,
the crude product was purified by flash chromatography (silica gel)
to yield ketone 11 in 92% yield.
Typical Procedure for NiO2 Oxidations: The procedure is identical
to the CrO2 oxidation with the exception that NiO2 was used in-
stead of CrO2.
[9]
Acknowledgments
This work was supported by the Fonds der Chemischen Industrie.
We thank Henkel KGaA (Düsseldorf, Germany), Evonik Indus-
tries AG (Essen, Germany) and IFF GmbH (München, Germany)
for financial or technical support. We thank Prof. T. Scheper and
co-workers (Institute of Technical Chemistry, Leibniz University of
Hannover) for designing and constructing PEEK reactors. ICP-
OES analyses for measuring the degree of leaching by T. Klande
(member of the group of Prof. C. Vogt, Institute of Inorganic and
Analytical Chemistry, Leibniz University of Hannover) is gratefully
acknowledged.
[10]
Before the flow system was used, the setup was calibrated with
respect to electric power and the induced temperature in dif-
ferent solvents at different flow rates (see ref.[1]).
[11] It has to be noted that the stream of reactants is rapidly heated
within the first cm of the reactor until the solution reaches the
desired temperature. After that first zone dispersion is kept to
a minimum. Furthermore, the design of the inductor allowed
maximum homogeneity of the magnetic field inside the flow
reactor, which guarantees homogeneous heating conditions all
through the reactor.
[12] a) M. Kröll, M. Pridöhl, G. Zimmermann, Mater. Res. Soc.
Symp. Proc. 2004, 788, L4.3.1–L4.3.6; b) M. R. Zachariah,
M. I. Aquino, R. D. Shull, B. E. Steel, Nanostruct. Mater. 1995,
5, 383–392; c) S. H. Ehrman, S. K. Friedlander, M. R. Zachar-
iah, J. Mater. Res. 1999, 14, 4551–4561.
[1] S. Ceylan, C. Friese, C. Lammel, K. Mazac, A. Kirschning,
Angew. Chem. 2008, 120, 9083–9086; Angew. Chem. Int. Ed.
2008, 47, 8950–8953.
[2] a) A.-H. Lu, E. L. Salabas, F. Schüth, Angew. Chem. 2007, 119,
1242–1266; Angew. Chem. Int. Ed. 2007, 46, 1222–1244; b) Y.-
W. Jun, J. S. Choi, J. Cheon, Chem. Commun. 2007, 1203–1214;
c) X. K. Zhang, Y. F. Li, J. Q. Xiao, E. D. Wetzel, J. Appl. Phys.
2003, 93, 7124–7126.
[3] R. Hiergeist, W. Andrä, N. Buske, R. Hergt, I. Hilger, U. Rich-
ter, W. J. Kaiser, J. Magn. Magn. Mater. 1999, 201, 420–422.
[4] Enabling technologies include techniques such as nonconven-
tional heating, solid-phase assistance and new reactor designs.
They allow carrying out syntheses faster and performing
workup more efficiently. We first reviewed and defined enabling
technologies for organic synthesis in: A. Kirschning, W. Solod-
enko, K. Mennecke, Chem. Eur. J. 2006, 12, 5972–5990.
[5] M. Lukasiewcz, D. Bogdal, J. Pielichowski, Adv. Synth. Catal.
2003, 345, 1260–1272.
[6] Besides magnetic nanoparticles 1, we also determined the heat-
ing profiles of other ferromagnetic materials such as Fe powder
and Fe3O4 but found that MagSilicaTM 1 had the best heating
properties (for details see ref.[1]).
[13] K. Nakagawa, R. Konaka, T. Nakata, J. Org. Chem. 1962, 27,
1597–1601.
[14] a) Androst-4-ene-3,17-dione (5), see: M. L. Almeida, P. Kocov-
sky, J.-E. Bäckvall, J. Org. Chem. 1996, 61, 6587–6590; b) 3-
phenylpropiolaldehyde (7), see: M. Noro, T. Masuda, A. S. Ich-
imura, N. Koga, H. Iwamura, J. Am. Chem. Soc. 1994, 116,
6179; c) 3-phenylpropenal (15), see: A. Gangloff, T. M. Judge,
H. Paul, J. Org. Chem. 1990, 4, 3679–3682; d) 2-methoxybenzo-
nitrile (17), see: R. K. Arvela, N. E. Leadbeater, J. Org. Chem.
2003, 68, 9122–9125; e) 3,4-dimethoxybenzonitrile (19), see: S.
Murahashi, T. Naota, N. J. Nakajima, J. Org. Chem. 1986, 51,
898–901; f) 2-phenylchromen-4-one (21), see: S. Cui, X. Lin, Y.
Wang, Org. Lett. 2006, 8, 3315–3318.
Received: May 4, 2010
Published Online: July 13, 2010
Eur. J. Org. Chem. 2010, 4372–4375
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
4375