Fullerene-promoted singlet-oxygen photochemical oxygenations in glass-polymer microstructured reactors

Article abstract of DOI:10.1002/adsc.200800459

In this paper we report the fabrication of thiolene-based microstructured reactors (MRs) that have been specifically designed to include solid-supported reagents within the microchannels network. We propose a convenient solution to realize reversible press-fit, leak-proof interconnects that greatly simplify the MR coupling to the external environment such as capillary tubing, sample reservoirs and pumps. The MRs have been used to carry out the oxidation of α-terpinene and methionine using [60]fullerene, covalently linked to Tentagel and silica gel matrices, as a singlet oxygen sensitizer. High conversions have been observed for both substrates although, in the case of α-terpinene, a partial photodegradation of the endo-peroxide product was detected. Interestingly, in the case of methionine, a quantitative conversion to the corresponding sulfoxides was achieved in about 40 seconds, using low-power, white LED illumination. The reaction time is considerably shorter when compared to the batch procedure that requires, for the same process, about one hour illumination and the use of a 300-W tungsten halogen lamp.

Full text of DOI:10.1002/adsc.200800459

FULL PAPERS
DOI: 10.1002/adsc.200800459
Fullerene-Promoted Singlet-Oxygen Photochemical Oxygenations
in Glass-Polymer Microstructured Reactors
a,
a
a,
Tommaso Carofiglio, * Paola Donnola, Michele Maggini, *
a
a
Massimiliano Rossetto, and Emiliano Rossi
Department of Chemical Sciences and ITM-CNR, University of Padua, Via Marzolo 1, 35131 Padua, Italy
a
Fax : (+39)-(0)49-826-5239; e-mail: tommaso.carofiglio@unipd.it
Received: July 24, 2008; Published online: November 19, 2008
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/adsc.200800459.
Abstract: In this paper we report the fabrication of strates although, in the case of a-terpinene, a partial
thiolene-based microstructured reactors (MRs) that photodegradation of the endo-peroxide product was
have been specifically designed to include solid-sup- detected. Interestingly, in the case of methionine, a
ported reagents within the microchannels network. quantitative conversion to the corresponding sulfox-
We propose a convenient solution to realize reversi- ides was achieved in about 40 seconds, using low-
ble press-fit, leak-proof interconnects that greatly power, white LED illumination. The reaction time is
simplify the MR coupling to the external environ- considerably shorter when compared to the batch
ment such as capillary tubing, sample reservoirs and procedure that requires, for the same process, about
pumps. The MRs have been used to carry out the ox- one hour illumination and the use of a 300-W tung-
idation of a-terpinene and methionine ꢀusing sten halogen lamp.
[
60]fullerene, covalently linked to Tentagel and
silica gel matrices, as a singlet oxygen sensitizer. Keywords: fullerene; microreactors; photooxidation;
High conversions have been observed for both sub- singlet oxygen
Introduction
terpinene in an MR, with Rose Bengal as a photosen-
sitizer. A cycloaddition of singlet oxygen to cyclopen-
[
6]
Organic photochemistry is playing an important role tadiene was studied by Jꢁhnisch and co-workers in a
[
7]
in organic synthesis as a green chemical processing falling-film MR. Fukuyama described the [2+2]
[
1,2]
technology.
In this connection, microstructured re- photocycloaddition of various cyclohexenones with
[3]
[8]
actors (MRs) represent an interesting alternative to vinyl acetates in a glass microflow system. Meyer re-
conventional photoreactors because the small size of ported the photosensitized oxidation of citronellol in
the fluidic channels leads to higher spatial illumina- an MR with light-emitting diodes (LED) as a source
tion homogeneity and better light penetration through of light.
the entire reactor depth. Also, the elevated surface-
In this paper we report the fabrication of polymer
to-volume ratio features of MRs allow an effective re- MRs and their application for heterogeneous singlet
agent mixing, enhanced heat- and mass-transfer rates oxygen photooxidations of model substrates, such as
and safer processing of toxic or highly reactive com- a-terpinene and methionine. The MRs have been
pounds. As an additional benefit, the chemical pro- properly designed to host, within the microchannels
cess under investigation can be scaled up either by network, the photosensitizer [60]fullerene that has
running the MR for an extended time, or by replicat- been immobilized on polystyrene beads or silica gel
ing the MR unit.
microparticles. The fabrication of MRs was achieved
Despite the mentioned advantages, only a few re- through a rapid prototyping technique that makes use
[
9]
ports on photochemical reactions in MRs have been of a thiolene polymer matrix and can be easily im-
[
4]
recorded in the literature. Jensen described the ap- plemented in the normal chemical laboratory, without
plication of MRs to synthesize pinacole from benzo- the need for expensive or dedicated lithographic fa-
[
5]
phenone and isopropyl alcohol. De Mello reported cilities. Since packed-solid MRs that use molecular
the singlet oxygen homogeneous photooxidation of a- oxygen can be subjected to back pressures, we de-
Adv. Synth. Catal. 2008, 350, 2815 – 2822
ꢂ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2815

Tommaso Carofiglio et al.
FULL PAPERS
signed effective female ports for reversible press-fit
The inclusion of a solid-supported [60]fullerene
interconnects and leak-proof operations also in the photosensitizer for singlet oxygen generation has
presence of high pressure drops.
been realized in a chamber within the microfluidic
network. Figure 2 illustrates a three-chambers (E, H,
Results and Discussion
Construction of Microreactors
The MRs used in this work have been fabricated
[
9]
through a procedure, introduced by Beers and co-
workers, that exploits commercial thiolene optical ad-
hesives. Accordingly, a layer of thiolene, acting as a
negative resist, was poured between two glass plates.
Upon partial UV light illumination through a mask
containing the desired microfluidic network, the adhe-
sive, in the regions not shielded by the mask, under-
goes cross-linking and solidifies. Conversely, the fluid,
uncured adhesive can be flushed away with ethanol or
acetone through proper interconnects, leaving the mi-
crofluidic channels behind. We modified Beerꢃs proce-
Figure 2. Microfluidic path for the photochemical MR.
Oxygen inlet (A), substrate solution inlet (B), oxygenation
zone (C), micro-chamber zone (D), photocatalyst feed ports
(
E, H, M), outlet/connection ports (F, G, I, L, N), paper
filter restrains (F1, F2, F3).
[9]
dure and introduced a new type of interconnects
that greatly simplify the coupling of the MR to the
external environment (e.g., capillary tubing, sample M) MR with filters (white stripes F1–F3) to confine
reservoirs and pumps). Our contribution to the gener- the photosensitizer (see the Supporting Information
al procedure of polymer MRs fabrication is a versatile for further fabrication details). Inlets A and B are
press-fit couple composed by a flanged tube (female used for reagent and oxygen, respectively. The con-
connector) and another smaller Teflon tube (male vergent geometry of the inlets governs the formation
[
5]
connector) with matching inside and outside diame- of oxygen bubbles within the reagent flow. The ser-
ters. Figure 1a shows the flanged tube, integrated in pentine channel (Figure 2, zone C) promotes the oxy-
genation of the reaction medium. The first chamber
(
Figure 2, zone E) is directly connected to the oxy-
genation serpentine whereas the others are not con-
nected for a modular use of the reactor. Therefore, it
is possible to employ one, two or three chambers on
demand, by series connection with Teflon tubing
bridges. In addition, one can collect intermediate sam-
ples of the mixture at the outlet of each chamber for
reaction monitoring purposes. Each chamber has its
own feeding port for an easy photocatalyst loading
and replacement. The whole process, from design to
working device, can be completed within a few hours.
Figure 1. a) Female port design in the MR head that serves
as a base for the microchannels network; b) mask used for
the reactor photo-patterning; c) glass-thiolene reactor with
three female ports: substrate inlet (A), oxygen inlet (B),
product outlet (C).
Photooxidation of a-Terpinene
The oxidation of a-terpinene to the endo-peroxide as-
caridole (Figure 3) has been carried out in the pres-
the MR head, and the male connector interconnect. ence of [60]fullerene which is a highly stable singlet
Figure 1b illustrates the mask that has been used for oxygen photosensitizer with an almost unitary quan-
[
10]
the photo-patterning and a thiolene reactor (Fig- tum yield over a broad range of wavelengths.
ure 1c) built on glass plates that has been employed
for the homogeneous photooxidation of a-terpinene
to the corresponding endo-peroxide ascaridole. Note
the flanged tubes A, B and C that constitute the
female ports. Further details on the MR fabrication
have been reported in the Supporting Information.
Figure 3. Photooxidation of a-terpinene to ascaridole.
2816
asc.wiley-vch.de
ꢂ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 2815 – 2822

Fullerene-Promoted Singlet-Oxygen Photochemical Oxygenations
FULL PAPERS
[a]
Table 1. Photooxidation of a-terpinene catalyzed by [60]fullerene and Tentagel-supported [60]fullerene (TG-C ).
6
0
Entry Photo cat- T [8C] Substrate flow
Oxygen flow
[mLh ]
Correlated residence
time [sec]
a-Terpinene conver- Ascaridole
[g] [g]
À1
À1
[f]
alyst
[mLh ]
sion [%]
[%]
[
c]
1
2
3
4
5
6
7
8
C60
C60
C60
C60
À5
À5
À5
2
1
1
1
1
1
0.5
0.25
5
3
2
0.2
0.7
2
1
0.5
78
38
70
97
45
29
94
97
96
30
40
51
32
25
52
53
53
138
180
138
324
27
[
d]
10
[e]
C60
>50
À5
[b]
TG-C60
TG-C60
TG-C60
À5
54
109
À5
[
a]
Reaction conditions – homogeneous reaction: a solution containing a-terpinene (195 mL, 1.0 mmol), n-tetradecane
130 mL, 0.5 mmol, internal standard for GC analysis) and fullerene (8.8 mg, 0.012 mmol, 1.2% with respect to the sub-
strate) in toluene (10 mL, total volume). Reaction conditions – heterogeneous reaction: a solution containing a-terpinene
256 mL, 1.5 mmol), n-tetradecane (174 mL, 0.7 mmol, internal standard for GC analysis), and toluene (25 mL, total
(
(
volume). The reactor contains 14.3 mg of TG-C60 corresponding to 0.45 mg of fullerene (0.6 mmol).
[
b]
Prepared by reacting Tentagel-NH (200 mg, 0.08 mmol amino groups based on a loading of 0.4 mmol of amino groups
2
for gram of dry resin) with [60]fullerene (6.4 mg, 8.8 mmol, 11% of Tentagel-NH loading) in toluene (5 mL).
Ice/sodium chloride mixture.
Water cooling.
Spontaneous temperature due to light source heating without any cooling control.
Calculated total volume of the reactor: 0.152 mL. Correlated residence time (sec) (CRT)=[reactor volume/(oxygen
flow+substrate flow]ꢄ3600. For hetereogeneous reactions the “reactor free volume” was evaluated as described in the
Experimental Section.
The yields of both residual a-terpinene and ascaridole were determined by GC analysis, using internal standard calibra-
tion.
2
[
[
[
[
c]
d]
e]
f]
[
g]
The reaction was first carried out in toluene under from “nominal” flows rates) is a useful parameter
oxygen gas/solution two-phases system, and dissolved that can be used to optimize product yields, reproduce
60]fullerene, onto the MR reported in Figure 1. Thus, a set of conditions or highlight a general process
[
a toluene solution of a-terpinene, [60]fullerene (1.2 trend. However, care must be taken when dealing
mol%) and n-tetradecane, as an internal standard for with solid-supported reagents whose packing charac-
gas-chromatographic quantitation, was introduced teristics have an important effect on the hydrodynam-
through inlet A, whereas pure oxygen was added ics of the microreactor and, therefore, on CRT. An in-
through inlet B. Owing to the “T” mixer configura- crease of the reaction mixture CRT into the MR from
tion, gas and liquid streams collided head-on generat- 78 (entry 1) to 180 seconds (entry 3), produces an a-
ing a gas microbubble train within the microchannel. terpinene consumption rise from 38% up to 97%.
The chip was layered on an ice/NaCl bath (À58C) However, the yield of ascaridole (30–51%) is less
and irradiated with a 300-W tungsten halogen lamp than the expected, for its known propensity to photo-
[
11]
that was placed 25 cm away from the MR (light inten- chemical degradation.
We noted that raising the
sity=19000 lux). Ascaridole was collected in brown temperature from À58C to 108C, while maintaining
glass vials and analyzed by GC. The oxidation was the same CRT (entries 2 and 4), led to a smaller a-
also carried out either at 108C, by pouring the MR terpinene conversion. This is probably due to a re-
into a water bath, or without any temperature control. duced oxygen solubility in toluene as the temperature
In this latter case the lamp heats the MR to about increases. This was further confirmed by the results of
508C. Results are reported in Table 1 (entries 1–5). It the experiment at T >508C in which an excessive
is worth noting that the contribution of the gaseous heating gave the lowest substrate conversion observed
component to the total flow is hard to estimate and in our study (25%, entry 5).
that the sum of oxygen and substrate flows does not
Table 1 reports also the heterogeneous photooxida-
represent the real residence time of reagents inside tion of a-terpinene that was carried out in the MR il-
the MR. We propose to use the term “correlated resi- lustrated in Figure 2. We used, as solid-supported
dence time” (CRT) as a practical parameter which de- photosensitizer, Tentagelꢀ resin beads that were func-
pends, according to the formula reported in the tionalized with [60]fullerene. The fullerene-Tentagelꢀ
paper, to the total flow value, calculated as the sum hybrid (TG-C , Figure 4) was prepared by nucleo-
6
0
of oxygen and substrate “nominal” flows as set on the philic addition of primary amino groups, that lie at
syringe pumps, being aware that “nominal” flows are, the surface of a commercial amino-Tentagelꢀ resin
in turn, proportional to the real ones. CRT (calculated (TG-NH ), to [60]fullerene (see the legend of Table 1
2
Adv. Synth. Catal. 2008, 350, 2815 – 2822
ꢂ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
2817

Tommaso Carofiglio et al.
FULL PAPERS
1
the thioether to sulfoxide conversion via H NMR
spectroscopy (see Supporting Information). The reac-
tion was first carried out under the irradiation of a
300-W tungsten halogen lamp that was kept 25 cm
away from the MR that, in turn, was layered on ice to
avoid overheating. Tentagel-based [60]fullerene (TG-
C , entries 1–5 in Table 2) or silica gel-based
6
0
Figure 4. Supporting [60]fullerene on Tentagelꢀ beads.
[
60]fullerene (Si-C , entries 6–8 in Table 2) were used
60
as photosensitizers. Inlets A and B of the MR in
We note that Figure 2 were used for methionine solution and
[
12,13]
for details on [60]fullerene loading).
[
60]fullerene polyfunctionalization is very likely to oxygen stream, respectively.
occur as the results of multiple fullerene-amine reac-
tions.
A 5ꢄ3 array of commercial white LED (Figure 6a)
was also employed as an alternative light source to
Entries 6–8 in Table 1 show that a-terpinene con- the classical tungsten halogen lamp. LEDs are small
version at À58C is almost quantitative (94–97%) in in size, cheap and represent a “cold” source of light
the CRT-range explored. Furthermore, the moderate that can be kept into close contact to the MR at room
amount of ascaridole that forms, as a result of some temperature without any overheating hazard. To this
decomposition discussed earlier, is constant (52– purpose, a polycarbonate case was prepared to con-
5
3%). The independence of the ascaridole production tain the MR and the LED array (Figure 6b) that de-
on CRT in the range 27–109 seconds, suggests a pla- livered a light intensity equal to 18000 lux.
teau condition that prompted us to carry out addition- The photoreaction was monitored by collecting ef-
al experiments with higher flow rates and, therefore, fluent portions that were diluted with D O (1:1 v/v)
2
1
shorter CRTs. Unfortunately, this led to enhanced and subjected to H NMR analysis to determine the
back-pressure that made it very difficult to achieve a extent of photooxygenation. The results are collected
stable segmented gas-liquid regime for the collection in Table 2.
of meaningful data.
For TG-C , under the irradiation of the tungsten
60
lamp, an increase of the CRT from 32 (entry 1) to 102
seconds (entry 5) rises the conversion of methionine
Photooxidation of l-Methionine
from 38 to 85%. The SiO -supported fullerene (Si-
2
C ) performs better than TG-C (entries 5–8) be-
6
0
60
As a second model reaction, we studied the singlet cause for a CRT of only 33 seconds a quantitative
oxygen photooxidation in water of l-methionine conversion of methionine was achieved. The oxida-
methyl ester to the corresponding sulfoxide with a tion worked nicely also under LED irradiation (en-
supported [60]fullerene photosensitizer that was graft- tries 9 and 10) since we observed 95% conversion of
ed to Tentagelꢀ beads and on SiO particles, com- methionine for a CRT of 42 seconds.
2
monly used for flash chromatography (Figure 5). We
used a functionalized fulleropyrrolidine-SiO hybrid
2
(
Si-C ), whose synthesis and characterization were Conclusions
60
[10c]
reported earlier,
ling that, for Tentagelꢀ resins, could be an issue in In this paper we have reported the fabrication of a
aqueous media. thiolene-based photo-microreactor specifically de-
to overcome the problem of swel-
The oxidation of l-methionine was carried out in signed for the inclusion, in the microfluidic network,
D O because it turned out very convenient to analyze of a supported photo-catalyst. This MR has been ap-
2
Figure 5. Photooxidation of methionine to methionine sulfoxide.
2818
asc.wiley-vch.de
ꢂ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 2815 – 2822

Fullerene-Promoted Singlet-Oxygen Photochemical Oxygenations
FULL PAPERS
Table 2. Singlet oxygen photooxidation of methionine methyl ester with Tentagelꢀ- and silica-supported [60]fullerene photo-
[a]
sensitizers. <W=3
[f]
[g,h]
Entry Photo cata-
lyst
T
Substrate flow
Oxygen flow
[mLh ]
Correlated residence time
[sec]
Methionine conversion
[%]
À1
À1
[8C] [mLh ]
[b]
[d]
1
2
3
4
5
6
7
8
9
1
TG-C60
TG-C60
TG-C60
TG-C60
TG-C60
Si-C60
Si-C60
Si-C60
Si-C60
Si-C60
0
0
0
0
0
0
0
1
1.5
1.5
1
0.8
0.6
1.5
2.5
2.0
2.5
2.0
32
37
48
74
102
21
33
42
30
42
38
40
55
59
85
0.7
0.7
0.3
0.2
0.7
0.7
0.5
1
[c]
83
100
100
89
0 [
25
25
e]
0
0.5
95
[
a]
Reaction conditions: a solution of l-methionine methyl ester hydrochloride (24.1 mg, 0.120 mmol) in D O (4 mL) contain-
2
ing also 2 drops of DCl in D O is loaded into a gastight Hamilton syringe and delivered to the microreactor at the proper
2
rate.
[
b]
Prepared by reacting Tentagel-NH (200 mg, 0.08 mmol amino groups based on a loading of 0.4 mmol of amino groups
2
for gram of dry resin) with fullerene (6.4 mg, 8.8 mmol, 11% of Tentagel-NH loading) in toluene (5 mL).
Prepared according to ref.
The reactor was layered on ice.
Room temperature.
2
[
[
[
[
c]
d]
e]
f]
[10c]
Calculated total volume of the reactor: 0.152 mL. The “reactor free volume” was evaluated as described in the Experi-
mental Section. Correlated residence time (sec) (CRT)=[reactor free volume/(oxygen flow+substrate flow)]ꢄ3600.
[
[
g]
h]
1
Determined by H NMR peak integration.
NMR characterization, that has been carried out on reaction mixture aliquots, showed only methionine ester sulfoxide as
oxidation product. Therefore, the conversion of methionine ester corresponds numerically to the sulfoxide yield.
Figure 6. a) LED 5ꢄ3 array; b) MR of Figure 2 with LED source turned on.
plied to the oxidation of a-terpinene to ascaridole lumination. Probably in a microreactor, the narrow
and of methionine to methionine sulfoxide with mo- channels space favours singlet oxygen diffusion thus
lecular oxygen, using solid-supported [60]fullerene as determining a faster reaction. Interestingly, the con-
singlet oxygen sensitizer. Nearly quantitative conver- version of methionine ester to the corresponding sulf-
sions were achieved for both substrates in about 50 oxide works nicely, in deuterium oxide, under white
and 40 seconds, respectively, although, in the case of LED illumination as well. On the other hand, LED ir-
a-terpinene, the moderate yield in ascaridole can be radiation did not produce any ascaridole in toluene.
ascribed to its sensitivity to photochemical degrada- Radiometric characterization gave a light intensity
tion. The reaction times are considerably shorter equal to 19000 lux and to 18000 lux for the 300-W
when compared to the batch reactor procedure that tungsten halogen lamp and the LED array, respective-
requires, for the conversion of a-terpinene and me- ly. However, the two sources deliver light in a differ-
thionine ester under comparable reaction conditions, ent spectral range. The LED array emits in the visi-
about one hour upon 300-W tungsten halogen lamp il- ble, whereas the tungsten halogen lamp has a substan-
Adv. Synth. Catal. 2008, 350, 2815 – 2822
ꢂ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
2819

Tommaso Carofiglio et al.
FULL PAPERS
tial emission also in the UV region. We believe that female Luer to female 10–32 adapters in conjunction with
1
0–32 fingertight fittings for 1/16 inch OD tubing (Up-
the failure of LED with a-terpinene could be ascribed
to their limited emission range that probably affects
the sensitized production of singlet oxygen, although
church). Proton nuclear magnetic resonance analysis
1
(
H NMR) was carried out on a Bruker AC 250F tuned at
50 MHz. According to the specification sheet available in
2
its lower lifetime in toluene, with respect to D O,
2
the internet (www.cnmic.com) for the white LED used
model. ML50W13H-BED) the emission spectrum is in the
cannot be ruled out. We are currently purchasing a
suitable equipment for an accurate control and meter-
ing of light sources to extend this promising photooxi-
dation study also to other functional substrates.
(
range 440–630 nm with two broad peaks at 480 and 580 nm.
The power supply for the LED-based light source was a
HAMEG instrument (model HM7042-S, Elpav Instruments,
The results reported in this paper were achieved Padua).
with a simple set-up in which the solid photosensitizer
was confined into suitably-designed chambers within
Fabrication of the Reactor Head with Integrated
Female Ports
the MR that was equipped with original female ports
for a quick and easy press-fit connection. The design
principles reported in this work should be applicable The structure of an integrated female port is depicted sche-
matically in Figure 1a. For its construction, access holes
to the inclusion of other supported molecular struc-
were drilled on two identical glass-slides using the corre-
tures, such as catalysts, scavengers or bead-based che-
sponding mask as a reference. One of the two glass-slides
mosensors, in the microfluidic network. As a final
was placed in horizontal position and flanged tubes were in-
remark, we believe that safety issues should make a
serted, flange up, into each of the access holes. Spacers were
microreactor a system to be considered for photo-
glued at the corners of the glass plate using a small amount
chemical oxygenations even if efficiency questions
would be in favour of a batch reactor
of thiolene resin (UV lamp exposition time of about a
minute). Next, the flanged holes were plugged with stoppers
and the adhesive was poured over the glass plate. A second
glass plate was aligned using the stopper tubes as a guide
and carefully layered on the thiolene resin avoiding air bub-
bles formation. The device was allowed to pre-cure beneath
the UV-lamp for two minutes. After removing the stopper
tubes with a tweezer, the reactor head was allowed to post
cure for 20 min. Finally, the device was thermally cured at
Experimental Section
Materials and General Methods
Chemical reagents were purchased from Aldrich and used
as received. The silica supported, fullerene-based sensitizer
5
08C for 12 h to complete the thiolene cross-linking and
[10c]
make the device impervious to organic solvents. The de-
tailed layout of the reactor head construction is reported in
the Supporting Information associated to this paper.
was prepared according to a published procedure.
Optical
adhesive NOA81 was obtained from Norland Products Inc.
A Spectroline SB-100P flood lamp (Spectronics, optimized
for 365 nm) was used for pre- and post-curing the optical ad-
hesive. The lamp was housed inside a home-made dark-box
that allowed us to place the device 60 and 10 cm away from
the lamp during pre-cure (to achieve a degree of beam colli-
mation) and post-cure, respectively. The fluidic path was
generated by using a computer drawing package (Canvas
ver. 9) and printed out in duplicate on transparency film
Fabrication of the Microstructured Reactor
The head described previously served as the bottom layer
for the construction of the photochemical MR. To this pur-
pose, the access holes of the reactor head were plugged with
stopper tubes to prevent contamination with the adhesive.
Using the mask as a reference, pieces of paper filters (about
2ꢄ5ꢄ0.3 mm) were placed in the appropriate locations (F1,
F2, and F3 in Figure 2) and glued to the glass using small
amounts of thiolene adhesive (curing time 1 min). The same
kind of paper filter was also used as spacer and glued to the
corners of the bottom layer. The adhesive was poured over
the glass plate then a second glass plate was carefully lay-
ered on the thiolene resin. The mask was placed on the top
and the assembly and pre-cured for 45 s beneath the UV
lamp at a distance of 60 cm. After removing the stoppers,
the channels were emptied by connecting the device to a
water aspirator, then extensively washed with methanol al-
ternated to short washings with acetone. The device was
dried with a flow of nitrogen. Post curing was carried out by
placing the microreactor under the UV lamp for 20 min at a
distance of 10 cm from the bulb, followed by thermal curing
at 508C for 12 h to attain mechanical stability and to
become impervious to organic solvents. The chambers of the
MR, filled with the sensitizer (vide infra), were connected in
series through small bridges of PTFE tubing. The light
(
KarnaK cod. 2676) using a 1200 dpi laser printer (Epson
EPL-6200). The two masks were aligned on top of each
other and taped together. Access holes (inlets and outlets)
were drilled on glass plates (25ꢄ75ꢄ1 mm or 50ꢄ100ꢄ
1
mm) using a diamond wheel point (Dremel item n. 7103,
bit diameter 2.0 mm, mounted in a variable speed multipro
rotary tool fixed to a 220–01 work station drill press). PTFE
tubings (OD=1/8 inch, ID=1/16 inch, Supelco, item no.
58699) were flanged with a flanging tool (female ports).
PTFE tubing (OD=1/16 inch, ID=1/32 inch, Supelco, item
no. 58701) were used for connecting the MR to sample res-
ervoirs and pumps. During the construction of the reactors,
in some cases it was necessary to prevent the adhesive to
enter into the flanged tubes. In such cases, small pieces of
the connection tubes were filled with optical glue, cured be-
neath the UV lamp, and used as stoppers.
Liquids and gas were delivered to the MR by syringe
pumps (KD-Scientific, Mod. KDS200 and KDS100)
equipped with 5-mL gas-tight Hamilton syringes. The con-
nection between tubing and syringes was realized with
2820
asc.wiley-vch.de
ꢂ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 2815 – 2822

Fullerene-Promoted Singlet-Oxygen Photochemical Oxygenations
source was an array of 15 commercial white LEDs (45 V,
FULL PAPERS
chambers, as illustrated in Figure 2. The solution collected
from the MR was diluted 1/10 with toluene and analyzed by
GC.
30 mA; Figure 6).
Fullerene Immobilization onto Silica Gel
Singlet-Oxygen Photooxidation of a-Terpinene under
Batch-Reactor Conditions
The procedure for covalently linking [60]fullerene onto
silica gel and the characterization of the hybrid singlet
oxygen photosensitizer have been reported previously.
A solution of a-terpinene (195 mL, 1 mmol), n-tetradecane
[10c]
(
130 mL, 0.5 mmol) and pristine [60]fullerene (0.005 mmol) in
The photosensitizer was suspended in toluene and intro-
duced in the MR through the ports (E, H, M in Figure 2). In
order to achieve uniform packing of the catalyst, a slight
vacuum was applied by a water aspirator at the correspond-
ing outlets. During the experiments, the feeding ports were
closed by stoppers. Replacement of the photosensitizer can
be achieved by applying a slight vacuum from the feeding
port and a concomitant flow of solvent through the outlet
port by means of a syringe.
toluene (10 mL) were loaded into a glass photochemical reac-
tor, maintained at 208C with water cooling, and irradiated
with a 300-W tungsten halogen lamp from a distance of
2
0 cm. During irradiation, oxygen was passed through the re-
action mixture via a 0.8 mm ID Teflon tube. 100 mL aliquots
of the reaction mixture were diluted 1:10 (v/v) with toluene
and analyzed by GC-MS to check substrate conversion.
Singlet-Oxygen Photooxidation in an MR of l-
Methionine Methyl Ester
À4
General Procedure for the Evaluation of the “Free
Reactor Volume”
A solution of l-methionine methyl ester (24.1 mg, 1.21ꢄ10
mol) in D O (4.0 mL) containing two drops of concentrated
DCl in D O were loaded into a gas-tight Hamilton syringe
2
2
An estimation of the “free reactor volume” is needed for
the calculation of the CRTs of the reactants inside the MR.
To this purpose, an empty reactor was filled with a solution
of tetraphenylporphyrin at a known concentration in tolu-
ene. The solution was then displaced by using a syringe and
collected in a flask and diluted to a known volume with tol-
uene. The absorbance of the solution was measured at the
Soret band and from the molecular extinction coefficient it
was possible to calculate the total volume of the reactor.
Subsequently, the reactor was filled with the resin and the
above procedure was repeated to find the total volume of
the reactor in the presence of the solid supported photo-cat-
alyst. The subtraction of this value from the previous one
allows an estimate of the volume not occupied by the cata-
lyst and the calculation of the CRT according the formula:
and delivered to the MR (inlet B, Figure 2) at a rate of
0
.5 mL/hour. In the meantime, oxygen was pumped with a
gas-tight Hamilton syringe, connected to inlet A at a rate of
.0 mL/hour. A steady formation of gas bubbles in the aque-
2
ous phase was observed (Figure 2, zone C). After this con-
tacting stage, the oxygenated solution with l-methionine
methyl ester passed throught the three packed chambers
(
each of them containing approssimatively 15 mg of
[
60]fullerene photosensitizer immobilized on silica gel)
under the illuminated zone (Figure 2, zone D). The solution
at the MR outlet (Figure 2, zone N) was collected, diluted
1
1
:1 (v/v) with D O and analyzed by H NMR to determine
2
substrate conversion.
Singlet-Oxygen Photooxidation of Methionine Methyl
Ester under Batch-Reactor Conditions
À4
A solution of l-methionine methyl ester (24 mg, 1.2ꢄ10
mol) in D O (4 mL) containing two drops of DCl in D O
Singlet-Oxygen Photooxidation in an MR of a-
Terpinene with [60]Fullerene under Homogeneous
Conditions
2
2
and [60]fullerene photosensitizer immobilized on silica silica
À6
gel (12.3 mg, 0.87 mg C , 1.2ꢄ10 mol, 1 mol%) were
6
0
charged into a glass photochemical reactor, maintained at
208C with water cooling, and irradiated with a 300-W tung-
sten halogen lamp from a distance of 10 cm. During irradia-
tion, oxygen was passed through the reaction mixture via a
0.8 mm ID Teflon tube. 100 mL aliquots of the reaction mix-
A solution of a-terpinene (195 mL, 1.0 mmol), n-tetradecane
(
130 mL, 0.5 mmol), and pristine [60]fullerene (0.012 mmol)
in toluene (10 mL) was passed through the MR reported in
Figure 2 for the contact times reported in Table 1 (entries 1–
5
1
). The solution that was collected from the MR was diluted
/10 with toluene and analyzed by GC.
ture were diluted 1:5 (v/v) with D O (500 mL) and analyzed
by H NMR to check substrate conversion.
2
1
Supporting Information
Singlet-Oxygen Photooxidation in an MR of a-
Terpinene with [60]Fullerene Supported onto
Tentagelꢀ Beads
1
Details for the reactor fabrication and H NMR monitoring
of methionine oxidation are reported in the Supporting In-
formation.
A solution of a-terpinene (256 mL, 1.5 mmol), n-tetradecane
(
174 mL, 0.7 mmol) in toluene (25 mL) was passed through
the MR reported in Figure 2 for the contact times reported
in Table 1 (entries 6–8). The MR had the fullerene grafted Acknowledgements
to Tentagelꢀ (90 mg of a TG-C60 resin containing 44 mmol
of [60]fullerene for gram of Tentagelꢀ, i.e., 4 mmol of
60]fullerene) which, in turn was placed in the appropriate
We thank Dr. S. Serafini, Dr. A. Castellin (FIS SpA) and Dr.
M. Natali (ICIS-CNR) for stimulating discussions and the
[
Adv. Synth. Catal. 2008, 350, 2815 – 2822
ꢂ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
2821

Tommaso Carofiglio et al.
FULL PAPERS
technical service of the Department of Chemical Sciences for
assistance. MIUR (PRIN 2006034372 and FIRB
RBNE033 KMA), FIS S.p.A., and Fondazione CARIPARO
[7] T. Fukuyama, Y. Hino, N. Kamata, I. Ryu, Chem. Lett.
2004, 33, 1430.
[8] S. Meyer, D. Tietze, S. Rau, B. Schꢁfer, G. Kreisel, J.
Photochem. Photobiol. A: Chemistry 2007, 186, 248.
[9] a) Z. T. Cygan, J. T. Cabral, K. L. Beers, E. J. Amis,
Langmuir 2005, 8, 3629; b) C. Harrison, J. T. Cabral,
C. M. Stafford, A. Karim, E. J. Amis, J. Micromech. Mi-
croeng., 2004, 14, 153.
(
MISCHA project 2008) are gratefully acknowledged for fi-
nancial support.
References
[10] a) M. Bonchio, M. Maggini, E. Menna, G. Scorrano, L.
Garlaschelli, A. Giacometti, F. Paolucci, F. Gasparrini,
D. Misiti, C. Villani, in: Recent Advances in the
Chemistry and Physics of Fullerenes and Related Mate-
rials, (Eds.: K. M. Kadish, R. S. Ruoff), The Electro-
chemical Society, Inc., Pennington, NJ, 1999, p 220;
b) M. Bonchio, M. Carraro, G. Scorrano, A. Bagno,
Adv. Synth. Catal. 2004, 346, 648; c) M. Maggini, G.
Scorrano, M. Prato, G. Brusatin, P. Innocenzi, M. Gu-
glielmi, A. Renier, R. Signorini, M. Meneghetti, R.
Bozio, Advanced Materials 1995, 7, 404.
[
1] P. Tundo, P. Anastas, D. S. Black, J. Breen, T. Collins, S.
Memoli, J. Miyamoto, M. Poliakoff, W. Tumas, Pure
Appl. Chem. 2000, 72, 1207.
[2] M. Oelgemçller, C. Jung, J. Ortner, J. Mattay, E. Zim-
mermann, Green Chem. 2005, 7, 35.
[
3] a) P. Watts, S. J. Haswell, Chem. Soc. Rev. 2005, 34, 235;
b) K. F. Jensen, Chem. Eng. Sci. 2001, 56, 293; c) B.
Ahmed-Omer, J. C. Brandt, T. Wirth, Org. Biomol.
Chem. 2007, 5, 733; d) W. Ehrfeld, V. Hessel, H. Lçwe,
in: Microreactors: New Technology for Modern
Chemistry, Wiley-VCH, Weinheim, 2000.
[11] a) K. K. Maheshwari, P. De Mayo, D. Wiegand, Can. J.
Chem. 1970, 48, 3265; b) C. Karapire, H. Kolancilar, U.
Oyman, S. Icli, J. Photochem. Photobiol. A: Chemistry
2002, 153, 173.
[
[
[
4] H. Lu, M. A. Schmidt, K. F. Jensen, Lab Chip 2001, 1,
2
2.
5] R. C. R. Wooton, R. Fortt, A.J de Mello, Org. Proc.
Res. & Develop. 2002, 6, 187.
6] K. Jꢁhnisch, U. Dingerdissen, Chem. Eng. Technol.
[12] C. Weis, C. Friedrich, R. Mꢅlhaupt, H. Frey, Macromo-
lecules 1995, 28, 403.
[13] S. Samal, B.-J. Choi, K. Geckeler, Chem. Commun.
2000, 1373.
2
005, 28, 426.
2822
asc.wiley-vch.de
ꢂ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 2815 – 2822