記住我
Oxidation plays a key role in synthesizing highly functionalized molecules . While Jones oxidation and oxidation using KMnO4 are classical and powerful methods, their harsh and hazardous conditions impede their application to complex molecules. Thus, a variety of mild and chemoselective oxidations have been developed, including Swern oxidation , tetrapropylammonium perruthenate (TPAP) oxidation , Pinnick oxidation , and 2,2,6,6-tetramethylpiperidinyl-1-oxyl (TEMPO) oxidation . TEMPO oxidation, in particular, has been successfully applied on a manufacturing scale as a low-cost and green oxidation method . However, these oxidation processes generally require stoichiometric oxidants, and reduced byproducts must be purged in a purification step, which diminishes the atom economy .
To overcome this limitation, the use of molecular oxygen (O2) present in air as an oxidant is one of the ideal solutions . The reduction of O2 generates only water as a byproduct, leading to high atom-economy processes. However, the use of O2 as an oxidant has safety risks due to its potential for explosions when employed on a manufacturing scale . These risks are due to the presence of two out of the three elements of combustion, namely combustibles, oxygen supply and an ignition source , and unexpected ignition caused by static electricity from operators or the equipment can have disastrous consequences. Because the large headspace of batch reactor aggravates these safety risks, the use of O2 in batch manufacturing is very limited .
Recently, continuous flow synthesis has recently been studied as a way to mitigate the safety risks . A compact and closed system improves the process safety of handling molecular oxygen by eliminating unexpected ignition. The safety advantage stimulates the development of various aerobic oxidation processes under continuous-flow conditions accompanied by dedicated devices such as tube-in-tube reactors or fixed bed reactors .
To maximize this advantage, the gas–liquid biphasic reaction must be controlled under continuous-flow conditions. This reaction requires high mixing efficiency to assure high mass transfer of the gas to the liquid phase and a consequent high reaction rate . The mechanism for mixing is categorized mainly as active and passive mixing . Active mixing requires an external force and a driving part. In using O2, active mixing increases the risk of ignition due to friction from the driving part, making it unsuitable for the continuous system to handle aerobic oxidation. Therefore, passive mixing should be more suitable for aerobic oxidation. The whole flow reactor with passive mixing can be immersed in incombustible medium such as water, leading to the improvement of the process safety. Passive mixing is commonly realized using slug-flow or a static mixer . Slug-flow is a simple method for passive mixing, but the formation of slug-flow depends on the tube diameter , and the reaction rate decreases as the tube diameter increases . The gas–liquid biphasic reaction also displays the same characteristic, and a static mixer needs to be developed to enable more robust aerobic oxidation under continuous-flow conditions. A static mixer is generally used by inserting it into a tube reactor. In the gas-liquid biphasic flow reaction, the static mixer has to be inserted into the full range of the tube reactor to maintain high mixing efficiency throughout the reaction. This complex equipment complicates its versatile application for the scale-up and manufacturing, which leads to higher costs.
To address this issue, we turned our attention to the use of porous material, which is widely used for exhaust gas treatment in automobiles . A porous device can improve gas contact efficiency in exhaust gas treatment. Cataler has developed gas–liquid mixing technology using a honeycomb reactor made of porous material (Figure 1) . Narrow channels separated by porous walls allow for high density accumulation in the reactor. As the gas–liquid reaction mixture passes through the honeycomb reactor, the porous material functions as a static mixer, maintaining high mixing efficiency throughout the aerobic oxidation process.
![[1860-5397-19-55-1]](https://www.beilstein-journals.org/bjoc/content/figures/1860-5397-19-55-1.png?scale=2.0&max-width=1024&background=FFFFFF)
Figure 1: Honeycomb reactor. (a) Photograph. (b) Schematic diagram.
We believe that this processing technology offers much potential for application as a flow reactor with a highly efficient and reliable static mixer in its full range. Such a reactor should become one of the ideal devices for aerobic oxidation under continuous-flow conditions and contribute to the further development of aerobic oxidation processes and related oxidation processes such as ozonolysis reactions . Herein, we describe the feasibility of the honeycomb reactor for aerobic oxidation.
Results and Discussion Reaction screening for aerobic oxidation under batch conditionsTo select the representative aerobic oxidation, the reaction conditions to oxidize 4-methoxybenzyl alcohol (1a) were screened (Table 1). The oxidized product p-anisaldehyde (2a) is a valuable substance in food chemistry and a valuable intermediate for synthesizing active pharmaceutical ingredients . This oxidation has been often applied for evaluating catalyst activity for aerobic oxidation , and its screening results can be transferred to obtain a wide variety of benzaldehydes from benzyl alcohols. The screening was conducted under batch conditions. Toward its application to continuous-flow synthesis, we considered the description of the reaction mixture as well as the reaction rate, conversion, yield and availability of the catalysts. Because the honeycomb reactor is made of porous material, the homogeneous reaction solution is a key factor.
Table 1: Reaction screening for aerobic oxidation of 4-methoxybenzyl alcohol (1a).
![[Graphic 1]](https://www.beilstein-journals.org/bjoc/content/inline/1860-5397-19-55-i5.svg?max-width=637&scale=1.0)
Entrya
Catalysts
Solvent
Description
Temp
Time
HPLC (area %)
Conv
(equiv)
(mL/g)
(°C)
(min)
1a
2a
3a
(%)b
1
TEMPO (0.05), Cu(MeCN)4OTf (0.05), 2,2’-bipyridyl (0.05), NMI (0.10)c
MeCN (12)
brown to green solution
25
30
0.0
97.7
0.4
100
2
nor-AZADO (0.01), NaNO2 (0.20)
AcOH (14)
pale yellow solution
25
15
8.9
90.3
0.0
42
60
0.0
99.9
0.0
100
3
TEMPO (0.05), Fe(NO3)3·9H2O (0.05)
AcOH (7)
orange solution
25
60
49.3
50.7
0.0
7
23 h
0.0
98.8
0.5
100
4
TEMPO (0.05), Cu(NO3)2·3H2O (0.075)
AcOH (7)
blue solution
25
60
26.9
64.2
0.0
14
20 h
0.0
98.9
0.4
100
5
TEMPO (0.05), Zn(NO3)2·6H2O (0.075)
AcOH (7)
pale yellow solution
25
60
68.8
29.9
0.0
3
20 h
0.0
98.3
0.7
100
6
PdOAc)2 (0.05), pyridine (0.10)
toluene (7)
pale brown slurry
50
60
24.8
72.8
0.0
17
20 h
4.4
88.4
7.2
59
7
Cu(OAc)2·H2O (0.05), pyridine (0.60)
toluene (7)
blue slurry
50
360
97.7
1.9
0.1
0
8
Ni(OH)2, (0.10)
toluene (7)
green slurry
50
60
98.1
1.8
0.0
0
9
DDQ (0.10), NaNO2 (0.10)
AcOH (7)
black slurry
25
60
1.1
95.9
0.0
86
360
0.0
97.0
0.0
100
10
DDQ (0.05), Fe(NO3)3·9H2O (0.05)
AcOH (7)
black solution
25
60
5.5
92.7
0.0
54
360
0.6
97.9
0.1
92
a1.0 mmol of 4-methoxybenxyl alcohol (1a) was used. The reaction was conducted under open air in EYELA ChemiStation (PPS-1511) with a cross-shaped stirring bar. bConv (%) = 2a (area %)/(2a (area %) + 1a (area %) × 14.083) × 100. 14.083: relative sensitivity coefficient on HPLC (factor). cNMI = N-methylimidazole.
Stahl and Steves have developed a highly reactive aerobic oxidation . This promising methodology enables completion of the reaction in 30 min at room temperature (Table 1, entry 1). However, four kinds of catalysts were used, and a simpler catalytic system would be preferable. The highly reactive catalyst, 9-azanoradamantane N-oxyl (nor-AZADO), was tried with NaNO2 as a cocatalyst, which resulted in completion of the reaction in 60 min (Table 1, entry 2) . While this led to a simpler catalyst system, nor-AZADO is expensive. Hong and co-workers have developed a low-cost catalyst system using TEMPO and nitrate salts . Fe(NO3)3 (Table 1, entry 3), Cu(NO3)2 (Table 1, entry 4), Zn(NO3)2 (Table 1, entry 5) worked as catalysts combined with TEMPO. The reactivities were significantly lower than those in Table 1, entries 1 and 2, but the reaction could be completed overnight at room temperature. The catalysts were completely dissolved in AcOH and the reaction mixture remained in the solution throughout the reaction in entries 3–5. From the viewpoint of application to pharmaceutical manufacturing, the residual amount of copper must be controlled according to ICH Q3D . Iron and zinc have low toxicity and are not listed in ICH Q3D. In comparison with the initial reaction rate of 60 min, Fe(NO3)3/TEMPO in Table 1, entry 3 shows high potential for further optimization. Aerobic oxidation using transition metals instead of TEMPO was also investigated. Pd(OAc)2 (Table 1, entry 6) and Cu(OAc)2 (Table 1, entry 7) , and Ni(OH)2 (Table 1, entry 8) left the starting material 1a. Pd(OAc)2 led to moderate conversion, but Pd(OAc)2 did not dissolve in toluene even with pyridine. As a substitute for TEMPO, 2,3-dichloro-5,6-dicyano-p-benzoquinone (DDQ) was tried (Table 1, entries 9 and 10) . Although the reactivity was improved compared with the TEMPO catalytic system in Table 1, entries 3–5, the DDQ catalytic system made it difficult to confirm the solubility due to the deep black color of the reaction mixture. This can increase the risk of clogging under continuous-flow conditions due to undissolved catalyst.
From this reaction screening, we concluded that Table 1, entry 3 was the most suitable catalytic system for low-cost and environmentally friendly aerobic oxidation and further optimized it to improve the reaction rate.
Reaction optimization for aerobic oxidation under batch conditionsBased on entry 3 in Table 1, we next focused on optimizing the reaction to increase the reaction rate (Table 2). Open air was switched to an O2 balloon to increase the partial pressure of O2 (Table 2, entry 1 vs 2 and 3 vs 4). At room temperature, the reaction rate did not increase although the partial pressure of O2 increased approximately five times with this change. In contrast, at 60 °C, the reaction rate improved greatly with the O2 balloon, and the reaction was completed in 20 min. The proposed catalytic cycle for aerobic oxidation is shown in Scheme 1 . At room temperature, the solubility of O2 in the reaction solution is relatively high, and the catalytic cycle A, which is associated with the solubility of O2, is not the rate-determining step. Therefore, the reaction rates under open air and O2 balloon did not differ. On the other hand, increasing the temperature decreased the solubility of O2, which converted the catalytic cycle A to the rate-determining step at 60 °C. When the amounts of catalysts were decreased to 0.02 equiv (Table 2, entry 5), the reaction completion took 60 min. When the reaction solution was heated to 80 °C, the reaction was completed within 20 min with 0.02 equiv of Fe(NO3)3/TEMPO (Table 2, entry 6). When the amounts of catalysts were decreased to 0.01 equiv at 80°C, the reaction time was significantly prolonged, which is not suitable for the flow synthesis (Table 2, entry 7). Based on these findings, entry 6 was considered to be the optimal reaction conditions, and the application to flow synthesis using the honeycomb reactor was tried under these conditions.
Table 2: Reaction optimization for aerobic oxidation of 4-methoxybenzyl alcohol (1a).
![[Graphic 2]](https://www.beilstein-journals.org/bjoc/content/inline/1860-5397-19-55-i6.svg?max-width=637&scale=1.0)
Entrya
Catalysts
Solvent
Oxidant
Temp
Time
HPLC (area %)
Conv
Fe(NO3)3·9H2O/TEMPO (equiv)
(mL/g)
(°C)
(min)
1a
2a
3a
(%)b
1
0.05/0.05
AcOH (7)
open air
25
30
65.1
34.9
0.0
4
60
49.3
50.7
0.0
7
150
23.8
75.6
0.0
18
23 h
0.0
98.8
0.5
100
2
0.05/0.05
AcOH (7)
O2 balloon
25
60
53.9
44.6
0.0
6
180
16.4
82.3
0.0
26
3
0.05/0.05
AcOH (7)
open air
60
10
7.5
90.9
0.0
46
20
2.6
96.1
0.0
72
30
0.4
98.2
0.0
95
40
0.0
98.4
0.2
100
4
0.05/0.05
AcOH (7)
O2 balloon
60
10
0.5
97.5
0.0
93
20
0.0
98.5
0.6
100
5
0.02/0.02
AcOH (3)
O2 balloon
60
10
33.3
62.4
0.0
12
20
16.3
81.0
0.0
26
30
8.5
86.5
0.0
42
留言 (0)