(283a) Comparative Study of Indan Oxidation in Microfluidic Reactors

Liquid‒phase
oxidation of hydrocarbons has broad applications in petrochemical production.
But the challenge is to control the conversion and product selectivity of such
complex free radical oxidation processes which is governed by the temperature,
pressure, residence time and local oxygen availability. Low-conversion is
practiced in industry to control the product selectivity. This study shows the
role of temperature and reactor hydrodynamics to manipulate the oxygen
availability to control oxidative conversion and product selectivity. The
experiments were performed in two glass microfluidic reactors of different
dimensions and volumes by using oxygen and indan, a five-member ring
naphthenic-aromatic compound that is more susceptible to form addition (dimer)
product. The microfluidic reactors were purchased from Dolomite having reactor
volumes of 62.5 µL (half circle shape; mixing channel dimension: 85 µm × 220 µm
× 532 mm; reaction channel dimension: 85 µm × 370 µm × 1912 mm) and 1000 µL
(rectangular shape; mixing channel dimension: 1240 µm × 161 µm × 536 mm;
reaction channel dimension:1240 µm × 391 µm × 1844 mm). The experiments were
performed at different experimental conditions maintaining slug (Taylor) flow (Figure
1). The image was captured during the experiments to calculate the
hydrodynamic parameters. The oxidized indan was analyzed in a GC-FID to
calculate the conversion and selectivity. The GC-FID was calibrated by using
1-indanol (alcohol of indan), 1-indanone (ketone of indan), 1,2-indandione and 1,3-indandione
model compounds and hexachlorobenzene as internal standard.

Oxidation performed in 62.5 µL reactor
showed consistent liquid slugs (approximately 7 × 10^-4 m) and gas
bubbles (approximately 2.5 × 10^-3 m) for the indan injection rate of
7 µL/ min at 30 psig ( 300 kPa) and different temperatures (100–160 ℃).
The resulted gas-liquid interfacial area, a, was approximately 9 × 10⁴
m²/m³. In contrast, oxidation performed in 1000 µL
reactor showed larger liquid slugs (approximately 3.5 × 10^-2 m) and
gas bubble (approximately 9.3 × 10^-2 m) resulting gas-liquid
interfacial area, a, of approximately 2 × 10⁴ m²/m³.
In both cases, conversion increased with temperature. In case of 62.5 µL
reactor, conversion increased from 1.2 wt/wt % to 15. 2 wt/wt % whereas
conversion was changed from 0.7 wt/wt % to 6.0 wt/wt % for the oxidation
performed in 1000 µL reactor. The ketone-to-alcohol selectivity in primary
oxidation product was also changed with temperature (100 ˗160 ℃)
from 3.3 to 6.9 and from 2.5 to 7.5 for the oxidation performed in 62.5 µL and
1000 µL reactor, respectively.

Indan oxidation was also performed in both
the 62.5 µL and 1000 µL reactors at different indan injection rates (2–10
µL/min) at 30 psig (300 kPa) and 150 ℃ to investigate the effect of oxygen
availability on conversion and selectivity. In both cases, the length of the
liquid slug was increased with the indan injection rates, but the size of the
slug was varied with reactor size. In case of the 62.5 µL reactor, the liquid
slug was increased from 2.9 × 10^-4 m to 8.4 × 10^-4 m
whereas liquid slug size was increased from 1.6 × 10^-3 m to 3.7 × 10^-3
m in case of 1000 µL reactor. Variation of liquid injection rates
resulted a decrease in gas bubble size, from 9.4 × 10^-3 m to 2.6 ×
10^-3 m (62.5 µL reactor) and from 3.2 × 10^-2 m to 8.8 ×
10^-3 m (1000 µL reactor). The variation of liquid slug and gas
bubble size obtained at 2 µL/min indan resulted maximum gas-liquid interfacial
area, a, 7.8 × 10⁵ m²/m³ (in 62.5 µL
reactor) and 1.4 × 10⁵ m²/m³ (in 1000 µL
reactor). The maximum indan conversion achieved at the higher interfacial area,
150 ℃ and 30 psig (300 kPa) was 28 wt/wt % (62.5 µL reactor) and 11.5
wt/wt % (1000 µL reactor). No addition product was noticed at the maximum
oxygen availability in 62.5 µL reactor whereas very high ketone-to-alcohol
(13:1) selectivity in primary oxidation product was obtained in at the maximum
oxygen availability in 1000 µL reactor (Figure 2). It can be
explained by the fact that higher interfacial area was obtained differently in
both reactors. In case of 62.5 µL reactor, the main contributor to get higher
oxygen availability was the smaller reactor volume whereas the length of the
liquid film surrounding the gas bubble was the main contributor to obtain the
higher oxygen availability.

In summary, it was possible to manipulate the oxygen
availability by using microfluidic reactor of different sizes and shapes that
influenced the conversion and product selectivity differently. It showed how
engineering could be used to control the chemistry. The understanding from the
study could be used in design and operation of liquid phase oxidation to
produce fine chemicals and pharmaceuticals.