Optimization of a scalable photochemical reactor for reactions with ...
Optimization of a scalable photochemical reactor for reactions with singlet oxygen Konstantin N. Loponov,1 Joao Lopes,1 Maciej Barlog,2 Ekaterina V. Astrova,3 Andrei V. Malkov,4 Alexei A. Lapkin1,*
1
Department of Chemical Engineering and Biotechnology, University of Cambridge, Cambridge CB2 3RA, United Kingdom 2
School of Chemistry, Glasgow University, Glasgow G12 8QQ, United Kingdom 3
4
Ioffe Physico-Technical Institute, St Petersburg 194021, Russian Federation
Department of Chemistry, University of Loughborough, Loughborough LE11 3TU, United Kingdom
Corresponding Author: *Prof. Alexei Lapkin; Email:[email protected]; Fax: +44 (0) 1223 334796.
Supporting Information
1
Actinometry. N2 was bubbling at 30 mL min-1 through the solution prior to and during each experiment. The solution was then irradiated for certain periods of time during recirculating/pumping through the reactor and aliquots were taken at the end of each period for F measuring optical density D. Intensity of light absorbed by ferrioxalate actinometer, I a (mol L
s-1) was calculated using the following equation: I aF =
1
V2 VΣ dD(t ) Θ(λ )εl V1 VR dt
(1)
Here VR is volume of reactor, VΣ is total volume of reaction mixture, V1 is volume of aliquot, V2=2×10-2 L is volume of solution prepared for analysis of aliquot,
dD(t ) were found as slopes dt
of the linear plots D(t), Θ(λ) is the quantum yield of Fe2+ formation at the emitted wavelength. In the case of recirculating mode VΣ=100 mL continuous flow in microreactor VΣ=VR. The Xe arc lamp and the 250 W MH lamps have continuous spectra from UV to IR region. Therefore, an F average quantum yields should be used for the calculation of I a :
Θ=
1 n j +1 ∑∑ N i Θ(∆λ j ) N j =1 i= j
(2)
Here Θ(∆λj) is the quantum yield of Fe2+ generation under illumination by photons from ∆λj interval, Ni is intensity of the emitted photons in interval ∆λi taken from the emission spectrum of lamp used, and N = ∑ N i ∆λi . Θ(λi) were estimated by interpolation using the original i
Hatchard and Parker data.1
To obtain the amounts of light absorbed inside the photoreactors during oxygenation of allylic substrate, I a (mol L-1 s-1) we have to account for the amounts of light absorbed by actinometer
2
and photosensitizer in the reactor. With taking into account, that Lambert-Beer equation is fulfilled only for each particular wavelength λi:
∑ N (1 − 10 ε ∑ N (1 − 10 ε −
S
( λi ) lc S
)
−
F
( λi ) lc F
)
i
Ia = I
F a
i
i
(3)
i
Here εS(λi) and εF(λi) are molar extinction coefficients of the sensitizer and ferrioxalate for each particular wavelength, cS and cF are concentrations of sensitizer and ferrioxalate, l is pathlength of light in the photoreactor. In the cases of 416 nm LED, 524 nm LED, and 420 nm actinic fluorescent lamp i=1 and λ=416, 524 and 420 nm correspondingly. All the parameters used for the calculation of the absorbed light inside the photoreactors are listed in Table S1.
Table S1. Parameters used for calculation of light absorbance by ferrioxalate actinometer and F TPP, I a / I a .
λ
Θ(λi)
nm
εF
εS
l
cF
cS
L mol-1 L mol-1 cm -1 -1 cm cm
M
M
I a / I aF
Annular reactor 420
1.07
110
399663
0.3
0.006 8.1×10-4 3.2
524
0.58
0.35
10900
0.3
0.15
Xe
-
-
0.15
1.3
416
1.08
140
345000
0.024
0.006
22.2
524
0.58
0.35
10900
0.024
0.15
2.6
Xe
-
-
0.024
0.006
130.0
HID
-
-
0.024
0.006
15.0
Microreactor
27.8
3
Kinetic modeling. The reaction scheme describing the photosensitized production of singlet oxygen and its interaction with other species in solution has been widely presented in the literature (see e.g. Ref 2). A simplified representation of this scheme can be depicted in Fig. S1. The steps considered are listed in Table S2, along with estimates for the respective kinetic constants for the alpha-pinene (substrate) – TPP (sensitizer) – dichloromethane (solvent) systems. Other reactions involving these species were considered negligible.
!
! !! !"
+ S9
S09
T19 !
!! !!
± O2 9
!
!! !∆
+ O29 (- S)9
1O
2*9
S + 3O29
! !!
+ R9
R + 3O29
!!
Product(s)9 9 3O 9 2
+ R9 !!
Fig. S1: Simplified kinetic scheme for the sensitized photo-oxygenation of substrate R by singlet oxygen (1O2*). The ground and triplet states of the sensitizer S are denoted by S0 and T1, respectively. Kinetic steps and associated constants are described in Table S2.
The kinetic expressions describing the consumption of reactant R and oxygen can be derived invoking pseudo-steady state assumptions for 1O2* and T1. Concerning the scheme presented above:
−ℛ = −ℛ =
∆
∆ + + + + ∆ +
Table S2: Estimates for the kinetic constants in the photo-oxygenation of -pinene. Reaction / Ref.
Kinetic constant
Reactive quenching of 1O2* by substrate R to oxidized products RO2 , 3
4.3 × 10 M "# s "#
Physical quenching of 1O2* by solvent (Solv), 4
%
9.5 × 10( s "#
Physical quenching of 1O2* by sensitizer S, 4
1.8 × 10* M "# s "#
Generation of 1O2* (reactive quenching of the sensitizer triplet state T1 by O2), 5
∆
2.1 × 10, M "# s "#
Decay of the sensitizer triplet state T1 to the ground state S0, 6
≈ 5 × 10. s "#
Physical quenching of T1 by O2
≈ 0 M "# s "#
0.88 ± 0.03 mol einst "#
Generation rate of T1 from the sensitizer singlet state (S1), , 7
4
Estimation of diffusion coefficients. According to the Wilke-Chang correlation, the diffusivity (in cmH /s) is given by JK = 7.4 × 10"M
(NO PO )Q/ RO STU.V
.
Alternatively, an estimate using the Hayduk-Minhas expression can be obtained from:
JK =
1.55 × 10"M Z #.H, (WX, \X )Y..Y Y.,H . (WK, \K )Y. H WXY.H( [X
The quantities required to estimate J and J at Z = 293.15 K (20ºC) are given in Table S3.
Table S3. Data required for the estimation of diffusion coefficients. Quantity
Unit
Value
Remarks
Dichloromethane (solvent)
Molar volume, WX
cm( /mol
71.4
Molar volume, WX,
cm( /mol
63.9 − 64.50
cP
0.44
g/mol
84.93
Viscosity, [X Molecular weight, aX Association parameter, X Surface tension, \X
at normal b. p. T = 25ºC; average value used
1 dyn/cm
27.2, 27.8, 28.1 T = 20ºC, average value used
-pinene Molar volume, W
cm( /mol
178.63
at normal b. p. 8
183.7
at normal b. p. 9
Molar volume, W,
cm( /mol
158.2
T = 20ºC 10
Surface tension, \
dyn/cm
27.24
T = 20ºC 10
cm( /mol
14.3
at normal b. p.
Oxygen
Molar volume, W
5
Both correlations yield results in the same order-of-magnitude for -pinene: 2.02 × 10", mH s"# and 2.56 × 10", mH s"#, respectively. Thus, results based on the Wilke-Chang expression are used.
Annular recirculating reactor-lamp-reactor geometries
4
Fig. S2. Annular photoreactor (a) and lamp-reactor geometry for (b)-75 W Xe arc lamp (1-lamp, 2-lens, 3-illuminated zone); (c)- 1.44 W 524 nm LED array cross-section (1-10×24 LEDs per strip, illuminated length is 15 cm) (d)-48 W Actinic fluorescent 420 nm (1-U-tube bulb ×2, 24 W each, illuminated length is 23 cm), 2-annular space of reactor, 3-heat transfer fluid, 4-reflector.
Microreactor drawings
6
(a) 12 13 11 10
9 8’ O2
9
(b)
10
1
2
Fig. S3. Microreactor rig (a) scheme, (b) photo. 1-anodically bonded glass-to silicon microreactor with 5×9 LED array, 2-heat exchanger, 3-reservoir, 4-magnetic stirrer, 5-ethylene glycol cooled condenser (connected to thermostat), 6-HPLC pump, 7-thermostat (coolant water), 8 and 8’-O2 metering valves, 9-inlet, 10-outlet, 11-switch valve, 12-back pressure regulator, 13relief valve.
7
Fig. S4. Crossections of heat exhchanger, (a)-vertical, (b)-horizontal. 1-LED array, 2microreactor, 3-microreactor inlet, 4-microreactor outlet, 5-heat exchanger outlets, 6-heat exchanger inlet, 7-groove with o-ring, 8-stainless steel sheet, 9-assembly holes, 10-base.
Fig. S5. Photo of the silicon-glass microreactor unit.
8
REFERENCES 1. Hatchard, C. G.; Parker, C. A., A New Sensitive Chemical Actinometer. II. Potassium Ferrioxalate as a Standard Chemical Actinometer. Proc. Roy. Soc. Lond. A 1956, 235 518-536. 2. Wilkinson, F.; Helman, W. P.; Ross, A. B., Quantum Yields for the Photosensitized Formation of the Lowest Electronically Excited Singlet State of Molecular Oxygen in Solution. Journal of Physical and Chemical Reference Data 1993, 22 (1), 113-262. 3. Wilkinson, F.; Helman, W. P.; Ross, A. B., Rate Constants for the Decay and Reactions of the Lowest Electronically Excited Singlet State of Molecular Oxygen in Solution. An Expanded and Revised Compilation. Journal of Physical and Chemical Reference Data 1995, 24 (2), 663-1021. 4. Gollnick, K.; Griesbeck, A., Solvent dependence of singlet oxygen / substrate interactions in ene-reactions, (4+2)- and (2+2)-cycloaddition reactions. Tetrahedron Letters 1984, 25 (7), 725-728. 5. Kearns, D. R., Physical and chemical properties of singlet molecular oxygen. Chemical Reviews 1971, 71 (4), 395-427. 6. Scholz, M.; Dedic, R.; Breitenbach, T.; Hala, J., Singlet oxygen-sensitized delayed fluorescence of common water-soluble photosensitizers. Photochemical & Photobiological Sciences 2013, 12 (10), 1873-1884. 7. Olmsted III, J., Photocalorimetric studies of singlet oxygen reactions. Journal of the American Chemical Society 1980, 102 (1), 66-71. 8. Silva, C. M.; Filho, C. A.; Quadri, M. B.; Macedo, E. A., Binary diffusion coefficients of α-pinene and β-pinene in supercritical carbon dioxide. Journal of Supercritical Fluids 2004, 32 (1-3), 167-175. 9. Mackay, D.; Shiu, W.-Y.; Ma, K.-C.; Lee, S. C., Handbook of Physical-Chemical Properties and Environmental Fate for Organic Chemicals. CRC Press, Taylor & Francis: Boca Raton, FL, 2006. 10. Zhang, H. Z.; Li, Y. Q.; Xia, J. R.; Davidovits, P.; Williams, L. R.; Jayne, J. T.; Kolb, C. E.; Worsnop, D. R., Uptake of Gas-Phase Species by 1-Octanol. 1. Uptake of α-Pinene, γTerpinene, p-Cymene, and 2-Methyl-2-hexanol as a Function of Relative Humidity and Temperature. The Journal of Physical Chemistry A 2003, 107 (33), 6388-6397.
9
1
Department of Chemical Engineering and Biotechnology, University of Cambridge, Cambridge CB2 3RA, United Kingdom 2
School of Chemistry, Glasgow University, Glasgow G12 8QQ, United Kingdom 3
4
Ioffe Physico-Technical Institute, St Petersburg 194021, Russian Federation
Department of Chemistry, University of Loughborough, Loughborough LE11 3TU, United Kingdom
Corresponding Author: *Prof. Alexei Lapkin; Email:[email protected]; Fax: +44 (0) 1223 334796.
Supporting Information
1
Actinometry. N2 was bubbling at 30 mL min-1 through the solution prior to and during each experiment. The solution was then irradiated for certain periods of time during recirculating/pumping through the reactor and aliquots were taken at the end of each period for F measuring optical density D. Intensity of light absorbed by ferrioxalate actinometer, I a (mol L
s-1) was calculated using the following equation: I aF =
1
V2 VΣ dD(t ) Θ(λ )εl V1 VR dt
(1)
Here VR is volume of reactor, VΣ is total volume of reaction mixture, V1 is volume of aliquot, V2=2×10-2 L is volume of solution prepared for analysis of aliquot,
dD(t ) were found as slopes dt
of the linear plots D(t), Θ(λ) is the quantum yield of Fe2+ formation at the emitted wavelength. In the case of recirculating mode VΣ=100 mL continuous flow in microreactor VΣ=VR. The Xe arc lamp and the 250 W MH lamps have continuous spectra from UV to IR region. Therefore, an F average quantum yields should be used for the calculation of I a :
Θ=
1 n j +1 ∑∑ N i Θ(∆λ j ) N j =1 i= j
(2)
Here Θ(∆λj) is the quantum yield of Fe2+ generation under illumination by photons from ∆λj interval, Ni is intensity of the emitted photons in interval ∆λi taken from the emission spectrum of lamp used, and N = ∑ N i ∆λi . Θ(λi) were estimated by interpolation using the original i
Hatchard and Parker data.1
To obtain the amounts of light absorbed inside the photoreactors during oxygenation of allylic substrate, I a (mol L-1 s-1) we have to account for the amounts of light absorbed by actinometer
2
and photosensitizer in the reactor. With taking into account, that Lambert-Beer equation is fulfilled only for each particular wavelength λi:
∑ N (1 − 10 ε ∑ N (1 − 10 ε −
S
( λi ) lc S
)
−
F
( λi ) lc F
)
i
Ia = I
F a
i
i
(3)
i
Here εS(λi) and εF(λi) are molar extinction coefficients of the sensitizer and ferrioxalate for each particular wavelength, cS and cF are concentrations of sensitizer and ferrioxalate, l is pathlength of light in the photoreactor. In the cases of 416 nm LED, 524 nm LED, and 420 nm actinic fluorescent lamp i=1 and λ=416, 524 and 420 nm correspondingly. All the parameters used for the calculation of the absorbed light inside the photoreactors are listed in Table S1.
Table S1. Parameters used for calculation of light absorbance by ferrioxalate actinometer and F TPP, I a / I a .
λ
Θ(λi)
nm
εF
εS
l
cF
cS
L mol-1 L mol-1 cm -1 -1 cm cm
M
M
I a / I aF
Annular reactor 420
1.07
110
399663
0.3
0.006 8.1×10-4 3.2
524
0.58
0.35
10900
0.3
0.15
Xe
-
-
0.15
1.3
416
1.08
140
345000
0.024
0.006
22.2
524
0.58
0.35
10900
0.024
0.15
2.6
Xe
-
-
0.024
0.006
130.0
HID
-
-
0.024
0.006
15.0
Microreactor
27.8
3
Kinetic modeling. The reaction scheme describing the photosensitized production of singlet oxygen and its interaction with other species in solution has been widely presented in the literature (see e.g. Ref 2). A simplified representation of this scheme can be depicted in Fig. S1. The steps considered are listed in Table S2, along with estimates for the respective kinetic constants for the alpha-pinene (substrate) – TPP (sensitizer) – dichloromethane (solvent) systems. Other reactions involving these species were considered negligible.
!
! !! !"
+ S9
S09
T19 !
!! !!
± O2 9
!
!! !∆
+ O29 (- S)9
1O
2*9
S + 3O29
! !!
+ R9
R + 3O29
!!
Product(s)9 9 3O 9 2
+ R9 !!
Fig. S1: Simplified kinetic scheme for the sensitized photo-oxygenation of substrate R by singlet oxygen (1O2*). The ground and triplet states of the sensitizer S are denoted by S0 and T1, respectively. Kinetic steps and associated constants are described in Table S2.
The kinetic expressions describing the consumption of reactant R and oxygen can be derived invoking pseudo-steady state assumptions for 1O2* and T1. Concerning the scheme presented above:
−ℛ = −ℛ =
∆
∆ + + + + ∆ +
Table S2: Estimates for the kinetic constants in the photo-oxygenation of -pinene. Reaction / Ref.
Kinetic constant
Reactive quenching of 1O2* by substrate R to oxidized products RO2 , 3
4.3 × 10 M "# s "#
Physical quenching of 1O2* by solvent (Solv), 4
%
9.5 × 10( s "#
Physical quenching of 1O2* by sensitizer S, 4
1.8 × 10* M "# s "#
Generation of 1O2* (reactive quenching of the sensitizer triplet state T1 by O2), 5
∆
2.1 × 10, M "# s "#
Decay of the sensitizer triplet state T1 to the ground state S0, 6
≈ 5 × 10. s "#
Physical quenching of T1 by O2
≈ 0 M "# s "#
0.88 ± 0.03 mol einst "#
Generation rate of T1 from the sensitizer singlet state (S1), , 7
4
Estimation of diffusion coefficients. According to the Wilke-Chang correlation, the diffusivity (in cmH /s) is given by JK = 7.4 × 10"M
(NO PO )Q/ RO STU.V
.
Alternatively, an estimate using the Hayduk-Minhas expression can be obtained from:
JK =
1.55 × 10"M Z #.H, (WX, \X )Y..Y Y.,H . (WK, \K )Y. H WXY.H( [X
The quantities required to estimate J and J at Z = 293.15 K (20ºC) are given in Table S3.
Table S3. Data required for the estimation of diffusion coefficients. Quantity
Unit
Value
Remarks
Dichloromethane (solvent)
Molar volume, WX
cm( /mol
71.4
Molar volume, WX,
cm( /mol
63.9 − 64.50
cP
0.44
g/mol
84.93
Viscosity, [X Molecular weight, aX Association parameter, X Surface tension, \X
at normal b. p. T = 25ºC; average value used
1 dyn/cm
27.2, 27.8, 28.1 T = 20ºC, average value used
-pinene Molar volume, W
cm( /mol
178.63
at normal b. p. 8
183.7
at normal b. p. 9
Molar volume, W,
cm( /mol
158.2
T = 20ºC 10
Surface tension, \
dyn/cm
27.24
T = 20ºC 10
cm( /mol
14.3
at normal b. p.
Oxygen
Molar volume, W
5
Both correlations yield results in the same order-of-magnitude for -pinene: 2.02 × 10", mH s"# and 2.56 × 10", mH s"#, respectively. Thus, results based on the Wilke-Chang expression are used.
Annular recirculating reactor-lamp-reactor geometries
4
Fig. S2. Annular photoreactor (a) and lamp-reactor geometry for (b)-75 W Xe arc lamp (1-lamp, 2-lens, 3-illuminated zone); (c)- 1.44 W 524 nm LED array cross-section (1-10×24 LEDs per strip, illuminated length is 15 cm) (d)-48 W Actinic fluorescent 420 nm (1-U-tube bulb ×2, 24 W each, illuminated length is 23 cm), 2-annular space of reactor, 3-heat transfer fluid, 4-reflector.
Microreactor drawings
6
(a) 12 13 11 10
9 8’ O2
9
(b)
10
1
2
Fig. S3. Microreactor rig (a) scheme, (b) photo. 1-anodically bonded glass-to silicon microreactor with 5×9 LED array, 2-heat exchanger, 3-reservoir, 4-magnetic stirrer, 5-ethylene glycol cooled condenser (connected to thermostat), 6-HPLC pump, 7-thermostat (coolant water), 8 and 8’-O2 metering valves, 9-inlet, 10-outlet, 11-switch valve, 12-back pressure regulator, 13relief valve.
7
Fig. S4. Crossections of heat exhchanger, (a)-vertical, (b)-horizontal. 1-LED array, 2microreactor, 3-microreactor inlet, 4-microreactor outlet, 5-heat exchanger outlets, 6-heat exchanger inlet, 7-groove with o-ring, 8-stainless steel sheet, 9-assembly holes, 10-base.
Fig. S5. Photo of the silicon-glass microreactor unit.
8
REFERENCES 1. Hatchard, C. G.; Parker, C. A., A New Sensitive Chemical Actinometer. II. Potassium Ferrioxalate as a Standard Chemical Actinometer. Proc. Roy. Soc. Lond. A 1956, 235 518-536. 2. Wilkinson, F.; Helman, W. P.; Ross, A. B., Quantum Yields for the Photosensitized Formation of the Lowest Electronically Excited Singlet State of Molecular Oxygen in Solution. Journal of Physical and Chemical Reference Data 1993, 22 (1), 113-262. 3. Wilkinson, F.; Helman, W. P.; Ross, A. B., Rate Constants for the Decay and Reactions of the Lowest Electronically Excited Singlet State of Molecular Oxygen in Solution. An Expanded and Revised Compilation. Journal of Physical and Chemical Reference Data 1995, 24 (2), 663-1021. 4. Gollnick, K.; Griesbeck, A., Solvent dependence of singlet oxygen / substrate interactions in ene-reactions, (4+2)- and (2+2)-cycloaddition reactions. Tetrahedron Letters 1984, 25 (7), 725-728. 5. Kearns, D. R., Physical and chemical properties of singlet molecular oxygen. Chemical Reviews 1971, 71 (4), 395-427. 6. Scholz, M.; Dedic, R.; Breitenbach, T.; Hala, J., Singlet oxygen-sensitized delayed fluorescence of common water-soluble photosensitizers. Photochemical & Photobiological Sciences 2013, 12 (10), 1873-1884. 7. Olmsted III, J., Photocalorimetric studies of singlet oxygen reactions. Journal of the American Chemical Society 1980, 102 (1), 66-71. 8. Silva, C. M.; Filho, C. A.; Quadri, M. B.; Macedo, E. A., Binary diffusion coefficients of α-pinene and β-pinene in supercritical carbon dioxide. Journal of Supercritical Fluids 2004, 32 (1-3), 167-175. 9. Mackay, D.; Shiu, W.-Y.; Ma, K.-C.; Lee, S. C., Handbook of Physical-Chemical Properties and Environmental Fate for Organic Chemicals. CRC Press, Taylor & Francis: Boca Raton, FL, 2006. 10. Zhang, H. Z.; Li, Y. Q.; Xia, J. R.; Davidovits, P.; Williams, L. R.; Jayne, J. T.; Kolb, C. E.; Worsnop, D. R., Uptake of Gas-Phase Species by 1-Octanol. 1. Uptake of α-Pinene, γTerpinene, p-Cymene, and 2-Methyl-2-hexanol as a Function of Relative Humidity and Temperature. The Journal of Physical Chemistry A 2003, 107 (33), 6388-6397.
9
