- Email: [email protected]
VIS-absorption spectra of photochemically active solutions in continuous flow
Author’s Accepted Manuscript Measurement of UV/VIS-Absorption Spectra of Photochemically Active Solutions in Continuous Flow Ümit Taştan, Johanna Dollinger, Dirk Ziegenbalg www.elsevier.com/locate/flowmeasinst
PII: DOI: Reference:
S0955-5986(17)30354-0 https://doi.org/10.1016/j.flowmeasinst.2017.12.012 JFMI1396
To appear in: Flow Measurement and Instrumentation Received date: 24 August 2017 Revised date: 19 November 2017 Accepted date: 28 December 2017 Cite this article as: Ümit Taştan, Johanna Dollinger and Dirk Ziegenbalg, Measurement of UV/VIS-Absorption Spectra of Photochemically Active Solutions in Continuous Flow, Flow Measurement and Instrumentation, https://doi.org/10.1016/j.flowmeasinst.2017.12.012 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Measurement of UV/VIS-Absorption Spectra of Photochemically Active Solutions in Continuous Flow Ümit Ta¸stana , Johanna Dollingera , Dirk Ziegenbalga a Institute
of Chemical Technology, University of Stuttgart, Stuttgart
Abstract Knowledge of UV/VIS absorption spectra is crucial for designing photochemical processes, but measurements become challenging for systems with fast photoreactions such as photochlorinations. Such challenges include the change in concentration during measurement, evolution of gaseous products, safety concerns as well as the requirement to avoid contact with moisture or oxygen. To overcome these points, a continuous flow method is developed for UV/VIS measurements. The method is applied to the determination of absorption spectra of chlorine dissolved in different organic solvents and the general feasibility is proven. Compared to non-aromatic solvents, the gathered data reveal significant hypsochromic and hyperchromic effects for aromatic solvents. Keywords: photochlorination, absorption spectra, continuous flow, gas-liquid-flow
1. Introduction Interaction of light with matter is the requirement for driving photochemical reactions. From the physical chemistry perspective, this interaction requires an overlap of the emission spectrum of the light source and the absorption spectrum of the reactant. Hence, knowledge of the absorption spectra is the foundation for designing intensified processes also from the reaction engineering point of view.[1–3] While the measurement of absorption spectra is easy for non-reactive species, the determination of absorption coefficients for photochemically active substances can become challenging when the measurements have to be conducted under reactive conditions. This is especially the case for systems where the photoreaction is fast and the photochemically active species is converted, meaning the reaction is non-sensitized and noncatalytic. A prominent example for such reactions is the photochlorination, usually possessing large quantum yields of up to φ = 106 and with this large reaction rates in very pure solvents.[2, 4, 5] Due to the changing concentration, stationary measurement conditions are hard to be achieved. Consequently, it is important to synchronize the measurement of the UV/VIS spectra with the ir∗ Corresponding
author Email address: [email protected] (Dirk Ziegenbalg) Preprint submitted to Flow Measurement and Instrumentation
radiation. It has to be ensured that the initial concentration of the reactive species in the reaction is known and the reaction is not initiated before the actual measurement starts. While this particular requirement can be achieved by triggering the light source, the remaining problems associated with photoreactions such as photochlorinations are not solved. Namely, these problems are the safety issues associated with the handling of chlorine, the requirement to avoid any water in the gas and liquid phase to preclude reaction of chlorine with water and the stoichiometric formation of hydrogen chloride during measurement. Owing to this, hardly any study on the absorption spectrum of chlorine in organic liquids was published till now. The small number of available studies either focuses on inert solvents such as carbon tetrachloride or apply low temperatures to suppress the reaction.[6, 7] Only information on the absorption spectra of chlorine and other halogens in the gas phase are readily available.[8–14] Nevertheless, a literature survey gives evidence for a non-neglectable influence of the nature of the solvent on the absorption characteristics of dissolved chlorine due to electronic interactions with the solvent.[15–18] This renders the measurement of absorption spectra of dissolved chlorine an important building block towards the development of efficient photochemical processes. During the last years, the use of microreactors, often December 30, 2017
referenced as chemistry in flow, attracted much attention owing to the advantages of this technology. Specifically, the small reaction volume, easy handling and the accompanied safety benefits render this technology beneficial for working with chlorine. The continuous operation enables a precise control of the residence and reaction times. Microreactors are also advantageous applied for conducting photochemical reactions.[19–23] The short optical path length allows a defined irradiation of the reactants and avoids shadowed volumes. Furthermore, reaction times can be reduced and selectivities can be increased. Consequently, a continuous flow approach is an attractive option for the investigation of the electronic absorption spectra of chlorine in solution.
varied by adjusting the flow rates. iv) After reaching steady state, the integration time of the spectrometer can be freely chosen. v) The flow rates can be adjusted quickly, minimizing the time required for measuring a single data point. Hence, data acquisition is sped up. vi) It is possible to temper the system to investigate different temperature ranges. vii) The system can be operated under elevated pressure, to further extend the parameter space exploitable. 3. Experimental The flow sheet diagram and a picture of the experimental setup are shown in Figures 1 and 2. For experiments, a certain amount (100 mL) of the solvent, in which the UV/VIS absorption spectra of dissolved chlorine would be measured, was given to a flask. The flask was equipped with a reflux cooler to avoid escape of solvent during degassing. Syringe pumps (neMESYS, CETONI GmbH, Korbussen, Germany) were used to pump the liquids with the desired volumetric flow rates. The UV-VIS measurements were carried out with a UV/VIS spectrometer from Avantes BV, Apeldoorn, The Netherlands (AvaSpec-ULS2048), which was connected to a light source (Avalight-DHS, Avantes BV, Apeldoorn, The Netherlands) and the measuring cell through fiber optical cables (400 µm). The UV-VIS-flow-throughcell used was designed and constructed in-house. Basically, the cell aligns the optical fibers to each other while a capillary was in between the optical fibers. With this, measurements under flow conditions were possible. Consequently, the inner diameter of the capillary was equal to the optical path length. After degassing the solvent with N2 , Cl2 was passed through the solvent. For measuring the volumetric flow rate a rotameter (DK47/PV, KROHNE Messtechnik GmbH) was used, which was especially made for Cl2 and calibrated (20 ◦C, 1 bar) for flows between 20 mL min−1 to 200 mL min−1 . The dissolved Cl2 concentration in the stock solution (c st ) was calculated from the volumetric flow rate of Cl2 (V˙ Cl2 ), the density of gaseous chlorine, the volume of the solvent (V sol ) and the duration of saturation t:
2. Methodology To tackle the challenges mentioned above, a continuous flow analytical methodology was developed. For this, the reactive species was dissolved in the solvent to be investigated to yield a stock solution of defined concentration first. UV/VIS measurements were conducted with a flow-through-cell to make use of the continuous flow approach. The stock solution could be pumped to this cell either with or without dilution by the pure solvent. Mixing of these two liquids was ensured by use of a static mixer. With this, the concentration in the cell could easily be adjusted by changing the ratio of the volumetric flow rates. The solution under investigation was continuously pumped through the UV/VIS-cell. This almost completely ruled out the problems associated with a changing concentration as the mean residence time in the UV/VIS-cell, or with other words the mean reaction time, is very short. Problems resulting from the formation of a gaseous reactant were solved by attaching a capillary to the flow-through-cell, which ended in a flask filled with an inert atmosphere at ambient pressure. The forming gas will lead to an increase in pressure in or right after the cell and consequently presses the reacting solution as well as the formed gasphase out of the capillary. The UV/VIS-measurement is not disturbed by this process since the reaction solution entering the cell is incompressible. The continuous approach allows the use of parts with small inner dimensions. Combining these two characteristics leads to a number of advantages: i) The small internal volume only requires the use of small amounts of substances, reducing experimental costs and improving safety. ii) Handling substances, which are sensitive to moisture or air can be realized easily. iii) Substances with a wide range of absorption coefficients can be measured with the same setup as the concentration can be
c st =
V˙ Cl2 · ρCl2 · t MCl2 · V sol
(1)
The volumetric flow rates were adjusted in such a manner, that no chlorine bubbles left the liquid phase. With that, full absorption of the gas was ensured. To adjust the desired concentration for the measurements, two syringe pumps were used. One pump was filled with the 2
tion solution with 1,2-dichlorobenzene (Sigma Aldrich 99.8 %) as internal standard was prepared. The reaction solution contained 5 vol.% of 1,2-dichlorobenzene in toluene. Chlorine (99.5 %) was purchased from Westfalen AG, Germany. Calculation of the absorption coefficients was based on at least two experimental series with different concentrations of dissolved chlorine. Each experimental series included at least five measurements with different concentrations, which were prepared by dilution with the pure solvent stream. With that, for each solvent at least 20 absorption spectra with different concentrations were recorded from which the absorption coefficients were calculated. To validate the concentration of chlorine in the stock solution, a fraction of the reaction solution was irradiated with a 365 nm LED to completely convert all dissolved chlorine. After irradiation, the concentration of benzyl chloride was determined by GC measurements. As benzyl chloride was the only reaction product, the amount of this substance can be directly correlated to the initial amount of chlorine dissolved. To calculate the concentrations, the gas chromatograph was calibrated prior to the experiments. For measurements, a gas chromatograph (Agilent 7890A) with a HP5 column and a 7693 autosampler was used. A volume of 2 µL was injected with a split ratio of 200:1. The initial oven temperature was set to 60 ◦C. After 8 min the temperature was raised to 110 ◦C with 15 ◦C min−1 . This temperature was kept for 10 min. Following this, the temperature was ramped to 250 ◦C at a heating rate of 25 ◦C min−1 . This temperature was held for 10 min. The two different methods for determining the chlorine concentration were found to be in agreement within ± 7 %. The accuracy of the rotameter used to adjust the volumetric chlorine flow rate was given to be ± 5 % by the manufacturer. The relative standard deviation for single data points was found to be strongly dependent on the solvent and being in the range of 7 % to 15 %. In general, deviations were found to be smaller in aromatic solvents as in non-aromatic solvents.
stock solution, in which chlorine was dissolved, and the other one was filled with pure solvent. The stock solution was pumped with a constant flow rate towards the UV/VIS-cell. By changing the flow rates of the pure solvent, the stock solution could be diluted to different concentrations. For that, solvent and stock solution were mixed in a T-mixer. The experiments were carried out with flow rates of 0.1 mL min−1 to 1 mL min−1 for the stock solution and 1 mL min−1 to 10 mL min−1 for the solvents. For all measurements a perfluoroalkoxy alkane (PFA) capillary with an inner diameter of 1.6 mm was used. The resulting velocities of the mixture were in the range of 8.3 mm s−1 to 91 mm s−1 . The corresponding mean residence times in the light path were ranging from approximately 50 ms to 4 ms. The (molar decadic) absorption coefficients λ were calculated with the Beer-Lambert law: I0 (λ) = λ cd (2) I (λ) where Aλ is the absorbance, c is the concentration, d is the optical path and I0 (λ) and I(λ) are the light intensities before and after passing the capillary. Calculation of the optical light path was done according to the approach used by Russo et al.[24] Because a round capillary was used, not all rays can travel through the middle of the capillary. Therewith, the optical path is shorter at the outside of the irradiated area. For that reason, the average light path was calculated to be d = 1.58 mm by taking into account the curvature of the capillary and the diameter of the optical fiber (for more details see the Supporting Information). To calculate the absorption coefficient, first the absorption spectra of a sample with dissolved chlorine of known concentration was measured. The influence of any external light on the photochemical reaction and thus on the results was eliminated by conducting the measurements in a dark room. The absorption measurements of dissolved chlorine were carried out for seven different solvents. Carbon tetrachloride CCl4 (Sigma Aldrich, 99 %), dichloromethane CH2 Cl2 (Sigma Aldrich, 99.8 %), chloroform CHCl3 (Sigma Aldrich, for analysis), benzene C6 H6 (Sigma Aldrich, for analysis), methylcyclohexane C7 H14 (Sigma Aldrich, for synthesis), p-xylene C8 H10 (Sigma Aldrich, for analysis) and toluene C7 H8 (Sigma Aldrich, HPLC grade 99.8 %) were used as solvents. The solvents were selected with respect to the polarity, stability against side chain photochlorination and the presence of an aromatic ring in the molecular structure. For determining the concentration of chlorine through gas chromatography (GC), a so called reacAλ = lg
4. Results and Discussion The determined absorption spectra of dissolved chlorine in non-aromatic solvents (CCl4 , CH2 Cl2 , CHCl3 , C7 H14 ) can be seen in Figure 3. For comparisons the absorption spectrum for chlorine in the gas phase is shown.[11] Figure 4 shows the results for aromatic solvents (C6 H6 , C8 H10 , C7 H8 ). In both Figures the error bars are only shown exemplarily for one solvent to ensure readability of the diagrams. All spectra can 3
Figure 3: Molar decadic absorption coefficients of chlorine in non aromatic solvents. Error bars are only shown exemplarily for one solvent. Gas phase data was taken from [11].
be found in the Supporting Information with the corresponding error indicators.
Figure 1: Flow sheet diagram of the experimental setup.
For the solvents under investigation, a shift of the absorption spectra as a function of polarity cannot be observed. Similar absorption spectra of dissolved chlorine were found for CCl4 as representative for non-polar solvents and CH2 Cl2 as a representative for polar solvents (see Figure 3). The absorption characteristics of dissolved chlorine in non-aromatic solvents are similar to that of chlorine in the gas phase.[9–12] While the absolute values are slightly smaller, no shift in the absorption maximum is observed. Thus, the absorption maximum of dissolved chlorine as well in the gas phase are found to be at 330 nm. In contrast, a significant hypsochromic shift of the absorption spectra was observed for aromatic organic solvents. Additionally, the absolute values of the absorption coefficients show a hyperchromic effect. The absorption coefficients increase by up to a factor of roughly 30. The observable decrease of the absorption coefficients starting at wavelengths below 280 nm to 290 nm is caused by the absorption of light by the solvent. In literature, this effect is attributed to an interaction of chlorine with the solvent.[18] Thus, the absorption maxima in aromatic solvents are found at approximately 290 nm.[18] The shift of the absorption maximum was interpreted as a charge transfer absorption caused by the formation of a σ- or π-complex of chlorine with the aromatic ring.[25–28] The interaction of the symmetrical orbitals of the aromatic ring with the chlorine atoms causes a shift towards short wavelengths.[29] Formation of such complexes additionally reduces the reactivity of chlorine atoms. As a result, the reaction rate is decreased while the selectivity is increased.[2, 15–17, 30]
Figure 2: Picture of the experimental setup.
4
6. Acknowledgement The authors gratefully acknowledge the financial support provided by the German Research Foundation within the Priority Program “Reactive Bubbly Flows”, SPP 1740 (ZI 1502/1-1). Dirk Ziegenbalg thanks Prof. Dr.-Ing. Elias Klemm (ITC) for continuous support.
References Figure 4: Molar decadic absorption coefficients of chlorine in aromatic solvents. Error bars are only shown exemplarily for one solvent.
[1] Horst Böttcher, Technical Applications of Photochemistry, Deutscher Verlag für Grundstoffindustrie, 1991. [2] A. M. Braun, M.-T. Maurette, E. Oliveros, Photochemical technology, Wiley, 1991. [3] M. Sender, D. Ziegenbalg, Light sources for photochemical processes – estimation of technological potentials, Chemie Ingenieur Technik (2017). [4] C. Lian, S.-Y. HO, Photochemical chlorination of methane, National Science Council, Proceedings 2 (1978) 4–7. [5] Ü. Ta¸stan, D. Ziegenbalg, Getting the most out of solar irradiation: Efficient use of polychromatic light for water splitting, Chemistry - A European Journal 22 (2016) 18824–18832. [6] J. E. Chateauneuf, Direct measurement of the absolute kinetics of chlorine atom in carbon tetrachloride, Journal of the American Chemical Society 112 (1990) 442–444. [7] J. E. Chateauneuf, Charge-transfer absorption spectra of chlorine atoms in halogenated solvents, Chemical Physics Letters 164 (1989) 577–580. [8] R. L. Wilkins, Continuous absorption coefficients of chlorine, Journal of Quantitative Spectroscopy and Radiative Transfer 4 (1964) 775–782. [9] G. E. Gibson, N. S. Bayliss, Variation with temperature of the continuous absorption spectrum of diatomic molecules: Part I. experimental, the absorption spectrum of chlorine, Physical Review 44 (1933) 188–192. [10] F. W. Jones, W. Spooner, Absorption of light by gaseous chlorine in the wavelength region 5040 a.u. to 5320 a.u., Trans. Faraday Soc. 31 (1935) 811–813. [11] D. J. Seery, D. Britton, The continuous absorption spectra of chlorine, bromine, bromine chloride, iodine chloride, and iodine bromide, The Journal of Physical Chemistry 68 (1964) 2263– 2266. [12] D. Maric, J. Burrows, R. Meller, G. Moortgat, A study of the UV—visible absorption spectrum of molecular chlorine, Journal of Photochemistry and Photobiology A: Chemistry 70 (1993) 205–214. [13] D. B. Kokh, A. B. Alekseyev, R. J. Buenker, Theoretical study of the UV photodissociation of Cl2 : Potentials, transition moments, extinction coefficients, and Cl*/Cl branching ratio, The Journal of Chemical Physics 120 (2004) 11549–11556. [14] S. Hubinger, J. Nee, Absorption spectra of Cl2 , Br2 and BrCl between 190 and 600 nm, Journal of Photochemistry and Photobiology A: Chemistry 86 (1995) 1–7. [15] G. A. Russell, Solvent effects in the reactions of free radicals and atoms, Journal of the American Chemical Society 79 (1957) 2977–2978. [16] G. A. Russell, Solvent effects in the reactions of free radicals and atoms. II. Effects of solvents on the position of attack of chlorine atoms upon 2,3-dimethylbutane, isobutane and 2deuterio-2-methylpropane1, Journal of the American Chemical Society 80 (1958) 4987–4996.
5. Conclusions
Developing efficient photochemical processes requires knowledge of basic physical properties of the reaction system. One of the most important properties in the context of photochemistry is the absorption coefficient of the reactants. Measuring the UV/VIS absorption spectra of photochemical active species can be challenging as they react during irradiation. As this is an intrinsic characteristic, analytical methods must be designed to take this into account, especially for systems showing fast reactions. To address these challenges, a continuous flow method was developed and applied to the measurement of absorption spectra of chlorine dissolved in different organic solvents. The results illustrate, that measurements of physical properties are of high relevance for photochlorinations and for the reaction engineering of photochemical processes in general. The absorption spectra of chlorine dissolved in aromatic solvents show significant hypsochromic as well as hyperchromic effects compared to non-aromatic solvents. Although this effect was already described in literature, the data published till now lack a thorough experimental investigation as well as an extension of the scope of solvents. This work fills this gap. While the presented methodology was developed for the determination of the absorption spectra of chlorine in different organic solvents, transferring the approach to other solvents, gases or reactants can be easily realized. This is especially relevant for the investigation of photochemical systems, which suffer from fast reaction rates, hazardous substances, sensitivity to moisture and/or oxygen or the evolving gases. Furthermore, it is possible to apply the methodology to reactions initiated by thermal energy as well. 5
[17] G. A. Russell, Solvent effects in the reactions of free radicals and atoms. III. Effects of solvents in the competitive photochlorination of hydrocarbons and their derivatives, Journal of the American Chemical Society 80 (1958) 4997–5001. [18] L. J. Andrews, R. M. Keefer, The interaction of chlorine with benzene and m-xylene, Journal of the American Chemical Society 73 (1951) 462–463. [19] Y. Su, Straathof, Natan J. W., V. Hessel, T. Noel, Photochemical transformations accelerated in continuous-flow reactors: Basic concepts and applications, Chem. Eur. J. 20 (2014) 10562– 10589. [20] K. A. Angelo Albini, Photochemistry: Volume 43, ROYAL SOCIETY OF CHEMISTRY, 2015. [21] K. Loubière, M. Oelgemöller, T. Aillet, O. Dechy-Cabaret, L. Prat, Continuous-flow photochemistry: A need for chemical engineering, Chemical Engineering and Processing: Process Intensification 104 (2016) 120–132. [22] D. Cambié, C. Bottecchia, N. J. W. Straathof, V. Hessel, T. Noël, Applications of continuous-flow photochemistry in organic synthesis, material science, and water treatment, Chemical Reviews 116 (2016) 10276–10341. [23] J. A. Lummiss, P. D. Morse, R. L. Beingessner, T. F. Jamison, Towards more efficient, greener syntheses through flow chemistry, The Chemical Record 17 (2017) 667–680. [24] D. Russo, D. Spasiano, M. Vaccaro, R. Andreozzi, G. L. Puma, N. M. Reis, R. Marotta, Direct photolysis of benzoylecgonine under UV irradiation at 254nm in a continuous flow microcapillary array photoreactor, Chemical Engineering Journal 283 (2016) 243–250. [25] G. Chiltz, R. Eckling, P. Goldfinger, G. Huynrechts, G. Martens, G. Simoens, The three body recombination of chlorine atoms, Bulletin des Sociétés Chimiques Belges 71 (1962) 747–758. [26] P. Goldfinger, Reactions of halogen atoms, Pure and Applied Chemistry 5 (1962) 423–440. [27] G. Chiltz, P. Goldfinger, G. Huybrechts, G. Martens, G. Verbeke, Atomic chlorination of simple hydrocarbon derivatives in the gas phase, Chemical Reviews 63 (1963) 355–372. [28] P. Skell, H. Baxter, J. Tanko, Reactions of BrCl with alkyl radicals, Tetrahedron Letters 27 (1986) 5181–5184. [29] R. A. Marcus, Relation between charge transfer absorption and fluorescence spectra and the inverted region, The Journal of Physical Chemistry 93 (1989) 3078–3086. [30] A. K. Croft, H. M. Howard-Jones, C. E. Skates, C. C. Wood, Controlling the action of chlorine radical: from lab to environment, Organic & Biomolecular Chemistry 9 (2011) 7439.
6
PII: DOI: Reference:
S0955-5986(17)30354-0 https://doi.org/10.1016/j.flowmeasinst.2017.12.012 JFMI1396
To appear in: Flow Measurement and Instrumentation Received date: 24 August 2017 Revised date: 19 November 2017 Accepted date: 28 December 2017 Cite this article as: Ümit Taştan, Johanna Dollinger and Dirk Ziegenbalg, Measurement of UV/VIS-Absorption Spectra of Photochemically Active Solutions in Continuous Flow, Flow Measurement and Instrumentation, https://doi.org/10.1016/j.flowmeasinst.2017.12.012 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Measurement of UV/VIS-Absorption Spectra of Photochemically Active Solutions in Continuous Flow Ümit Ta¸stana , Johanna Dollingera , Dirk Ziegenbalga a Institute
of Chemical Technology, University of Stuttgart, Stuttgart
Abstract Knowledge of UV/VIS absorption spectra is crucial for designing photochemical processes, but measurements become challenging for systems with fast photoreactions such as photochlorinations. Such challenges include the change in concentration during measurement, evolution of gaseous products, safety concerns as well as the requirement to avoid contact with moisture or oxygen. To overcome these points, a continuous flow method is developed for UV/VIS measurements. The method is applied to the determination of absorption spectra of chlorine dissolved in different organic solvents and the general feasibility is proven. Compared to non-aromatic solvents, the gathered data reveal significant hypsochromic and hyperchromic effects for aromatic solvents. Keywords: photochlorination, absorption spectra, continuous flow, gas-liquid-flow
1. Introduction Interaction of light with matter is the requirement for driving photochemical reactions. From the physical chemistry perspective, this interaction requires an overlap of the emission spectrum of the light source and the absorption spectrum of the reactant. Hence, knowledge of the absorption spectra is the foundation for designing intensified processes also from the reaction engineering point of view.[1–3] While the measurement of absorption spectra is easy for non-reactive species, the determination of absorption coefficients for photochemically active substances can become challenging when the measurements have to be conducted under reactive conditions. This is especially the case for systems where the photoreaction is fast and the photochemically active species is converted, meaning the reaction is non-sensitized and noncatalytic. A prominent example for such reactions is the photochlorination, usually possessing large quantum yields of up to φ = 106 and with this large reaction rates in very pure solvents.[2, 4, 5] Due to the changing concentration, stationary measurement conditions are hard to be achieved. Consequently, it is important to synchronize the measurement of the UV/VIS spectra with the ir∗ Corresponding
author Email address: [email protected] (Dirk Ziegenbalg) Preprint submitted to Flow Measurement and Instrumentation
radiation. It has to be ensured that the initial concentration of the reactive species in the reaction is known and the reaction is not initiated before the actual measurement starts. While this particular requirement can be achieved by triggering the light source, the remaining problems associated with photoreactions such as photochlorinations are not solved. Namely, these problems are the safety issues associated with the handling of chlorine, the requirement to avoid any water in the gas and liquid phase to preclude reaction of chlorine with water and the stoichiometric formation of hydrogen chloride during measurement. Owing to this, hardly any study on the absorption spectrum of chlorine in organic liquids was published till now. The small number of available studies either focuses on inert solvents such as carbon tetrachloride or apply low temperatures to suppress the reaction.[6, 7] Only information on the absorption spectra of chlorine and other halogens in the gas phase are readily available.[8–14] Nevertheless, a literature survey gives evidence for a non-neglectable influence of the nature of the solvent on the absorption characteristics of dissolved chlorine due to electronic interactions with the solvent.[15–18] This renders the measurement of absorption spectra of dissolved chlorine an important building block towards the development of efficient photochemical processes. During the last years, the use of microreactors, often December 30, 2017
referenced as chemistry in flow, attracted much attention owing to the advantages of this technology. Specifically, the small reaction volume, easy handling and the accompanied safety benefits render this technology beneficial for working with chlorine. The continuous operation enables a precise control of the residence and reaction times. Microreactors are also advantageous applied for conducting photochemical reactions.[19–23] The short optical path length allows a defined irradiation of the reactants and avoids shadowed volumes. Furthermore, reaction times can be reduced and selectivities can be increased. Consequently, a continuous flow approach is an attractive option for the investigation of the electronic absorption spectra of chlorine in solution.
varied by adjusting the flow rates. iv) After reaching steady state, the integration time of the spectrometer can be freely chosen. v) The flow rates can be adjusted quickly, minimizing the time required for measuring a single data point. Hence, data acquisition is sped up. vi) It is possible to temper the system to investigate different temperature ranges. vii) The system can be operated under elevated pressure, to further extend the parameter space exploitable. 3. Experimental The flow sheet diagram and a picture of the experimental setup are shown in Figures 1 and 2. For experiments, a certain amount (100 mL) of the solvent, in which the UV/VIS absorption spectra of dissolved chlorine would be measured, was given to a flask. The flask was equipped with a reflux cooler to avoid escape of solvent during degassing. Syringe pumps (neMESYS, CETONI GmbH, Korbussen, Germany) were used to pump the liquids with the desired volumetric flow rates. The UV-VIS measurements were carried out with a UV/VIS spectrometer from Avantes BV, Apeldoorn, The Netherlands (AvaSpec-ULS2048), which was connected to a light source (Avalight-DHS, Avantes BV, Apeldoorn, The Netherlands) and the measuring cell through fiber optical cables (400 µm). The UV-VIS-flow-throughcell used was designed and constructed in-house. Basically, the cell aligns the optical fibers to each other while a capillary was in between the optical fibers. With this, measurements under flow conditions were possible. Consequently, the inner diameter of the capillary was equal to the optical path length. After degassing the solvent with N2 , Cl2 was passed through the solvent. For measuring the volumetric flow rate a rotameter (DK47/PV, KROHNE Messtechnik GmbH) was used, which was especially made for Cl2 and calibrated (20 ◦C, 1 bar) for flows between 20 mL min−1 to 200 mL min−1 . The dissolved Cl2 concentration in the stock solution (c st ) was calculated from the volumetric flow rate of Cl2 (V˙ Cl2 ), the density of gaseous chlorine, the volume of the solvent (V sol ) and the duration of saturation t:
2. Methodology To tackle the challenges mentioned above, a continuous flow analytical methodology was developed. For this, the reactive species was dissolved in the solvent to be investigated to yield a stock solution of defined concentration first. UV/VIS measurements were conducted with a flow-through-cell to make use of the continuous flow approach. The stock solution could be pumped to this cell either with or without dilution by the pure solvent. Mixing of these two liquids was ensured by use of a static mixer. With this, the concentration in the cell could easily be adjusted by changing the ratio of the volumetric flow rates. The solution under investigation was continuously pumped through the UV/VIS-cell. This almost completely ruled out the problems associated with a changing concentration as the mean residence time in the UV/VIS-cell, or with other words the mean reaction time, is very short. Problems resulting from the formation of a gaseous reactant were solved by attaching a capillary to the flow-through-cell, which ended in a flask filled with an inert atmosphere at ambient pressure. The forming gas will lead to an increase in pressure in or right after the cell and consequently presses the reacting solution as well as the formed gasphase out of the capillary. The UV/VIS-measurement is not disturbed by this process since the reaction solution entering the cell is incompressible. The continuous approach allows the use of parts with small inner dimensions. Combining these two characteristics leads to a number of advantages: i) The small internal volume only requires the use of small amounts of substances, reducing experimental costs and improving safety. ii) Handling substances, which are sensitive to moisture or air can be realized easily. iii) Substances with a wide range of absorption coefficients can be measured with the same setup as the concentration can be
c st =
V˙ Cl2 · ρCl2 · t MCl2 · V sol
(1)
The volumetric flow rates were adjusted in such a manner, that no chlorine bubbles left the liquid phase. With that, full absorption of the gas was ensured. To adjust the desired concentration for the measurements, two syringe pumps were used. One pump was filled with the 2
tion solution with 1,2-dichlorobenzene (Sigma Aldrich 99.8 %) as internal standard was prepared. The reaction solution contained 5 vol.% of 1,2-dichlorobenzene in toluene. Chlorine (99.5 %) was purchased from Westfalen AG, Germany. Calculation of the absorption coefficients was based on at least two experimental series with different concentrations of dissolved chlorine. Each experimental series included at least five measurements with different concentrations, which were prepared by dilution with the pure solvent stream. With that, for each solvent at least 20 absorption spectra with different concentrations were recorded from which the absorption coefficients were calculated. To validate the concentration of chlorine in the stock solution, a fraction of the reaction solution was irradiated with a 365 nm LED to completely convert all dissolved chlorine. After irradiation, the concentration of benzyl chloride was determined by GC measurements. As benzyl chloride was the only reaction product, the amount of this substance can be directly correlated to the initial amount of chlorine dissolved. To calculate the concentrations, the gas chromatograph was calibrated prior to the experiments. For measurements, a gas chromatograph (Agilent 7890A) with a HP5 column and a 7693 autosampler was used. A volume of 2 µL was injected with a split ratio of 200:1. The initial oven temperature was set to 60 ◦C. After 8 min the temperature was raised to 110 ◦C with 15 ◦C min−1 . This temperature was kept for 10 min. Following this, the temperature was ramped to 250 ◦C at a heating rate of 25 ◦C min−1 . This temperature was held for 10 min. The two different methods for determining the chlorine concentration were found to be in agreement within ± 7 %. The accuracy of the rotameter used to adjust the volumetric chlorine flow rate was given to be ± 5 % by the manufacturer. The relative standard deviation for single data points was found to be strongly dependent on the solvent and being in the range of 7 % to 15 %. In general, deviations were found to be smaller in aromatic solvents as in non-aromatic solvents.
stock solution, in which chlorine was dissolved, and the other one was filled with pure solvent. The stock solution was pumped with a constant flow rate towards the UV/VIS-cell. By changing the flow rates of the pure solvent, the stock solution could be diluted to different concentrations. For that, solvent and stock solution were mixed in a T-mixer. The experiments were carried out with flow rates of 0.1 mL min−1 to 1 mL min−1 for the stock solution and 1 mL min−1 to 10 mL min−1 for the solvents. For all measurements a perfluoroalkoxy alkane (PFA) capillary with an inner diameter of 1.6 mm was used. The resulting velocities of the mixture were in the range of 8.3 mm s−1 to 91 mm s−1 . The corresponding mean residence times in the light path were ranging from approximately 50 ms to 4 ms. The (molar decadic) absorption coefficients λ were calculated with the Beer-Lambert law: I0 (λ) = λ cd (2) I (λ) where Aλ is the absorbance, c is the concentration, d is the optical path and I0 (λ) and I(λ) are the light intensities before and after passing the capillary. Calculation of the optical light path was done according to the approach used by Russo et al.[24] Because a round capillary was used, not all rays can travel through the middle of the capillary. Therewith, the optical path is shorter at the outside of the irradiated area. For that reason, the average light path was calculated to be d = 1.58 mm by taking into account the curvature of the capillary and the diameter of the optical fiber (for more details see the Supporting Information). To calculate the absorption coefficient, first the absorption spectra of a sample with dissolved chlorine of known concentration was measured. The influence of any external light on the photochemical reaction and thus on the results was eliminated by conducting the measurements in a dark room. The absorption measurements of dissolved chlorine were carried out for seven different solvents. Carbon tetrachloride CCl4 (Sigma Aldrich, 99 %), dichloromethane CH2 Cl2 (Sigma Aldrich, 99.8 %), chloroform CHCl3 (Sigma Aldrich, for analysis), benzene C6 H6 (Sigma Aldrich, for analysis), methylcyclohexane C7 H14 (Sigma Aldrich, for synthesis), p-xylene C8 H10 (Sigma Aldrich, for analysis) and toluene C7 H8 (Sigma Aldrich, HPLC grade 99.8 %) were used as solvents. The solvents were selected with respect to the polarity, stability against side chain photochlorination and the presence of an aromatic ring in the molecular structure. For determining the concentration of chlorine through gas chromatography (GC), a so called reacAλ = lg
4. Results and Discussion The determined absorption spectra of dissolved chlorine in non-aromatic solvents (CCl4 , CH2 Cl2 , CHCl3 , C7 H14 ) can be seen in Figure 3. For comparisons the absorption spectrum for chlorine in the gas phase is shown.[11] Figure 4 shows the results for aromatic solvents (C6 H6 , C8 H10 , C7 H8 ). In both Figures the error bars are only shown exemplarily for one solvent to ensure readability of the diagrams. All spectra can 3
Figure 3: Molar decadic absorption coefficients of chlorine in non aromatic solvents. Error bars are only shown exemplarily for one solvent. Gas phase data was taken from [11].
be found in the Supporting Information with the corresponding error indicators.
Figure 1: Flow sheet diagram of the experimental setup.
For the solvents under investigation, a shift of the absorption spectra as a function of polarity cannot be observed. Similar absorption spectra of dissolved chlorine were found for CCl4 as representative for non-polar solvents and CH2 Cl2 as a representative for polar solvents (see Figure 3). The absorption characteristics of dissolved chlorine in non-aromatic solvents are similar to that of chlorine in the gas phase.[9–12] While the absolute values are slightly smaller, no shift in the absorption maximum is observed. Thus, the absorption maximum of dissolved chlorine as well in the gas phase are found to be at 330 nm. In contrast, a significant hypsochromic shift of the absorption spectra was observed for aromatic organic solvents. Additionally, the absolute values of the absorption coefficients show a hyperchromic effect. The absorption coefficients increase by up to a factor of roughly 30. The observable decrease of the absorption coefficients starting at wavelengths below 280 nm to 290 nm is caused by the absorption of light by the solvent. In literature, this effect is attributed to an interaction of chlorine with the solvent.[18] Thus, the absorption maxima in aromatic solvents are found at approximately 290 nm.[18] The shift of the absorption maximum was interpreted as a charge transfer absorption caused by the formation of a σ- or π-complex of chlorine with the aromatic ring.[25–28] The interaction of the symmetrical orbitals of the aromatic ring with the chlorine atoms causes a shift towards short wavelengths.[29] Formation of such complexes additionally reduces the reactivity of chlorine atoms. As a result, the reaction rate is decreased while the selectivity is increased.[2, 15–17, 30]
Figure 2: Picture of the experimental setup.
4
6. Acknowledgement The authors gratefully acknowledge the financial support provided by the German Research Foundation within the Priority Program “Reactive Bubbly Flows”, SPP 1740 (ZI 1502/1-1). Dirk Ziegenbalg thanks Prof. Dr.-Ing. Elias Klemm (ITC) for continuous support.
References Figure 4: Molar decadic absorption coefficients of chlorine in aromatic solvents. Error bars are only shown exemplarily for one solvent.
[1] Horst Böttcher, Technical Applications of Photochemistry, Deutscher Verlag für Grundstoffindustrie, 1991. [2] A. M. Braun, M.-T. Maurette, E. Oliveros, Photochemical technology, Wiley, 1991. [3] M. Sender, D. Ziegenbalg, Light sources for photochemical processes – estimation of technological potentials, Chemie Ingenieur Technik (2017). [4] C. Lian, S.-Y. HO, Photochemical chlorination of methane, National Science Council, Proceedings 2 (1978) 4–7. [5] Ü. Ta¸stan, D. Ziegenbalg, Getting the most out of solar irradiation: Efficient use of polychromatic light for water splitting, Chemistry - A European Journal 22 (2016) 18824–18832. [6] J. E. Chateauneuf, Direct measurement of the absolute kinetics of chlorine atom in carbon tetrachloride, Journal of the American Chemical Society 112 (1990) 442–444. [7] J. E. Chateauneuf, Charge-transfer absorption spectra of chlorine atoms in halogenated solvents, Chemical Physics Letters 164 (1989) 577–580. [8] R. L. Wilkins, Continuous absorption coefficients of chlorine, Journal of Quantitative Spectroscopy and Radiative Transfer 4 (1964) 775–782. [9] G. E. Gibson, N. S. Bayliss, Variation with temperature of the continuous absorption spectrum of diatomic molecules: Part I. experimental, the absorption spectrum of chlorine, Physical Review 44 (1933) 188–192. [10] F. W. Jones, W. Spooner, Absorption of light by gaseous chlorine in the wavelength region 5040 a.u. to 5320 a.u., Trans. Faraday Soc. 31 (1935) 811–813. [11] D. J. Seery, D. Britton, The continuous absorption spectra of chlorine, bromine, bromine chloride, iodine chloride, and iodine bromide, The Journal of Physical Chemistry 68 (1964) 2263– 2266. [12] D. Maric, J. Burrows, R. Meller, G. Moortgat, A study of the UV—visible absorption spectrum of molecular chlorine, Journal of Photochemistry and Photobiology A: Chemistry 70 (1993) 205–214. [13] D. B. Kokh, A. B. Alekseyev, R. J. Buenker, Theoretical study of the UV photodissociation of Cl2 : Potentials, transition moments, extinction coefficients, and Cl*/Cl branching ratio, The Journal of Chemical Physics 120 (2004) 11549–11556. [14] S. Hubinger, J. Nee, Absorption spectra of Cl2 , Br2 and BrCl between 190 and 600 nm, Journal of Photochemistry and Photobiology A: Chemistry 86 (1995) 1–7. [15] G. A. Russell, Solvent effects in the reactions of free radicals and atoms, Journal of the American Chemical Society 79 (1957) 2977–2978. [16] G. A. Russell, Solvent effects in the reactions of free radicals and atoms. II. Effects of solvents on the position of attack of chlorine atoms upon 2,3-dimethylbutane, isobutane and 2deuterio-2-methylpropane1, Journal of the American Chemical Society 80 (1958) 4987–4996.
5. Conclusions
Developing efficient photochemical processes requires knowledge of basic physical properties of the reaction system. One of the most important properties in the context of photochemistry is the absorption coefficient of the reactants. Measuring the UV/VIS absorption spectra of photochemical active species can be challenging as they react during irradiation. As this is an intrinsic characteristic, analytical methods must be designed to take this into account, especially for systems showing fast reactions. To address these challenges, a continuous flow method was developed and applied to the measurement of absorption spectra of chlorine dissolved in different organic solvents. The results illustrate, that measurements of physical properties are of high relevance for photochlorinations and for the reaction engineering of photochemical processes in general. The absorption spectra of chlorine dissolved in aromatic solvents show significant hypsochromic as well as hyperchromic effects compared to non-aromatic solvents. Although this effect was already described in literature, the data published till now lack a thorough experimental investigation as well as an extension of the scope of solvents. This work fills this gap. While the presented methodology was developed for the determination of the absorption spectra of chlorine in different organic solvents, transferring the approach to other solvents, gases or reactants can be easily realized. This is especially relevant for the investigation of photochemical systems, which suffer from fast reaction rates, hazardous substances, sensitivity to moisture and/or oxygen or the evolving gases. Furthermore, it is possible to apply the methodology to reactions initiated by thermal energy as well. 5
[17] G. A. Russell, Solvent effects in the reactions of free radicals and atoms. III. Effects of solvents in the competitive photochlorination of hydrocarbons and their derivatives, Journal of the American Chemical Society 80 (1958) 4997–5001. [18] L. J. Andrews, R. M. Keefer, The interaction of chlorine with benzene and m-xylene, Journal of the American Chemical Society 73 (1951) 462–463. [19] Y. Su, Straathof, Natan J. W., V. Hessel, T. Noel, Photochemical transformations accelerated in continuous-flow reactors: Basic concepts and applications, Chem. Eur. J. 20 (2014) 10562– 10589. [20] K. A. Angelo Albini, Photochemistry: Volume 43, ROYAL SOCIETY OF CHEMISTRY, 2015. [21] K. Loubière, M. Oelgemöller, T. Aillet, O. Dechy-Cabaret, L. Prat, Continuous-flow photochemistry: A need for chemical engineering, Chemical Engineering and Processing: Process Intensification 104 (2016) 120–132. [22] D. Cambié, C. Bottecchia, N. J. W. Straathof, V. Hessel, T. Noël, Applications of continuous-flow photochemistry in organic synthesis, material science, and water treatment, Chemical Reviews 116 (2016) 10276–10341. [23] J. A. Lummiss, P. D. Morse, R. L. Beingessner, T. F. Jamison, Towards more efficient, greener syntheses through flow chemistry, The Chemical Record 17 (2017) 667–680. [24] D. Russo, D. Spasiano, M. Vaccaro, R. Andreozzi, G. L. Puma, N. M. Reis, R. Marotta, Direct photolysis of benzoylecgonine under UV irradiation at 254nm in a continuous flow microcapillary array photoreactor, Chemical Engineering Journal 283 (2016) 243–250. [25] G. Chiltz, R. Eckling, P. Goldfinger, G. Huynrechts, G. Martens, G. Simoens, The three body recombination of chlorine atoms, Bulletin des Sociétés Chimiques Belges 71 (1962) 747–758. [26] P. Goldfinger, Reactions of halogen atoms, Pure and Applied Chemistry 5 (1962) 423–440. [27] G. Chiltz, P. Goldfinger, G. Huybrechts, G. Martens, G. Verbeke, Atomic chlorination of simple hydrocarbon derivatives in the gas phase, Chemical Reviews 63 (1963) 355–372. [28] P. Skell, H. Baxter, J. Tanko, Reactions of BrCl with alkyl radicals, Tetrahedron Letters 27 (1986) 5181–5184. [29] R. A. Marcus, Relation between charge transfer absorption and fluorescence spectra and the inverted region, The Journal of Physical Chemistry 93 (1989) 3078–3086. [30] A. K. Croft, H. M. Howard-Jones, C. E. Skates, C. C. Wood, Controlling the action of chlorine radical: from lab to environment, Organic & Biomolecular Chemistry 9 (2011) 7439.
6