Standard Article

You have free access to this content

Photochemical and Photo-induced Reactions in Supercritical Fluid Solvents

Part 4. Supercritical Solvents

  1. James M. Tanko

Published Online: 15 MAR 2010

DOI: 10.1002/9783527628698.hgc043

Handbook of Green Chemistry

Handbook of Green Chemistry

How to Cite

Tanko, J. M. 2010. Photochemical and Photo-induced Reactions in Supercritical Fluid Solvents. Handbook of Green Chemistry. 4:10:399–417.

Publication History

  1. Published Online: 15 MAR 2010

Chemistry Terms

Choose one or more boxes to highlight terms.

Abstract

  1. Top of page
  2. Introduction
  3. Photochemical Reactions in Supercritical Fluid Solvents
  4. Photo-initiated Radical Chain Reactions in Supercritical Fluid Solvents
  5. Conclusion
  6. Acknowledgment
  7. References

The sections in this article are

  • Introduction
    • “Solvent” Properties of Supercritical Fluids
    • Scope of This Chapter
    • Experimental Considerations
  • Photochemical Reactions in Supercritical Fluid Solvents
    • Geometric Isomerization
    • Photodimerization
    • Carbonyl Photochemistry
    • Photosensitization and Photo-induced Electron Transfer
    • Photo-oxidation Reactions
  • Photo-initiated Radical Chain Reactions in Supercritical Fluid Solvents
    • Free Radical Brominations of Alkyl Aromatics in Supercritical Carbon Dioxide
    • Free Radical Chlorination of Alkanes in Supercritical Fluid Solvents
  • Conclusion
  • Acknowledgment

10.1 Introduction

  1. Top of page
  2. Introduction
  3. Photochemical Reactions in Supercritical Fluid Solvents
  4. Photo-initiated Radical Chain Reactions in Supercritical Fluid Solvents
  5. Conclusion
  6. Acknowledgment
  7. References

10.1.1 “Solvent” Properties of Supercritical Fluids There are several potential advantages that may be realized with the use of supercritical fluids (SCFs) as solvents for chemical reactions from the standpoint of reactivity and selectivity. As many of the examples discussed in this chapter illustrate, the unique features of SCFs can be exploited to control the behavior (i.e. kinetics and selectivity) of many chemical processes in a way not possible with conventional liquid solvents.

Changes in reaction rates arising from direct effects of temperature and pressure on the kinetics of a reaction are governed by transition-state theory, and the same considerations pertain to reactions both in SCF media and in conventional solvents (1-4). However, a unique feature of SCFs is that solvent properties, such as polarity (dielectric constant), viscosity, and solubility parameter, vary with temperature and pressure, and changes in these properties may alter reaction rates. The polarity of the reaction medium will exert an effect on the rate of a chemical reaction if the polarities of the reactants and transition state are different. Solvent viscosity will exert its influence on reactions that are diffusion controlled or on reactions in which cage effects are important. Hence control of these solvent properties, via manipulation of temperature and pressure, provides a way of adjusting the kinetics of a chemical process unique to the SCF medium (1-3). As many of the studies cited here demonstrate, with an SCF solvent it is possible to study solvent effects on reaction rates without varying the molecular functionality of the solvent.

However, there are still additional factors which may affect reactivity. Numerous studies have shown that, for SCFs in the compressible region of the phase diagram, the local solvent density about a solute is often enhanced relative to the bulk solvent density (3, 5). The term “solvent–solute clustering” has been coined to describe this phenomenon. Because of enhanced local solvent density, the rotational and translational motion of a solute may be restricted (attributable to increased local viscosity). In this scenario, reaction rate and selectivity will be perturbed only when the reactions are extremely rapid (diffusion-controlled), or if cage effects are important – as is often the case with photochemical reactions. There is also evidence that, in some cases, “solute–solute clustering” may be important (6-12), and conceivably reaction rates could be affected because of locally higher concentrations. In general, clustering is most important near the critical point, and there is considerable interest in what effect this phenomenon has on reaction rates and selectivities(13, 14) (see below). An excellent review of clustering and solvation in supercritical fluids appeared in 1999 (15).

10.1.2 Scope of This Chapter This chapter summarizes the literature through 2007 pertaining to photochemical and photo-initiated organic reactions in SCF solvents. Photochemical reactions involving organometallic compounds in SCF solvents have been extensively studied by Poliakoff and co-workers (16, 17). In this chapter, the emphasis is on reaction chemistry in which stable products are formed and isolated, and how the unique features of the SCF medium influence reaction yields, rates, and/or selectivities.

What are not discussed at length are photophysical phenomena in SCF solvents (e.g. fluorescence quenching, triplet–triplet annihilation, charge transfer, and exiplex formation), which have been extensively used to probe SCF properties, in general, and have been especially informative regarding the existence of clusters (solvent–solute and solute–solute) and their effect on reactivity. Absorption and fluorescence spectroscopy (both steady-state (9, 18-35) and time-resolved) (11, 36-50), vibrational spectroscopy (51-56), pulse radiolysis (57, 58), and electron paramagnetic resonance (EPR) (59-62) have all been utilized in this regard. The interested reader is directed to the references provided above for more information on these topics.

10.1.3 Experimental Considerations Photochemical reactions conducted under supercritical conditions require high-pressure reaction vessels equipped with a “window” which permits light to enter, and the necessary hardware/plumbing to generate high pressures and maintain constant temperature. A typical reaction vessel (63) used for photochemical reactions conducted in supercritical carbon dioxide (scCO2) (Figures 10.1 and 10.2) is fabricated from a strong (inert) alloy such as stainless steel or Hastelloy, and is equipped with a sapphire window. Sapphire is especially suited for high-pressure work and is optically transparent in the region 150–6000 nm. CaF2 and quartz have also been used. Provisions for stirring may also be included; in many cases, a simple magnetic stir bar suffices. Temperature control is typically achieved through the use of a resistive heater, thermocouple, and a temperature controller. Utilizing such a reactor, pressures up to 70–100 MPa can be achieved.

original image

Figure 10.1. Cross-section of scCO2 reactor.

original image

Figure 10.2. Photograph of scCO2 reactor.

A complete system for generating scCO2 (Figure 10.3) (63) requires a device to generate high pressures (compressor, high-pressure piston, or HPLC pump), a pressure transducer, and the necessary plumbing. Often, a high-pressure release valve (rupture disk) is used to ensure that pressures in the reactor do not exceed specification. Because many photochemical and free radical reactions require the exclusion of oxygen, provisions can be made for purging the system with an inert gas such as argon. Solid and liquid samples can be added to the reactor under an argon backflush; volatile liquids or gases can be introduced in glass vials which rupture when the reactor is pressurized.

original image

Figure 10.3. Schematic diagram of apparatus for generating scCO2.

Generally, light sources typically used for photochemical reactions in conventional solvents (e.g. medium- or high-pressure mercury lamps) are used for reactions in SCF solvents. Because many popular SCFs used for photochemical experiments do not absorb UV–visible light (e.g. CO2, CHF3, low molecular weight alkanes), and because these experiments are usually conducted at low substrate concentrations, heating of the reactor and the concomitant increase in pressure are generally not a problem. Occasionally, heating may occur because of the heat given off by the lamp itself, but separating the lamp and reactor and providing ample ventilation can avoid this.

After irradiation, the reactor must be vented so as to bring the system to ambient pressure. Products that are solids often precipitate from solution as the pressure is lowered, and are readily recovered. Separation of volatile compounds from an SCF is often more challenging, and usually entails “bubbling” the contents of the reactor into an organic solvent. Another strategy, applicable to SCFs that are not gases at room temperature (e.g. H2O), is to cool the reactor to room temperature where the system is no longer pressurized.

Two procedures that illustrate some of the experimental protocols associated with these experiments are highlighted below:

The photodimerization of isophorone in scCO2 (64) (see below) was accomplished first by pressurizing (to ∼0.4 MPa) and depressurizing the reactor with CO2 several times to purge the system of air. The reagent was introduced into the reactor via a syringe under a positive pressure of CO2. Subsequently, the reactor was sealed, brought to the desired temperature and pressure, and irradiated with a 450 W medium-pressure mercury lamp. At completion of the reaction, the reactor contents were bubbled into methylene chloride, the reactor and lines were rinsed with solvent, and the combined solutions were analyzed by gas chromatography (GC).

For the chlorination of cyclohexane in scCO2, the following procedure was followed (65, 66). The appropriate volume of cyclohexane was placed in a 1 ml ampoule. The ampoule was degassed by several freeze–pump–thaw cycles (freezing to −198 °C, evacuating to less than 0.01 mbar, and warming to room temperature), sealed under vacuum, and placed in the reactor. A second sealed ampoule containing the appropriate amount of Cl2 (similarly degassed) was added to the reactor. The reactor was then sealed, covered with aluminum foil (to prevent premature initiation of the reaction via action of ambient laboratory light), and brought to 40 °C (the desired reaction temperature). Following several argon purges, the reactor was pressurized with CO2 and allowed to equilibrate at 40 °C for several minutes. The aluminum foil was removed and the reactor was illuminated as with a 450 W mercury arc lamp. Following illumination, the contents of the reactor were bubbled slowly into hexanes cooled to 0 °C. An internal standard was added, and direct analyses by GC were performed to assess product yields.

10.2 Photochemical Reactions in Supercritical Fluid Solvents

  1. Top of page
  2. Introduction
  3. Photochemical Reactions in Supercritical Fluid Solvents
  4. Photo-initiated Radical Chain Reactions in Supercritical Fluid Solvents
  5. Conclusion
  6. Acknowledgment
  7. References

10.2.1 Geometric Isomerization Aida and Squires examined the photoisomerization of (E)-stilbene (Equation 10.1) in a conventional organic solvent (cyclohexane) and scCO2. This system was selected for study because the solvent effect on the isomerization was already documented; increased viscosity facilitates the E[RIGHTWARDS ARROW]Z conversion (67).

original image(10.1)

For liquid CO2 at 25 °C, a change in pressure from 8.3 to 21.4 MPa (corresponding to a change in viscosity from ∼0.07 to 0.1 cP) changes the Z:E ratio from 5.5 to 6.8. For scCO2, where an analogous pressure variation changes the viscosity from 0.02 to 0.08 cP, the effect is more dramatic, with the Z:E ratio changing from 1.4 to 7.0 (67). This study provided one of the first examples of how the outcome of a photochemical reaction can be altered by varying the solvent properties of the SCF via manipulation of pressure.

10.2.2 Photodimerization In 1989, Fox and co-workers (64). studied the [2 + 2] photodimerization of isophorone (Equation 10.2) in scCO2 (38 °C) and scCHF3 (34.5 °C). Three dimers were produced, a head-to-head dimer (H-Hanti), and two diastereomeric head-to-tail dimers (H-Tanti and H-Tsyn). In conventional solvents, Chapman et al. found that more polar solvents favor production of the more polar product: the ratio H-Hanti:H-Ttotal was 1 : 4 in cyclohexane compared with 4 : 1 in methanol (68).

original image(10.2)

Analogous results were observed in SCF solvents: The more polar product (H-Hanti) was a major product in the more polar solvent CHF3 (where the H-H:H-Ttotal ratio varied from 0.75 to 1.0 with increasing pressure) and only a minor product in CO2 (in which the H-H:H-Ttotal ratio was essentially 0.10, independent of pressure) (64). These observations are explicable on the basis that over the range of pressures examined, the dielectric constant (a measure of solvent polarity) varies more for CHF3 (from 2.5 to 8.4) than it does for CO2 (from 1.34 to 1.54).

An unexpected result of this study was that for the head-to-tail dimers, the anti:syn ratio varied with pressure (Figure 10.4). In conventional solvents, both are formed in approximately equal amounts. The authors suggested that “differential solvent reorganization” was responsible (i.e. that more desolvation must occur to form the syn isomer compared with the anti isomer, Scheme 10.1) (64). Thus, at higher pressures (higher solvent densities), solvent reorganization was important thereby favoring the anti isomer.

original image

Figure 10.4. Anti:syn ratio for the “head-to-tail” dimers formed in the dimerization of isophorone in scCHF3 (34.5 °C) and scCO2 (38 °C) as a function of pressure. Data taken from (72).

original image

Scheme 10.1.

Sun and co-workers reported that in scCO2, the quantum yield for anthracene dimerization (Equation 10.3) was (a) 10 times greater in scCO2 than in conventional liquid solvents at comparable anthracene concentrations, and (b) pressure dependent, with the yield decreasing at higher pressures (69). The key step in the dimerization process involves the formation of an eximer via diffusion-controlled reaction of anthracene in its ground and singlet excited states, A and inline image, respectively (Equation 10.4). In conventional liquid solvents, the rate constant for a diffusion-controlled reaction is on the order of 1010 M−1 s−1. However, because of the higher diffusivity of the SCF medium, the limit for diffusion control is higher (∼1011 M−1 s−1). Thus, because the quantum yield for anthracene dimerization is directly proportional to kdim, a 10-fold increase in efficiency is achieved in SCF media. The pressure effect arises because the CO2 viscosity increases with increase in pressure (69).

original image(10.3)
original image(10.4)

10.2.3 Carbonyl Photochemistry In 1992, Kraus and Kirihara reported that acylhydroquinones could be synthesized photochemically from the corresponding aldehyde and quinone (Scheme 10.2) (70), and in 2001 the same group successfully transferred this chemistry into scCO2 solvent (71). Irradiation of the quinone generates the triplet state, which abstracts the relatively weak C[BOND]H bond of the formyl group, generating radical pair 3. Radical–radical coupling forms 4, which, after enolization, yields product 5. This chemistry provides an environmentally benign alternative to the Friedel–Crafts acylation reaction because it eliminates the use of acid chlorides, strong Lewis acids, and benzene as a solvent for the synthesis of acylhydroquinones.

original image

Scheme 10.2.

The Norrish Type I photo-cleavage represents a classic process in organic photochemistry, and has been extensively studied in conventional solvents. In 1991, Fox and co-workers extended this reaction to SCF solvent: in an attempt to probe for cage effects and possibly enhanced cage effects attributable to solvent–solute clustering, the photolysis of an unsymmetrical dibenzyl ketone was examined in scCO2 and scC2H6 (72). Dibenzyl ketone photolysis had been shown to lead to cage effects in conventional solvents. The rationale behind this experiment is outlined in Scheme 10.3. Photolysis of an unsymmetrical dibenzyl ketone, A(C[DOUBLE BOND]O)B, leads to the formation of two benzyl radicals and carbon monoxide in the solvent cage, depicted as [A[FREE RADICAL] CO [FREE RADICAL]B]cage. In-cage coupling is expected to yield exclusively the cross-coupling product A–B, whereas cage escape will lead to all possible coupling products, A–A, A–B, and B–B, in a statistical ratio of 1 : 2 : 1.

original image

Scheme 10.3.

Over a pressure range from just above Pc to 30.0 MPa, this reaction yielded only a statistical distribution of products (72). Hence no cage effect (enhanced or otherwise) was observed for this reaction.

This problem was later discussed by Chateauneuf and co-workers, who examined the decarbonylation of the phenylacetyl radical (PhCH2C[FREE RADICAL][DOUBLE BOND]O [RIGHTWARDS ARROW] PhCH2[FREE RADICAL] + C[DOUBLE BOND]O) by laser flash photolysis (44). These workers again found no evidence for a cage effect (enhanced or otherwise) in SCF solvent. Moreover, to explain the absence of a cage effect in these reactions, they went on to suggest that the integrity of the cage is maintained for only a few picoseconds, whereas decarbonylation occurs in the time regime of a few hundred nanoseconds (i.e. the cage disintegrates long before in-cage coupling can occur) (44).

In order to address this issue, a process which involves a much shorter-lived radical pair needed to be examined. Toward this end, Weedon and co-workers examined the photo-Fries rearrangement of naphthyl acetate (Scheme 10.4) in scCO2 at 35 and 46 °C (73). Photolysis of 6 leads to caged-pair (7, 8); reaction in-cage yields the photo-Fries products, 2- or 4-acetylnaphthol (9). On the other hand, cage escape, followed by hydrogen abstraction (2-propanol was present as a hydrogen atom donor), leads to α-naphthol (10).

original image

Scheme 10.4.

A plot of the product ratio 9 : 6 as a function of pressure is presented in Figure 10.5. This plot exhibits a dramatic spike at pressures near the critical pressure, which the authors attributed to the onset of solvent–solute clustering; disintegration of the caged pair is inhibited because the viscosity at the molecular level is much greater than the bulk viscosity.

original image

Figure 10.5. Ratio of products produced from photolysis of α-naphthyl acetate in CO2. Data taken from (81).

Tanko and Pacut examined the behavior of geminate and diffusive caged radical pairs, both generated by the photolysis of dicumyl ketone in scCO2 (74). A geminate caged radical pair arises when the two radicals are generated simultaneously from a common precursor, such as caged pair 12 generated from photolysis of dicumyl ketone (11, Scheme 10.5). Geminate caged pair 12 partitions between two pathways: cage escape (kesc, the magnitude of which is viscosity dependent) and in-cage hydrogen abstraction (kH). Because distinct products arise from each of these two competing pathways, the rate constant ratio kesc/kH is readily determined, and provides a measure of solvent viscosity at the molecular level. In these experiments, the magnitude of the cage effect was found to be greater than expected based on extrapolations from conventional solvents (74). These extrapolations were based on the relationship between the diffusion coefficient (D) and viscosity as described by the Stokes–Einstein equation, which may overestimate D and hence underestimate the cage effect in scCO2.

original image

Scheme 10.5.

In this system, a diffusive caged radical pair (14) is also formed by two cumyl radicals diffusing together. Radical pair 14 also partitions between two pathways: dimerization (kdim) and disproportionation (kdisp). The rate constant ratio kdim/kdisp is also viscosity dependent, and easily measured by product yields. The rate constant kdim decreases with increasing viscosity because the two radicals must rotate and align in order for bond formation to occur. In contrast, the geometric constraints for disproportionation are less rigid, and kdisp is less sensitive to viscosity. Again, the magnitude of the cage effect was found to be greater than expected. For both the geminate and diffusive caged pairs in this system, the magnitude of the cage effect was found to increase near the critical pressure, possibly the result of enhanced local viscosity attributable to solvent–solute clustering (74).

10.2.4 Photosensitization and Photo-induced Electron Transfer Using scCO2 as both a solvent and reactant, Chateauneuf et al. (75) reported the synthesis of 9,10-dihydroanthracene-9-carboxylic acid from anthracene via photo-induced electron transfer. Irradiation of anthracene to its triplet state in the presence of a good electron donor such as N,N-dimethylaniline leads to the anthracene radical anion. In the presence of a hydrogen atom donor such as 2-propanol, the carboxylic acid is formed in 57% yield, presumably by the mechanism depicted in Scheme 10.6. Consistent with the electrophilic nature of CO2, trapping by anthracene radical anion is best viewed as nucleophile–electrophile coupling, rather than radical addition to CO2.

original image

Scheme 10.6.

Using a chiral sensitizer, Inoue and co-workers (76-78) found that photoaddition of alcohols to 1,1-diphenylpropene was enantioselective in scCO2 (Equation 10.5). The enantiomeric excess (ee) exhibited a jump as the pressure was increased from below to above the critical pressure, the magnitude of which also varied with the structure of ROH. For example, at 8 and 18 MPa, the observed ee with MeOH was 7 and 21%, respectively. With 2-propanol at these same pressures, ees of 21 and 42% were observed, respectively. This peculiar behavior near the critical pressure was attributed to clustering of ROH around the sensitizer. In a related study, these authors also reported an anomalous pressure dependence on the enantioselectivity of the photosensitized Z[RIGHTWARDS ARROW]E isomerization of cyclooctene in scCO2, which was attributed to CO2 clustering near the critical pressure (79).

original image(10.5)

10.2.5 Photo-oxidation Reactions Koda and co-workers reported the photo-induced (KrF laser, 248 nm) oxidations of ethylene (80) and benzene (81) in scCO2 (Equations 10.6 and 10.7). In the context of product distribution and yields, no unusual behavior was noted near the critical point.

original image(10.6)
original image(10.7)

10.3 Photo-initiated Radical Chain Reactions in Supercritical Fluid Solvents

  1. Top of page
  2. Introduction
  3. Photochemical Reactions in Supercritical Fluid Solvents
  4. Photo-initiated Radical Chain Reactions in Supercritical Fluid Solvents
  5. Conclusion
  6. Acknowledgment
  7. References

10.3.1 Free Radical Brominations of Alkyl Aromatics in Supercritical Carbon Dioxide In 1994, Tanko and Blackert reported that the free radical bromination of alkylaromatics (e.g. toluene) could be carried out in scCO2 (82). This reaction is photo-initiated, and proceeds via the chain process outlined in Scheme 10.7 (83).

original image

Scheme 10.7.

Reaction yields were analogous to those observed using conventional solvents (e.g. CCl4). Via competition experiments, the relative reactivity of the “secondary” hydrogens of ethylbenzene versus the “primary” hydrogens of toluene on a per hydrogen basis, r(2o/1o), were assessed. Within experimental error, the selectivity did not vary over a pressure range of 7.5–42.3 MPa [r(2o/1o) = 30 ± 2] at 40 °C (82). In retrospect, this result is reasonable because (a) the hydrogen abstraction step is insensitive to solvent polarity effects, and is well below the diffusion-controlled limit so that viscosity effects are unimportant, and (b) the difference in the volume of activation for hydrogen abstraction from toluene versus ethylbenzene is small (∼4.8 cm3 mol−1), so that over the range of pressures examined the selectivity change would be of the same magnitude as experimental error (63, 82).

The observed selectivity in scCO2 is nearly identical with that found in conventional organic solvents: r(2o/1o) = 35 ± 1, 34 ± 1, and 29 ± 1 for CCl4, Freon 113, and CH2Cl2, respectively. at 40 °C (63, 82). These results confirm the role of Br[FREE RADICAL] as the chain carrier in these experiments, as depicted in Scheme 10.7, and suggest that Br[FREE RADICAL] selectivity is not altered by complexation to CO2. It is noteworthy that Br[FREE RADICAL]does form a complex with CS2 (which is isoelectronic with CO2) and that this complex does exhibit enhanced selectivities in hydrogen atom abstractions (84).

With molecular bromine (Br2) as the brominating agent, a small amount of p-bromotoluene is formed, arising from the competing electrophilic aromatic substitution (EAS) process. However, with the use of N-bromosuccinimide (NBS) as the brominating agent in direct analogy with the classical Ziegler reaction (Equation 10.8), the EAS side-product is completely eliminated. Reaction yields and selectivities are identical with those observed in CCl4, the solvent most widely used for the Ziegler reaction (63, 82).

original image(10.8)

Competition experiments (ethylbenzene versus toluene) confirm the role of Br[FREE RADICAL] as chain carrier in the Ziegler bromination in scCO2 (63, 82). The role of NBS in this reaction is to maintain a low, steady-state concentration of Br2, by scavenging HBr as it is produced during the course of the reaction (Equation 10.9).

original image(10.9)

10.3.2 Free Radical Chlorination of Alkanes in Supercritical Fluid Solvents The free radical chlorination of alkanes is a classic procedure for the functionalization of alkanes. Many of the details of this reaction have been well understood for more than half a century (85). In the laboratory, this reaction is initiated by action of visible light, with product formation occurring via the propagation steps outlined in Scheme 10.8: Chlorine atom abstracts hydrogen from the alkane yielding an alkyl radical and HCl. The alkyl radical subsequently reacts with molecular chlorine yielding the product alkyl chloride and regenerating chlorine atom.

original image

Scheme 10.8.

A chlorine atom is a highly reactive species and exhibits low selectivity in hydrogen abstractions. In solution, the preference decreases in the order tertiary C[BOND]H (4.2) > secondary C[BOND]H (3.6) > primary C[BOND]H (1.0), on a per hydrogen basis (25 °C) (83). Absolute rate constants for hydrogen abstraction are slightly below the diffusion-controlled limit (86).

The chlorine atom cage effect, first discovered by Skell and Baxter in 1983 (87), has been the subject of numerous investigations (88, 89). Put briefly, for the chlorine atom abstraction step in the free radical chlorination of an alkane (RH2), the geminate RHCl[BOND]Cl[FREE RADICAL] caged pair is partitioned between three pathways (Scheme 10.9): diffusion apart (kdiff), abstraction of hydrogen from RH2 comprising the cage walls (inline image), and a second in-cage abstraction of hydrogen from the alkyl chloride (kRHCl). Although the kdiff and inline image steps result in the formation of monochloride (RHCl), the kRHCl step results in the formation of polychlorides. In conventional solvents, the ratio of mono- to polychlorinated products (M/P) has been shown to depend on solvent viscosity (90).

original image

Scheme 10.9.

Tanko and co-workers utilized the chlorine atom cage effect as a highly sensitive probe for studying the effect of SCF viscosity and the possible role of solvent clusters on cage lifetimes and reactivity (65, 66). These experiments were conducted in scCO2 (40 °C at various pressures), with parallel experiments in conventional solvents and in the gas phase.

Cage effects are typically quantified in terms of the Noyes model, which predicts that the efficiency of cage escape should vary linearly with the inverse of viscosity (1/η) (91). In Figure 10.6, the ratio of mono- to polychlorides (M/P) observed in the chlorination of 2,3-dimethylbutane, neopentane, and cyclohexane is plotted as a function of 1/η for the experiments conducted in scCO2and in conventional solvents. Overall, these plots are linear over a range of viscosities spanning 1.7 orders of magnitude (from conventional solvents to scCO2) and provide no indication of an enhanced cage effect (unusually low observed M/P ratio) near the critical pressure. It is also worth noting that the best straight line through the solution-phase results successfully predicts the SCF phase results (65, 66).

original image

Figure 10.6. Ratio of mono- to polychlorides produced in the free radical chlorination of 2,3-dimethylbutane (23DMB), neopentane, and cyclohexane in conventional and supercritical fluid solvents as a function of inverse viscosity at 40 °C. Data taken from (73) and (74).

Based on these observations, there was no indication of an enhanced cage effect near the critical point in scCO2 solvent. The magnitude of the cage effect observed in scCO2 at all pressures examined is well within what is expected based on extrapolations from conventional solvents. It was suggested that for instances where enhanced cage effects have been observed attributable to solvent–solute clustering, this enhancement may be unique to the specific systems studied (65, 66).

The experiments with 2,3-dimethylbutane (23DMB) provided insight into the extent that Cl[FREE RADICAL] selectivity varies as a function of pressure. In the gas phase (40 °C), the relative reactivity of the tertiary and primary hydrogens of 23DMB [r(3o/1o)] is 3.97. In the condensed phase (neat 23DMB, 40 °C), r(3o/1o) = 3.27. In scCO2, r(3o/1o) varies with pressure and falls between the gas- and liquid-phase values. The fact that r(3o/1o) is so close to the solution and gas-phase values suggests that Cl[FREE RADICAL] selectivity is not altered by complexation to CO2 (65).

The slight variation in r(3o/1o) can be explained as follows. The rate constants for primary, secondary, or tertiary hydrogen abstractions by Cl[FREE RADICAL] from alkanes are nearly diffusion controlled in conventional solvents. Consequently, the intrinsic selectivity of Cl[FREE RADICAL] is diminished in conventional solvents because of the onset of diffusion control. In the gas phase, selectivity is slightly higher because the barrier imposed by diffusion is eliminated. The viscosity of a supercritical fluid (a) lies between those of conventional liquid solvent and the gas phase and (b) varies with pressure. Because of the low viscosity of supercritical fluids, bimolecular rate constants greater than the 1010 M−1 s−1 liquid-phase diffusion-controlled limit can be realized in SCFs and, as a consequence, enhanced selectivity is achieved. Consistent with this interpretation is the observation that the plot of r(3o/1o) versus inverse viscosity is approximately linear (Figure 10.7) (65).

original image

Figure 10.7. Chlorine atom selectivity in scCO2 solvent at 40 °C. Data taken from (73).

10.4 Conclusion

  1. Top of page
  2. Introduction
  3. Photochemical Reactions in Supercritical Fluid Solvents
  4. Photo-initiated Radical Chain Reactions in Supercritical Fluid Solvents
  5. Conclusion
  6. Acknowledgment
  7. References

The examples discussed here demonstrate that the unique nature of SCFs provides a means of “dialing up” the selectivity of a chemical process in a manner which is simply impossible using conventional solvents, that is by manipulation of temperature and pressure. Reaction rates may vary with temperature and pressure depending on the magnitude of the activation energy and activation volume, respectively, for reactions conducted in both SCF and conventional solvents. What is unique about SCF solvents is that the actual nature of the solvent (polarity, viscosity, etc.) also varies with temperature and pressure. In addition, numerous studies have demonstrated surprisingly large variations in selectivity with pressure (density) near the critical pressure – generally attributed to clustering. However, from a practical perspective (chemical manufacturing and synthesis), large variations in selectivities resulting from small changes in pressure are not likely to be useful. Rather, it is the fact that selectivities tend to be either constant, or at least predictable, at pressures sufficiently above the critical pressure that will make SCFs desirable for chemical manufacturing purposes. SCF solvents such as CO2 and H2O are especially attractive as they are “environmentally benign” alternatives to a number of classical solvents that pose hazards to either health or the environment. Coupled with the tunable properties of a supercritical fluid, these solvents emerge not only as viable alternatives to conventional organic solvents, but in some cases at least, also as superior alternatives. Finally, SCF solvents are superb tools for probing solvent effects in chemical processes, as it is possible to vary (via manipulation of pressure) pertinent solvent properties (e.g. viscosity and polarity) without changing the molecular functionality of the solvent.

Acknowledgment

  1. Top of page
  2. Introduction
  3. Photochemical Reactions in Supercritical Fluid Solvents
  4. Photo-initiated Radical Chain Reactions in Supercritical Fluid Solvents
  5. Conclusion
  6. Acknowledgment
  7. References

Support from the National Science Foundation (CHE-0548129) during the writing of this chapter is acknowledged and appreciated.

References

  1. Top of page
  2. Introduction
  3. Photochemical Reactions in Supercritical Fluid Solvents
  4. Photo-initiated Radical Chain Reactions in Supercritical Fluid Solvents
  5. Conclusion
  6. Acknowledgment
  7. References