Original PhD Thesis (1987) by Horst Hintze
Refurbished (2026) by Horst Hintze-Brüning
Abstract
Dihydroaromatics are of particular interest as hydrogen donors because they are readily available and highly reactive. From a mechanistic perspective, 1,4-dihydroaromatics are particularly interesting, as there is ongoing debate regarding whether their aromatization involves synchronous or two-step ionic dehydrogenation.
The subject of this work is mechanistic studies of two types of hydrogen transfer from 1,4-cyclhexadiene:
- thermal transfer to quinones,
- homogeneous catalyzed disproportionation.
Five regio- and stereoselectively deuterated 1,4-cyclohexadienes were synthesized with the high isotopomeric and steric purity required for kinetic measurements and tracer studies.
H-transfer to both quinones ‒ 2,3-dichloro-5,6-dicyano-1,4-benzoquinone (DDQ) and 3,4,5,6-tetrachloro-1,2-benzoquinone (OCA) ‒ occurs via primary hydride transfer followed by proton abstraction. The high primary isotope effects (13.6 (OCA), 11.5 (DDQ) indicate a tunneling component in the hydride transfer. Tracer studies showed that dehydrogenation proceeds with high cis-selectivity. Activation parameter deduced from the temperature dependency of the primary isotope effects suggest an angled transition state for the hydride transfer. At the same time, the limitations of the exclusive use of activation parameters as a criterion for the geometry of transition states were demonstrated in this work.
Disproportionation on collidal nickel at 60°C in DMF proceeds without by-product formation: first-order kinetics with respect to the catalyst and second-order with respect to the reactant. A highly cis-selective dehydrogenation is followed by a rate-limiting hydrogenation in which the molecular identity of the abstracted hydrogen is lost.
Disproportionation of 1,4-cyclohexadiene by the nickel(II) acetylacetonate/triethylaluminum Ziegler catalyst at 25°C in dioxane and toluene, resp. proceeds without isomerization or byproduct formation involving a distinct induction period which is not observed with 1,3-cyclohexadiene. The latter reacts zero-order in dioxane and first-order in toluene and is preferentially consumed in competition with the 1,4-isomer despite its lower reactivity. Following a rapid, left-handed equilibrium of ligand exchange, the diene is gradually dehydrogenated in a rate determining oxidative addition followed by β- or γ-H abstraction, respectively.
Main Part
1. Introduction
Intermolecular hydrogen transfers between organic molecules are of interest for various reasons (Johnstone et al. 1985). On the one hand, the mechanisms are the subject of controversy; on the other hand, these reactions are used as both efficient and selective preparative methods in both laboratory and industrial synthesis (Fu and Harvey 1978). More recently, hydrocarbons have also been gaining importance as potential hydrogen donors in coal and petroleum technology (Virk 1978). Furthermore, model reactions for biologically relevant hydride transfers (e.g., NAD/NADH) can provide helpful information about their possible mechanisms, as recent studies on the following system demonstrate (Verhoeven et al. 1983),(Verhoeven and Gerresheim 1983):
Based on the fundamental work by E.A. BRAUDE et al. (Braude and Linstead 1954), a distinction is made between intramolecular (H migration within a molecule), intermolecular H transfers (donor/acceptor), and disproportionations (donor = acceptor), all of which can occur both thermally and catalytically. Both reaction types can proceed via various intermediate stages or transition states:
- Concerted mechanism Synchronous group transfers are frequently proposed for H-transfer reactions that are symmetry-permitted according to the Woodward–Hoffmann rules (Woodward and Hoffmann 1970),(Anh 1972). An example of this mechanism is the reduction of olefinic double bonds by diimines (Hünig et al. 1965):
- Stepwise mechanism The transfer of two hydrogen atoms occurs in two temporally distinct steps, forming an intermediate with a more or less long-lived nature. The hydrogen atoms can be transferred as radicals (H*), as ions (H-, H+), or via a SET (single-electron transfer) mechanism (e-; H+, e-). An example of a radical reaction is the thermal disproportionation of 1,2-dihydronaphthalene:
Studies conducted in our research group using stereospecifically deuterated reactants demonstrated a stereo-nonselective course (Heesing and Müllers 1980a) and thus ruled out a pericyclic transfer, as had previously been postulated by GILL and HAWKINS (Gill et al. 1974) based solely on kinetic data. In contrast, W. MÜLLERS was able to demonstrate the existence of ionic intermediates in the stereospecific cis-dehydrogenation of 1,2-dihydronaphthalene by tetracyanoethylene using isotope effects (Heesing and Müllers 1980c);
SET mechanisms can be regarded as a limiting case of hydride transfer in which the primary electron transfer is rate-determining. Thus, the dehydrogenation of NADH analogs can occur via both mechanistic extremes, depending on the oxidation potentials of the acceptors (Verhoeven et al. 1986). Among thermally induced hydrogen transfers, the dehydrogenation of dihydroaromatics by electonegatively substituted quinones has received particular attention, e.g., in the synthesis of reference compounds for metabolic derivatives of polycyclic aromatic hydrocarbons (Harvey 1986). For these systems, concerted and ionic mechanisms are predominantly discussed in the literature. While R. PAUKSTAT was able to demonstrate a two-step process for the aromatization of 1,2-dihydroarenes (Paukstat et al. 1985), a conclusive study of the mechanism for the analogous reaction of the more reactive 1,4-dihydroaromatics — in which, in addition to the symmetry rules, the steric conditions for a pericyclic transition state are also favorable (Anh 1972) — a conclusive study of the mechanism is still pending.
The focus of the first part of this work is a mechanistic study of the dehydrogenation of 1,4-cyclohexadiene by electrophilic quinones, with the aim of elucidating the nature of the rate-determining step. For this purpose, several regio- and stereoselectively deuterated starting materials were synthesized and dehydrogenated with quinones under mild conditions. The isotope effects and stereochemistry of the reactions were determined using various methods.
Catalytic hydrogen transfer reactions represent an interesting alternative to catalytic hydrogenation with molecular hydrogen in organic chemistry because they often proceed chemo- and regioselectively under mild conditions (Brieger and Nestrick 1974). Hydroaromatics are of particular importance as H donors due to their easy availability and high reactivity. In contrast to reactions catalyzed by soluble transition-metal complexes — whose mechanisms are consistent with elementary reaction modes and bonding relationships in transition-metal chemistry (Gessner and Heesing 1985) — there are comparatively few mechanistic studies available for the transfers catalyzed by metals and metal oxides that allow for an exact formulation of the reaction mechanism.
Of particular interest is whether the H-acceptor is involved in the dehydrogenation of the donor, or whether dehydrogenation and hydrogenation are two separate steps. The palladium-catalyzed heterogeneous disproportionations of cyclohexene (Carrà and Ragaini 1967) and 1,2-dihydronaphthalene (Heesing and Müllers 1980b) each represent examples of one of the two reaction pathways.
In the second part of this work, the homogeneously catalyzed disproportionation of 1,4-cyclohexadiene over suitable nickel catalysts is investigated; in particular, the dehydrogenation step was subjected to a detailed study using regio- and stereoselectively deuterated reactants.
2. Syntheses of Deuterated 1,4-Cyclohexadienes
2.1. Selection of Starting Materials and Synthesis Routes for Labeled 1,4-Cyclohexadienes
The quality of the starting materials—in terms of regio- and stereospecificity as well as the degree of labeling of the introduced isotope—is crucial for the interpretability of isotope effects and tracer studies. In particular, kinetic studies are increasingly distorted by proportions of the lower-indexed isotopomer exceeding 5% due to its higher reactivity. This source of error has a greater impact on isotopomers that are deuterium-labeled at multiple equivalent positions than on compounds in which only some of these hydrogen atoms are substituted. These constraints often greatly reduce the number of possible synthetic routes for the target molecule. The following isotopomers of 1,4-cyclohexadiene were synthesized in this work:
Of these, the isotopomers 1a–1d are known from the literature. However, for the crucial, stereoselectively indexed compounds 1c and 1d in particular, insufficient purities were achieved with regard to stereochemistry and the degree of indexing (Müller et al. 1984). In particular, 1c was synthesized in this work for the first time as a “pure substance” with a steric purity of > 90 percent. Compound 1e has not been described to date. The syntheses of the precursors for 1a, 1b, and 1e were optimized with regard to the degree of indication and chemical purity. In addition to compounds 1a–1d, the isotopomers 1f–1i are known from the literature:
The isotopomers 1h and 1i (Müller 1973), which are readily accessible via BIRCH reduction of the aromatics, are hardly suitable for mechanistic studies due to the accumulation of isotopic effects and the lack of stereospecificity in 1h. The isotopomer 1f can be obtained by functional group conversion from 1,4-cyclohexanedione only as a mixture with the isomeric 1,3-cyclohexadiene (Franzus 1963). Furthermore, the synthesis of 1g from the corresponding deuterated 1,3-butadiene was omitted, as it does not provide any more mechanistic information than is already available from 1a–1e. All syntheses of the isotopomers of 1,4-cyclohexadiene used in this work proceed via a Diels–Alder reaction of the corresponding indexed 1,3-butadienes to build the C6 carbon skeleton. Reactive dienophiles such as maleic and fumaric acid derivatives are suitable for this key reaction; from their cycloaddition products, the cyclohexadienes can then be obtained via oxidative bis-decarboxylation with lead tetraacetate. However, S. WOLFFE observed H/D scrambling when using maleic anhydride for the synthesis of 1a (Wolfe and Campbell 1979). Furthermore, our own preliminary experiments (Hintze 1984), in agreement with findings in the literature (Cimarusti and Wolinsky 1968), showed that the reactions with lead tetraacetate yielded products with benzene contents of up to 50 percent. Instead of a time-consuming and loss-prone purification via preparative gas chromatography or via the tetrabromo adduct (Wibaut and Haak 1948), preference was given to the following alternative developed by W. P. NORRIS (Norris 1968):
Although the cycloaddition proceeds with only moderate yields — variations in experimental parameters did not yield any significant improvements (Hintze 1984) — thermal intramolecular β-elimination, however, yields a very pure product (> 99.5%) in good yields and under mild conditions, so that further purification was not necessary. Our own earlier studies (Hintze 1984) using the isotopomers 1a and 1b were also able to rule out H/D scrambling during the synthesis. The position of the deuterium atoms in 1,4-cyclohexadiene was determined from the integration and coupling patterns in the 1H-NMR spectra, as well as from the number and multiplicity of the signals in the noise- and off-resonance-decoupled 13C-NMR spectra. The degree of isomerization of the isotopomers was determined by mass spectrometric isotope ratio measurements (see Section 4.1). The steric arrangement of the deuterium atoms in the isotopomers 1c and 1d was determined by several independent methods (see Section 4.3).
2.2. Syntheses of Labeled 1,3-Butadienes
The syntheses of the indexed 1,3-butadienes were carried out according to methods known from the literature or, for individual steps, were performed by analogy with literature reports. In doing so, the procedures described in the literature—some of which were incomplete and/or incorrect—had to be optimized to meet the high standards for the degree of labeling and steric purity required for our investigations (see Experimental Section).
[1,1,4,4-D4]-1,3-butadiene
The degree of indication was increased to 99.6% D4 (Ref. (Cope et al. 1961) 95.5% D4).
[2,3-D2]-1,3-butadiene
Hexadeuterio-1,3-butadiene was synthesized by analogy to the method described by CRAIG and FOWLER (Craig and Fowler 1961), with the yield increased to > 90%. After repeated D/H exchange, the butadiene dideuterated at the 2,3-position exhibited a deuterium content of 98.4% D2, which is significantly higher than the value of 86.0% D2 reported in the literature (Charlton and Agagnier 1973). This isotopomer is thus present at the purity required for kinetic measurements.
[(Z)-1-D1]-1,3-butadiene
The reduction of (Z)-1-chloro-1,3-butadiene was carried out in accordance with a procedure described by STEPHENSON (Stephenson et al. 1977). The degree of indication of 95.0% D1 reported by P. MÜLLER (Müller et al. 1984) was increased to 99.3% D1.
[(Z,Z)-1,4-D2]-1,3-butadiene
(Z,Z)-1,4-Dichloro-1,3-butadiene was obtained in pure form (97.5%) by distillation from the isomer mixture prepared according to PORRI and AGLIETTO (Porri and Aglietto 1976). The reduction according to STEPHENSON (Stephenson et al. 1977) to deuterated butadiene proceeded with increased yield and retention: according to the 1H-NMR spectrum, the proportion of the (E,Z)-isomer was < 5% at a degree of deuteration of 96.0% D2 (determined in the CHD). This synthetic route to (Z,Z)-configured 1,3-butadiene had proven difficult to reproduce in the literature with regard to stereochemical purity and the degree of deuteration: P. MÜLLER (Müller et al. 1984) was only able to synthesize the desired isotopomer using this method at a purity level that was unusable for our purposes. The proportion of the (E,Z) isomer was approximately 30% with a degree of labeling of approximately 85% D2.
[(E,Z)-1,4-D2]-1,3-butadiene
The synthetic route analogous to that for the (Z,Z)-isomer via (E,Z)-dichlorobutadiene did not appear feasible, since, according to STEPHENSON, the reduction of E-configured chlorobutadienes proceeds with partial loss of stereochemistry (Stephenson et al. 1977). P. MÜLLER also obtained only 1d via this route, which was contaminated with approximately 30% 1c and had a D2 content of only 85% (Müller et al. 1984), which is insufficient for kinetic measurements. Therefore, this butadiene was synthesized in sufficient purity using the synthetic route described by I. FLEMING (Fleming and Wildsmith 1970).
The (E,Z)-configured 1,3-butadiene is obtained through a stereospecific, palladium-catalyzed cis-deuteration of a cyclobutene unit incorporated into a tricyclic system, followed by two successive pericyclic reactions (retro-Diels-Alder reaction, conrotatory ring opening). However, this synthesis is only incompletely documented in the literature with experimental data and has also proven to be difficult to reproduce. For example, P. MÜLLER obtained a product with an insufficient degree of deuteration of 85% D2 (Hagemann et al. 1985). In contrast, the values of 92% (and 95% in the second run) achieved in this work are sufficiently high for kinetic measurements and correspond approximately to the value of 96% D2 reported by FLEMING (Fleming and Wildsmith 1970). The stereochemical purity was determined by NMR spectroscopy. The ratio of E- to Z-hydrogens in the 1H-NMR spectrum corresponded to the theoretical value of 1.00. However, this analysis should be viewed critically, since an equimolar mixture of the (E,Z) and (E,E) isomers yields the same integration ratio. Additional confirmation of stereochemical purity is provided by the pyrolysis of [trans-3,6-D2]-1,4-cyclohexadiene to > 85% (see Section 4.3).
2.3. Alternative Synthesis Route for Stereoselectively Labeled 1,4-Cyclohexadienes
The objective of the following synthetic route was to prepare the isotopomers 1c and 1d from a precursor that is readily available for both compounds. The target molecules should be obtained by nucleophilic or electrophilic substitution of 1,4-cyclohexadienes bearing suitable leaving groups (R) at the 3,6-positions.
Among the 1,4-cyclohexadienes known from the literature that are disubstituted at the 3- and 6-positions, the bistrimethylsilyl derivative is a suitable candidate; its synthesis and separation into stereoisomers have been described (Keil and Effenberger 1982). The silyl groups can be substituted under both basic and acidic conditions. The basic hydrolysis described by DUNOGUES et al. (Dunogues et al. 1972) was applied by PLATT and OESCH (Platt and Oesch 1977) for the synthesis of tritium-labeled 1,4-cyclohexadiene from benzene and 3H₂O; the mechanism and stereochemistry of the reaction have not been investigated in detail. In contrast, there are several recent studies on the stereochemistry of SE’-reactions involving cyclic and acyclic allylsilanes (Chan and Fleming 1979). T. HAYASHI et al. (Hayashi, Konishi, et al. 1982),(Hayashi, Ito, et al. 1982) were able to demonstrate anti-stereoselectivity for the protodesilylation of a chiral, acyclic allylsilane with trifluoroacetic acid [D]. They obtained the same results in other SE’-reactions of chiral cyclopentene and cyclohexene allylsilanes (Hayashi et al. 1983). The protodesilylation of 3,6-bistrimethylsilyl-1,4-cyclohexadiene is not reported in the literature. In this work, [3,6-D2]-1,4-cyclohexadiene was to be synthesized via both routes according to the following scheme:
While the reaction of trans-3,6-bistrimethylsilyl-1,4-cyclohexadiene with potassium deuteride proceeds uniformly, protodesilylation yields a 1:1 mixture of the two isomeric cyclohexadienes. The formation of 1,3-cyclohexadiene can be explained by a deuterium attack at the 5-position of the monosilyl derivative formed as an intermediate:
After isolation, the deuterated 1,4-cyclohexadienes were analyzed for steric purity using preparative gas chromatography (see Section 4.3). Mass spectrometric and 1H-NMR spectroscopic analysis show that more than 90% of both products are dideuterated at the 3,6-position. In contrast, stereochemical studies reveal that in both reaction products, the isotopomers 1c and 1d are present in a ratio of approximately 1:1. The lack of stereoselectivity could be due to the fact that, in this cyclic system, the optimal antiplanar arrangement of the electrophile and the leaving group is not possible (Rabideau et al. 1985), or that the extended π-system influences the substitution mechanism. Since neither synthesis exhibited any discernible stereoselectivity during the substitution, the reaction conditions were not optimized.
2.4. Summary of the Synthesized 1,4-Cyclohexadienes
By reacting the 1,3-butadienes obtained in Section 2.2 according to the procedure described in Section 2.1, all required isotopomers were thus synthesized with the purity necessary for kinetic and stereochemical studies. The cyclohexadienes obtained, along with their deuterium content and steric purity, are summarized below. The chemical purity was > 99.5% in all cases:
| 1a | 1b | 1c | 1d | 1e | |
|---|---|---|---|---|---|
| % Dn | 99.6 D4 | 99.3 D1 | 96.0 D2 | 95.0 D2* | 98.6 D2 |
| cis(trans) | > 90% | > 85% |
*: 91.6% D2 in a preceding synthesis
3. Syntheses of Deuterated Benzenes
Isotopomerically pure benzenes were required as reference substances for the mass spectrometric analysis of mixtures of labeled benzenes. Monodeuterated benzene was synthesized via the Grignard compound obtainable from bromobenzene:
The degree of labeling is sufficient at 95.I% D1 and is higher than the 91.5% D1 reported in the literature (Todd 1973). Other synthetic routes yielded, in some cases, lower degrees of deuteration (reduction of iodobenzene with D2O and zinc/copper (Stephenson et al. 1977), electrochemical reduction of bromobenzene (Koppang et al. 1986); 90–95% D1) or were experimentally too complex (deuterolysis of phenyltrimethylstannane (Asomaning et al. 1973), > 99% D1; photolysis of halogenated benzenes (Mueller et al. 1976), > 98% D1). [1,4-D2]-benzene and [1,2-D2]-benzene were obtained by quantitative dehydrogenation of the isotopomers 1a and 1e with DDQ (see Chapter 6). The labeling efficiency is derived from that of the cyclohexadienes used (1a: 99.6% D2; 1e: 98.6% D2). This synthetic route yields significantly higher degrees of deuteration than the preparation of [1,4-D2]-benzene from the Grignard compound (Angelini et al. 1980) described in the literature (88% D2, 8% D1, 4% D0).
4. Analysis of Deuterated Compounds
4.1. Mass Spectrometric Isotope Ratio Measurements
4.1.1. Methods for Isotope Analysis
Isotope ratio measurements of the deuterated substances were performed, depending on the compound and the required measurement accuracy, at excitation energies of 70–80 eV or under low-voltage conditions at approximately 18 eV. Only the molecular peak groups were measured and evaluated, since hydrogen and skeletal rearrangements can occur during the formation of fragment ions under the distorting influence of H/D isotope effects.
Low-voltage method
In aromatic hydrocarbons and their fully or partially hydrogenated derivatives, strong [M-n]+ peaks occur at the usual ionization energies as a result of hydrogen abstraction. Particularly in the case of the labile 1,4-dihydroaromatics, these peaks approach the intensity of the M+ peak, thereby complicating reliable analysis. Under low-voltage conditions, with excitation energies just above the ionization potential, these fragmentations are suppressed to approximately 4% of the relative intensity of the M+ peak in 1,4-cyclohexadiene and are completely suppressed in benzene. A disadvantage of this method is that the greatly reduced ion yield affects the detection limit of the ions and thus the measurement accuracy, which must be compensated for by using relatively large amounts of sample (at least 0.5 mg) and increasing the cathode heating. The consequences are rapid fouling of the spectrometer and a shortened cathode lifespan. For technical reasons, GC/MS measurements cannot be performed under low-voltage conditions at our institute. Sample introduction was therefore performed using the direct injection method, i.e., the compounds were analyzed as pure substances. To do this, it was necessary to isolate them from the reaction mixtures using preparative gas chromatography. The separated substance was condensed at the preparative outlet into specially manufactured sample vials. Due to the high volatility of the substances, the samples were connected to the mass spectrometer’s push rod via a coolable sample sleeve immediately after the condensation process. By cooling with liquid nitrogen, a continuous ion current suitable for measurement could be established at pressures of < 10-6 Torr, which allowed for a slow scan of the mass range of interest. The resulting high measurement accuracy is offset by the extreme time required for the method—approximately two hours per sample (including preparative GC). Another disadvantage of this method is the time lag between the experiment and the analysis: Only a few measurement slots are available because the instrument is unavailable for routine GC/MS operation for an extended period due to the equipment changeover. For each measurement session, a large number of unprocessed reaction products therefore had to be collected over an extended period, which severely limited efficient experimental planning that incorporated isotope analysis. The low-voltage method was therefore reserved for cases in which the GC/MS method we developed at 70 eV did not yield sufficiently precise results.
GC/MS measurements at 70 eV
The high ionization energy in this method leads to pronounced [M-n]+ peaks; however, these can be accounted for arithmetically if the molecular peak groups of the pure isotopomers — which are present in the samples as mixtures — are known and sufficiently reproducible. In the GC/MS coupling, the C6 hydrocarbons were separated as a reaction mixture in a low-boiling solvent using a capillary column. For each GC peak, 4–5 mass spectra were recorded; the signals were stored using the CAT method and then averaged across all spectra. This procedure compensates for systematic errors caused by isotopomer retention in the capillary as well as significant variations in the ion current on the flanks of the GC peaks. The [M-n]+ peaks for benzene and cyclohexene were < 20% and 30%, respectively, and were reproducible with unchanged instrument settings (ionization energy, focusing). The molecular peak groups proved to be characteristic of each of the benzene isotopomers synthesized for comparison, as evidenced by the relative intensities of the individual peaks. At the beginning of each measurement session, the molecular peak patterns of the reference samples were therefore determined, and quantitative analyses of the reaction products were performed using these values, taking the indexing degrees into account. For each sample, the GC/MS data from at least ten such determinations were evaluated to average out the greater variation in the GC/MS data. The time required for a single GC/MS determination of a sample, including evaluation, is approximately 10 minutes. The percentage content of the isotopomers could be determined with sufficient accuracy (+/‒ 1.5% abs.) with an acceptable investment of time and material. However, the applicability of the GC/MS method to C6 hydrocarbons is limited by the following factors: - Quantitative analyses of benzene are only possible if it consists exclusively, or at least predominantly, of the isotopomers available as reference substances. - The cyclohexene GC peaks are too sharp to provide a sufficient number of spectra. Furthermore, the isotopomers required for the reference samples are difficult to obtain. - For cyclohexadiene, the [M-n]+ peaks (>> 50% of the M+ peak) are too intense to be taken into account in the calculations. - For samples with a high excess of cyclohexadiene relative to benzene, separation problems occurred due to column overload. When smaller amounts of the substance were injected, the peak intensities of benzene were too low to achieve sufficiently high measurement accuracies.
4.1.2. Error Analysis
Systematic errors in the described methods are primarily caused by memory peaks (instrument contamination from the previous sample during push-rod measurements) and by peak discrimination (= overemphasis of low-intensity peaks; inertia of the compensation recorder during low-voltage measurements; exceeding the measurement range during GC/MS measurements). For push-rod measurements, residual intensities were therefore checked between samples, and, if necessary, the procedure was paused until these were no longer detectable. Peak discrimination was ruled out by checking the 13C peaks of the non-indexed substances and, if necessary, eliminated by reducing the scan speed. Comparative measurements using both methods on a single sample at different measurement dates yielded identical compositions of the mixture within the error limits. Both methods are therefore, in principle, to be regarded as equivalent. The absolute statistical error for the percentage composition of an isotopomer mixture was approximately 0.5 – 1.0 percent for low-voltage and approximately 1.0 – 1.5 percent for GC/MS measurements. The errors represent the standard deviations of a sample obtained from at least ten measurements: \(s{_x} = (x{_i} - x){^2}/(n - 1){^{1/2}}\). The relative error of the percentage isotopomer content therefore increases as peak intensity decreases. For low-voltage measurements, this trend applies only to a limited extent, since with this method the relative error increases as peak intensity increases.
4.2. Gas-Chromatographic Isotopomer Separation
The separation of isotopes and isotopomers of simple molecules using gas chromatography is a long-established and frequently used method for isotope analysis and isotope enrichment (Atkinson et al. 1967). In addition to complexation chromatography (Schurig and Wistuba 1983),(Schurig 1976) — which is particularly effective but also very labor-intensive and limited to specific problems (the separation effect is based on isotope effects during reversible complexation in specially synthesized stationary phases) — separation is usually based on relative mass differences (Bruner et al. 1969),(Bruner et al. 1972). Thus, separations of deuterated naphthalenes on capillary columns were also observed in our own research group. For the C6 hydrocarbons investigated in this study, this method appeared to be a viable alternative to wet-spectrometric isotope analysis due to the larger relative mass differences. A prerequisite for successful separations is an extremely long capillary with a small inner diameter and uniform distribution of the stationary phase (Schulte 1984), whereby the choice of the optimal phase polarity is of particular importance. For practical reasons, the initial separation experiments were conducted using a nonpolar silicone phase (GC-SE-52), which was deposited within a 130-meter-long (0.2 mm inner diameter) glass capillary using a 0.2% solution in n-pentane. The chromatograms shown in figure 1 depict the (optimized) separations of mixtures of cyclohexadiene (unlabeled and [3,3,6,6-D4]-1,4-CHD) and benzene ([D6]-benzene and unlabeled benzene).
While the isotopomers of 1,4-cyclohexadiene leave the capillary completely separated, hexadeuterobenzene appears only as a shoulder in the benzene peak. The separation efficiency therefore does not correlate with the relative mass differences of 5.0 and 7.7 percent, respectively. Since the deuterated compounds also exhibit shorter retention times, the separation is based on other effects — probably the different polarities of the molecules. Further attempts to separate hexadeuterobenzene/benzene mixtures on moderately polar (OV 225, 50 m quartz capillary) and highly polar (FFAP, 90 m glass capillary) phases revealed no separations. Since mechanistic studies require separations of benzene mixtures containing low-indexed substances, the experiments on gas chromatographic isotopomer separation were not continued. However, the good separation of the 1,4-cyclohexadienes is of only limited use for determining kinetic isotope effects via intermolecular competition experiments (see Section 5.2). The determination of isotope effects in the unreacted reactant is, in principle, subject to higher systematic errors and was therefore not performed in this work.
4.3. On the Stereochemistry of [cis-3,6-D2]- and [trans-3,6-D2]-1,4-Cyclohexadiene
The characterization of the stereoselectively deuterated cyclohexadienes must include not only the isotope content and position but, above all, the stereochemical arrangement of the isotopes. Therefore, it was necessary to confirm the stereochemistry determined by 1H-NMR spectroscopy in the 1,3-butadiene precursors (which was unambiguous only for the [(Z,Z)-D2] isotopomer) through further investigations of the labeled cyclohexadienes. Although H/D scrambling does not occur during the synthetic route (see Section 2.1) determining the stereochemistry in the cyclohexadiene was intended to rule out the possibility of thermal cis/trans isomerization of the butadienes via 1,2-diradicals (Stephenson et al. 1972),(Stephenson et al. 1975) having occurred during the cycloaddition. By 1H-NMR spectroscopy, the stereoisomers 1c and 1d can be distinguished — due to the high molecular symmetry — only on the basis of the homoallylic 5JH,H coupling constants for cis-positioned (5Jcis = 9.63 Hz) and trans-positioned (5Jtrans = 8.04 Hz) hydrogen atoms in the 13C-H satellite spectrum. These data were obtained from measurements of the BIRCH reduction product of hexadeuterated benzene (1h, a 1:1 mixture of cis and trans) under heteronuclear broadband decoupling of the deuterium atoms (Grossel and Perkins 1979),(Grossel 1980),(Garbisch and Griffith 1968). Measurements (Bandmann 198y) on the 1h isotopomer mixture — obtained via a basic SN reaction from hexadeuterated trans-3,6-bistrimethylsilyl-1,4-cyclohexadiene (analogous to the synthetic route described in Section 2.3) — confirm these findings and simultaneously demonstrate the stereo-unspecific nature of the substitution (Fig. 2).
The figure shows the higher-field side of the 13C satellite spectrum of the methylene protons. The absorption at δ = 2.20 ppm is due to traces of 1,3-cyclohexadiene present as an impurity. The two outer absorption lines at 2.28 and 2.32 ppm correspond to the cis-positioned hydrogen atoms, while the two inner peaks correspond to the trans-positioned hydrogen atoms at the 3- and 6-positions of 1,4-cyclohexadiene. This analytical method was also applied to a mixture of 1c and 1d obtained in the same manner (Fig. 3). Shown here is the complete 13C satellite spectrum of the methylene protons.
The coupling with the vinyl hydrogens, which is effective here, leads to a strong splitting of the signals, which does not allow for an accurate analysis of such isotopomer mixtures (additional decoupling is not possible for technical reasons).
4.3.1. IR and Raman Spectroscopic Analysis
Vibrational spectroscopic studies of 1,4-cyclohexadienes have been conducted in the sixties and seventies exclusively on non- and perdeuterated derivatives as conformational analyses to distinguish between the discussed molecular symmetries C2v and D2h (Stidham 1965),(Laane and Lord 1971),(Carreira et al. 1973). Together with force-field (Lipkowitz et al. 1982) and ab initio calculations (Saboe and Boggs 1981), as well as 1H-NMR studies of the homoallylic coupling constants (see Section 4.3), these results support a planar conformation (D2h symmetry). The Raman spectra of the isotopomers 1a, 1b, 1d, 1g, as well as a 2:1 mixture of 1c:1d, published by H. HAGEMANN et al. (Hagemann et al. 1985), confirm these results. The spectral analysis was performed not from a stereochemical perspective but rather from a spectroscopic perspective regarding band assignment. Compared to that of the pure trans isomer, the spectrum of the stereoisomer mixture showed an additional band at 835 cm‒1. Although the authors emphasized their usefulness, IR spectra could not be recorded due to the limited amounts of substance (a few μL!). In this work, the IR and Raman spectra of the isotopomers 1c and 1d, synthesized as described in (see Section 2.1), were recorded for the purpose of stereochemical analysis (Schnöckel 198z). Due to the different molecular symmetries (inversion center in 1d), significantly different spectra were expected, particularly in the “fingerprint” region and, to a lesser extent, in the C–H and C–D valence vibrations. The overview spectra are shown in the appendix. The clearly distinguishable regions of both isotopomers are shown in Figures 4 (IR spectra) and 5 (Raman spectra). The bands of each of the other stereoisotopomer, present as impurities, are marked with arrows.
It is evident, particularly from the IR fingerprint regions, that the two cyclohexadienes were obtained with high stereochemical purity. The impurity level of the respective other stereoisomer is estimated to be less than ten percent. For quantitative measurements, the plotting of a necessary calibration curve would have required too much material. Unfortunately, the differences in the Raman spectra are too small for such analyses.
4.3.2. Pyrolysis of Labeled 1,4-Cyclohexadienes
The thermal elimination of hydrogen from 1,4-cyclohexadiene is a reaction permitted by the WOODWARD-HOFFMANN rules and has been well studied in the past: kinetic studies by ELLIS and FREY (Ellis and Frey 1966), as well as by FREY, KRANTZ, and STEVENS (Frey et al. 1969), tracer studies by TARDY, GORDON, and NORRIS (Tardy et al. 1976) using per- and hexadeuterated CHD, and stereochemical studies by FLEMING and WILDSMITH (Fleming and Wildsmith 1970) using 1d demonstrate a homogeneous, unimolecular, and cis-specific hydrogen elimination, for which the following transition state is likely:
To corroborate the spectroscopic findings, the isotopomers 1c and Id, as well as 1b, were pyrolyzed at 340 °C. The values obtained from 1b allow for a correction of the analytical data for the stereoisomers with respect to their degree of indication. The resulting benzene was analyzed by mass spectrometry for its Dn composition of [D0]-, [D1]- and [D2]-isotopomeres. The steric purity of the reactants is determined from the analytical data (after correction for the D1 contents of the reactants): for 1c, from the percentage of the sum of [D0]- and [D2]-benzene, and for 1d, from [D1]-benzene. The table below shows the already corrected data for the reactants examined:
| 1,4-Cyclohexadiene | Percentage Stereoisomer | |
|---|---|---|
| cis | trans | |
| 1c: [cis-3,6-D2] | 90 | 10 |
| 1d: [trans-3,6-D2] | 15 | 85 |
| 1d: [trans-3,6-D2]⁕ | 14 | 86 |
| ⁕: reproduction | ||
| Products from trans-3,6-Bistrimethylsilyl-1,4-CHD: | ||
| treated with KOD/D2O | 51 | 49 |
| treated with CF3CO2D | 58 | 42 |
Gas chromatographic analysis of the pyrolysis products consistently yielded approximately 7% cyclohexane and cyclohexene. Accordingly, in addition to intramolecular hydrogen elimination, a radical, sterically nonselective disproportionation occurred to an extent of approximately 15 percent (Heesing and Müllers 1980a). This side reaction causes a systematic error in the reported analytical values, which gives the false impression of insufficient steric purity. Due to a lack of information regarding the isotopic effects occurring during pyrolysis, this aspect was not further included in the evaluation. Within the margin of error, the results provide a picture consistent with the other findings and the steric purity of the butadiene precursors.
5. H/D Isotope Effects
5.1. Fundamentals of Isotope Effects
5.1.1. Primary H/D Isotope Effects
If a C–H bond in a molecule is replaced by a C–D bond, the vibrational frequency of the bond in question — and thus its zero-point energy — decreases due to the increased reduced mass. If this bond is broken in the transition state of a reaction step, the activation energy for this process is, in a first approximation, greater by the difference in zero-point energies in the reactants. The indexed substance therefore reacts more slowly due to this primary isotope effect. In semiclassical primary isotope effects, H-transfer occurs exclusively along the potential hyperplane. They can be calculated from the molecular data of the reactants and the activated complex, based on the fundamental work by EYRING, LAIDLER, and GLASSTONE (Glasstone et al. 1941) on transition-state theory. MELANDER (Melander and Saunders 1980) and WESTHEIMER (Westheimer 1961) demonstrated, using a three-center model for the transition state, that the magnitude of the primary isotope effect depends on the geometry and symmetry of the activated complex:
Linear symmetric [A‒H‒B]
This transition state exists only if A and B represent identical groups. The H atom participates in no vibration other than the decay motion, so that the difference in activation energies [ΔEA]H/0 is equal to the difference in zero-point energies. The isotope effect is then calculated to be 6.9 at 25°C (Melander and Saunders 1980) (page 130).
Linear asymmetric [A-H‒B]
If A and B are different groups, one of the bonds (here the A–H bond) remains strengthened in the transition state. It follows that [ΔEA]H/D < [ΔE0]H/D. The isotope effect at 25°C ranges from 1.4 to 6.9.
Angled [A‒H‒B, ⦛AHB < 180°]
The groups A and B can be identical or different here. In addition to the decay vibration, three deformation vibrations can be excited in the transition state, so that [ΔEA]H/D < [ΔE0]H/D. At 25°C, the isotope effect lies within the same range as for a linearly asymmetric complex. Small semiclassical isotope effects can also be caused by the presence of heavy atoms, which lead to an increased reduced mass.
In quantum mechanical tunneling, H transfer occurs even from unexcited vibrational states when the energy barrier is sufficiently narrow, as a result of the overlap of the vibrational wave functions of the reactant and product. The contribution of tunneling to H transfer depends on the probability of the H atom being present in the overlap region and does not constitute an independent mechanism. Rather, it can be involved to a greater or lesser extent in every H transfer. As a mass-dependent phenomenon, it amplifies the semiclassical isotope effect and can lead to values of >> 6.9 at 25 °C (Melander and Saunders 1980) (page 140). It follows from this that no conclusions regarding the transition state may be drawn based solely on the magnitude of isotope effects, especially since tunneling contributions may be present even at values < 6.9. However, these are in principle recognizable by the temperature dependence of the isotope effect (see Section 7).
5.1.2. Secondary Isotope Effects
Secondary isotope effects occur when the bond to the isotope is preserved, but the bond ratios have changed in the transition state. Depending on the position of the isotope relative to the dissolved C–X bond, a distinction is made between α, β and γ effects (D is the isotope):
sec-α effect: C-CD-X / sec-β effect: CD-C-X / sec-γ effect: CD-C-C-X
Their magnitude depends on the extent to which the C-D bond is weakened (1.0–1.3) or strengthened (0.8–1.0) in the transition state (Shiner et al. 1968),(Llewellyn et al. 1960),(Halevi 1963). While α-effects are usually attributable to rehybridization of the C atom (Streitwieser et al. 1958), inductive (Shiner and Humphrey 1963), hyperconjugative (Sunko et al. 1977), and steric effects (Shiner et al. 1963) often play a role in β-effects as well. If molecules are indexed multiple times at a single position, or if there are equivalent positions of which only one is indexed, multiple isotope effects come into play simultaneously. The observed total isotope effect is then the product of the individual isotope effects, since the activation energies are additive.
5.2. Measurement Methods
5.2.1. Total Kinetic Isotope Effects
If a reaction step is rate-determining in a reaction, the isotope effect at work in that step is referred to as the total kinetic isotope effect. In this work, these are determined in two different ways.
a) Determination of rate constants
Rate constants for the labeled and unlabeled substances are determined in separate experiments using kinetic measurements. The total isotope effect is obtained by dividing the constants kH and kD. Isotope effects in upstream or downstream fast steps or equilibria can also affect the kinetic data as a result of equilibrium isotope effects, even though these steps are not rate-determining. A prerequisite for this method is very good reproducibility of the kinetic data in order to keep the error from the isotope effect small. The determination of very large isotope effects (> 5) is critical, as systematic errors become more pronounced as the difference in reaction times between the labeled and unlabeled substances increases. Another disadvantage is the high amount of substance required for this method.
b) Intermolecular competition
This method also captures only the rate-determining step of the reaction. The labeled and unlabeled substances are simultaneously reacted in competition with each other in a known ratio; the reaction is stopped after a conversion sufficient for the analysis, and the ratio of the products formed from the isotopomers is determined. If the reactants were present as a 1:1 mixture and the conversion was < 10 percent, this ratio of the products corresponds to the isotope effect kH/kD. As the conversion increases, the labeled reactant accumulates in the mixture as a result of isotope effects, leading to distorted, underestimated values for kH/kD. A correction can be made using BIGELEISEN’s formula (Melander and Saunders 1980) (page 95), which also accounts for any ratio of the reactants. The advantages over kinetic analysis are the reduced material and time requirements, as well as the extensive elimination of systematic errors in reaction control, since both isotopomers are affected to the same extent in a first approximation.
5.2.2. Product Isotope Effects
Intramolecular competition occurs in labeled substances in which two or more equivalent positions are substituted with different isotopes. These compete intramolecularly during abstraction, resulting in a mixture of products with different indices. The resulting product isotope effect arises from their relative proportions, which, in a multistep reaction, can be influenced by the isotope effects of all reaction steps.
6. Hydrogen Transfer to Quinones
6.1. Literature Findings and Selection of Redox Systems
Fundamental studies from the 1950s on the mechanism of thermal hydrogen transfer from dihydroaromatics to quinones by BRAUDE et al.(Braude, Jackman, et al. 1954),(Braude, Brook, et al. 1954),(Braude, Jackman, Linstead, and Shannon 1960),(Braude, Jackman, Linstead, and Lowe 1960a),(Braude, Jackman, Linstead, and Lowe 1960b) and BANARD (Barnard and Jackman 1960) suggested a two-step ionic mechanism:
\[ \ce{RH_2 + Q -> RH^+ + QH^- -> R + QH_2} \]
This mechanism was supported by the dependence of the reaction rate on the solvent and the oxidation potential of the quinone, the reaction order, and the proton-catalyzed reaction in the case of quinones with low reactivity. TROST (Trost 1967) (acenaphthene system) and, within our own research group, PAUKSTATT and BROCK (Paukstat et al. 1985) were able to confirm this mechanism for the dehydrogenation of 1,2-dihydroaromatics. Furthermore, the regio- and stereoselectivities determined using a series of deuterated 1,2-dihydronaphthalenes demonstrate the existence of a close ion pair as an intermediate. For 1,4-dihydroaromatics, THUMMEL et al. [(Thummel et al. 1980)] found a clear dependence of the aromatization tendency on -I and +I substituents and attributed this to an ionic reaction mechanism. In addition, a SET in the rate-determining step — proposed in 1976 by HASHISH and HOODLESS (Hashish and Hoodless 1976) and refuted in 1980 by MÜLLER (Müller and Joly 1980) — and, above all, two synchronous mechanisms for the dehydrogenation of 1,4-cyclohexadiene have been discussed: a pericyclic mechanism permitted by symmetry rules and favored by steric factors, and a concerted termolecular H-transfer mechanism involving the solvent:
The following experimental findings were cited to support a synchronous hydrogen elimination in 1,4-cyclohexadiene:
The reactivity order 1,4-CHD > 1,3-CHD > cyclohexene and the increased reactivity compared to other dihydro compounds that yield nonaromatic products suggest considerable aromatic stabilization of the transition state (Müller and Rocek 1972)
The reactivity hierarchy cis-3,6-dimethyl-1,4-CHD > 1,4-CHD > trans-3,6-dimethyl-1,4-CHD >> 3,3-dimethyl-1,4-CHD, as well as the identical behavior of these cyclohexadienes toward DDQ and the triphenylmethyl cation, support the termolecular mechanism (Stoos and Rocek 1972),(Müller 1973).
The isotopomers 1h and 1i (see Section 2.1)) exhibit greater isotope effects upon dehydrogenation with 2,3-dichloro-5,6-dicyano-1,4-benzoquinone (DDQ) in benzene than in the equally rapid reaction with the triphenylmethyl cation in acetonitrile (DDQ/1h: 1.70, DDQ/1i: 10.0; (Ph)3C+/1h: 2.82: (Ph)3C+/1i: 4.20). The high value for 1i in the DDQ dehydrogenation is considered evidence of a pericyclic mechanism (Müller 1973).
Hydrogen elimination in 1,4-CHD proceeds with high cis-selectivity, as MÜLLER was able to demonstrate using enriched mixtures of the isotopomers 1c and 1d (Müller et al. 1984).
However, the insufficient purity of the stereoisomers and unsatisfactory analytical data call into question the conclusiveness of these results. Cis selectivity was also reported by CARTER et al. (Carter et al. 1981) for the DDQ oxidation of 1d — though without experimental data. The results regarding the cis-selectivity of the reaction largely refute the termolecular mechanism.
Although the findings listed are consistent with a synchronous mechanism, they do not prove it, as they can also be interpreted as a two-step process involving a narrow ion pair. Of interest in this context is the unspecified “suggestion” of stepwise dehydrogenation published by CARTER et al. (Carter et al. 1981). The role of intermediates in a rapid pre-equilibrium state also remains unclear:
\[ \ce{RH_2 + Q <=>[fast] Intermediate <=>[slow] R + QH_2} \]
Possible intermediates include charge-transfer complexes (Paukstat et al. 1985) and ene-products, which were observed in the reaction of 1,4-CHD with tetracyanoethene (Jacobsen 1980),(Haselbach and Rossi 1976). In the reaction of 1,4-CHD with DDQ, no intermediates were detected within the limits of detection (Müller and Joly 1983). Since rapid equilibrium establishment does not affect the reaction kinetics or isotope effects, these methods cannot distinguish between a true intermediate and a secondary equilibrium.
In this work, the isotopomers 1a–1e of 1,4-CHD were synthesized and used to thoroughly investigate the nature of the rate-determining step in the dehydrogenation reaction with 2,3-dichloro-5,6-dicyano-1,4-benzoquinone (DDQ) and 3,4,5,6-tetrachloro-1,2-benzoquinone (OCA):
These quinones were selected based on their reactivity and importance in preparative chemistry. They allow for a comparative study of ortho- and para-quinones, as well as the inclusion of literature findings on DDQ oxidation in the evaluation.
6.2. Reaction Conditions and Product Analysis
The reactions were carried out under standardized reaction conditions that were optimized in our own preliminary experiments (Hintze 1984) from both experimental and analytical perspectives. All experiments were conducted at 25°C in dioxane and, to investigate solvent effects, in some cases also in acetonitrile and N-methylformamide (NMF). The reactions are irreversible and proceed without the formation of byproducts. This fulfills the basic prerequisite for determining isotope effects — a uniform reaction course. A complex reaction mechanism (reversibility and parallel reactions), on the other hand, would carry the risk of abnormal pseudo-isotope effects (Thibblin 1983).
6.3. Determination of Total Isotope Effects
6.3.1 Kinetic Measurements
The rate constants were determined by UV spectroscopy by monitoring the time-dependent decrease in quinone absorbance. The dehydrogenations proceeded under pseudo-first-order conditions with respect to 1,4-cyclohexadiene, which was used in a 90- to 200-fold excess depending on the reactivity of the isotopomer: \[ -\frac{d[\text{Q}]}{dt} = k' \cdot [\text{Q}] \] with \(k' = k \cdot[\text{CHD}]\). Assuming the validity of Lambert-Beer’s law, the integrated rate equation is: \[ lnE-lnE_0 = -k' \cdot t \] The UV absorption (‘extinction’ E ) of the reaction mixtures DDQ/1,4-CHD and OCA/1,4-CHD, respectively, recorded at various times t in the wavelength range between 320 and 500 nm, are shown below:
The time-dependent decrease in DDQ absorbance was monitored at 390 nm; for evaluation, the measured absorbance values had to be corrected for the — albeit slight — absorption of the hydroquinone formed (see Experimental Section). The time-dependent increase in absorbance of the hydroquinone band at 350 nm could not be interpreted using a simple time law — even after correction for the decreasing quinone absorption (Hintze 1984) — and was therefore not evaluated. The decrease in absorbance of OCA was recorded at 425 nm and corrected for the very weak absorption of hydroquinone. The pseudo-first-order straight lines obtained for both quinones showed a very good correlation (r > 0.999) over several half-lives as exemplary shown in the figure.
Each rate constant determined in this manner is the statistical mean of at least four individual measurements. The errors of approximately 2% listed in Table 1 are the standard deviations determined for a sample (see Section 4.1.2).
| 1,4-Cyclohexadiene | k \(\mathbf{[L\cdot mol^{-1}\cdot s^{-1}]\cdot 10^2}\) | |
|---|---|---|
| Isotopomer | DDQ | OCA |
| 1 | 1.32 ± 0.03 | 0.492 ± 0.009 |
| 1a | 0.106 ± 0.02 | 0.0337 ± 0.0007 |
| 1b | 0.978 ± 0.02 | 0.365 ± 0.007 |
| 1c | 0.670 ± 0.01 | 0.250 ± 0.005 |
| 1d | 0.667 ± 0.01 | 0.241 ± 0.005 |
| 1e | 1.38 ± 0.03 | 0.558 ± 0.009 |
6.3.2 Intermolecular Competition
To investigate the overall kinetic isotope effects on systematic errors, the isotopomers 1a and 1b were each reacted with DDQ in intermolecular competition with unlabeled CHD. In the case of isotopomer 1b, intramolecular competition between H and D also occurs during hydrogen abstraction at the C3 carbon atom. The ratio of the benzenes formed in this process must be taken into account in the intermolecular competition and was therefore determined in an additional intramolecular competition experiment.
6.3.3 Values for the Total Isotope Effects
The total isotope effects determined using the methods outlined in the previous two sections are listed below. Values for the isotopes 1h and 1i are reported in the literature.
| Isotopomer | Quinone | IE type / technique | value |
|---|---|---|---|
| 1a | DDQ | kinetic | 12.2 ± 0.5 |
| 1a | DDQ | inter | 13.5 ± 1.5 |
| 1a | OCA | kinetic | 14.6 ± 0.6 |
| 1b | DDQ | kinetic | 1.35 ± 0.03 |
| 1b | DDQ | inter / intra | 1.33 ± 0.09 |
| 1b | DDQ | intra | 2.1 ± 0.06 |
| 1b | OCA | kinetic | 1.35 ± 0.04 |
| 1b | OCA | intra | 2.1 ± 0.08 |
| 1c | DDQ | kinetic | 1.97 ± 0.06 |
| 1c | OCA | kinetic | 1.98 ± 0.07 |
| 1d | DDQ | kinetic | 1.98 ± 0.07 |
| 1d | OCA | kinetic | 2.04 ± 0.09 |
| 1e | DDQ | kinetic | 0.95 ± 0.03 |
| 1e | OCA | kinetic | 0.88 ± 0.04 |
| 1h⁕ | DDQ | kinetic | 1.7 |
| 1i⁕ | DDQ | kinetic | 10.0 |
⁕: literature data from P. MÜLLER (Müller 1973)
6.4. Determination of the Stereochemistry of the Reaction
To date, there have been no satisfactory, systematic studies in the literature on the stereochemistry of the dehydrogenation of 1,4-cyclohexadiene by quinones (see Section 6.1). The dehydrogenation with DDQ was carried out by MÜLLER only with 2:1 mixtures of the stereoisomers 1c and 1d, which were contaminated with approximately 15% 1b, in dioxane (Müller et al. 1984). The low purity of the starting materials and the inaccuracy of the GC/MS analysis used in this study allow only the conclusion that the reaction exhibits high cis-selectivity. Further information also does not include the unpublished results on the DDG dehydrogenation of 1d cited by CARTER et al. in a footnote (Carter et al. 1981).
In this work, the isotopomers 1c and 1d were reacted in intramolecular competition with DDQ and OCA at 25°C in three solvents of varying polarity, and the isotopomer patterns of the dehydrogenation products were determined via GC/MS coupling. In the case of exclusive cis abstraction, a mixture of only [D2]- and [D0]-benzene is expected for 1c, and only [D1]-benzene for 1d. In Table 6.4 the measured values are listed as relative fractions of the theoretical values for an isotopomer’s exclusive cis elimination. The error in the percentage values is ±1.5 percent (absolute). If the data are based on a steric purity of the reactants of > 90% cis for 1c and > 86% trans for 1d (see Section 4.3), dehydrogenation with OCA occurs highly stereoselectively in nonpolar and moderately polar solvents, and in the highly polar N-methylformamide (NMF), approximately 95% is still achieved as the abstraction of two cis-adjacent hydrogen atoms. In contrast, the DDQ exhibits significantly lower stereoselectivity in the dehydrogenation reaction, which also decreases sharply with increasing solvent polarity. A detailed discussion of the observed stereoselectivities, as well as the differing values for 1c and 1d, is provided in (Section 6.6).
| solvent, rel. permittivity | Quinone / Isotopomer | ||||
|---|---|---|---|---|---|
| ε-values: (Reichardt 1969) | DDQ | OCA | |||
| ε | 1c | 1d | 1c | 1d | |
| 1,4-Dioxane | 2.21 | 82 (91) | 71 (83) | 90 (> 95) | 88 (> 95) |
| Acetonitrile | 37.5 | 80 (89) | 60 (70) | 91 (> 95) | 88 (> 95) |
| NMF | 182.4 | 71 (79) | 56 (65) | 87 (95) | 84 (93) |
NMF = N-methylformamide
6.5. Individual Isotope Effects
6.5.1 Calculation of the Individual Isotope Effects of the Rate-Determining Step
The total isotope effects obtained in Chapter 6.3 are composed of the individual isotope effects (primary, sec-α, sec-β and sec-γ ‒ see Section 5.1) in a manner that takes into account not only the number and position of the deuterium-labeled hydrogens but also whether the reaction is single- or two-step as well as the steric course of hydrogen abstraction. The equations listed below apply to the relationship between the total and individual effects for a one-step (synchronous elimination of two cis-positioned hydrogens) and a two-step (elimination of one hydrogen atom in the rate-determining step) process. The stereochemistry of the reactants and that of the reaction are irrelevant in the first step when calculating the individual isotope effects for the two-step mechanism.
Single-step mechanism:
kH is the rate constant for attack from one side of the cyclohexadiene.
Two-step reaction:
kH is the rate constant for the abstraction of one of the four equivalent hydrogen atoms in cyclohexadiene.
| Isotopomer | Single-Step | Dual-Step |
|---|---|---|
| 1 | \(k = 2k_\text{H}\) | \(k = 4k_\text{H}\) |
| 1a | \(k = 2k_\text{H}\cdot (p \cdot α)^{-2}\) | \(k = 4k_\text{H}\cdot (p \cdot α)^{-1}\) |
| 1b | \(k = k_\text{H}\cdot (p^{-1} + α^{-1})\) | \(k = k_\text{H}\cdot (2 + p^{-1} + α^{-1})\) |
| 1c | \(k = k_\text{H}\cdot (p^{-2} + α^{-2})\) | \(k = 2k_\text{H}\cdot (p^{-1} + α^{-1})\) |
| 1d | \(k = 2k_\text{H}\cdot (p \cdot α)^{-1}\) | \(k = 2k_\text{H}\cdot (p^{-1} \cdot α)^{-1}\) |
| 1e | \(k = 2k_\text{H}\cdot (β \cdot γ)^{-2}\) | \(k = 4k_\text{H}\cdot β^{-1} \cdot γ^{-1}\) |
Using approximation, values for p and α were calculated for both mechanisms such that the sum of the squares of the relative errors compared to the experimental total isotope effects is minimized. The table below lists the calculated individual isotope effects for which the deviations are minimal.
| Quinone | IE | Dual-Step | Synchronous |
|---|---|---|---|
| (< 3) | (3,000) | ||
| DDQ | p | 11.5 | 2.5 |
| α | 1.08 | 1.1 | |
| β·γ | 0.95 | 0.97 | |
| OCA | p | 13.6 | 2.7 |
| α | 1.08 | 1.0 | |
| β·γ | 0.88 | 0.94 | |
A good correlation of the measured total isotope effects is only observed with a two-step mechanism, as indicated by the identical values of the two stereoisotopomes . The deviations of the calculated values from the experimental values, assuming a synchronous process, are a factor of 1,000 larger and lie far outside the experimental error limits.
6.5.2 Discussion of the Individual Isotope Effects of the Rate-Determining Step
The calculated individual isotope effects of the two-step process can be interpreted as follows, based on the theoretical principles described in Section 5.1):
The high primary isotope effects lie outside the range for semiclassical isotope effects. The first step of hydrogen abstraction therefore proceeds with a significant contribution from tunneling.
The magnitude of the secondary α effects corresponds to a rehybridization of the C atom from sp³ to sp² in the first step of the reaction.
The inverse total isotope effects of 1e consist of secondary β- and γ-effects. Both are inverse because only inductive effects are effective. The two effects that would lead to a normal isotope effect are absent. Hyperconjugation of the β-C-H(D) bond with the free p orbital that forms on the α-C atom in the rate-determining first step is impossible for steric reasons (Sunko et al. 1977),(Braude, Jackman, Linstead, and Shannon 1960),(Braude, Jackman, Linstead, and Lowe 1960a),(Braude, Jackman, Linstead, and Lowe 1960b). Furthermore, the hybridization of the β-C atom remains unchanged.
The inverse β-isotope effects explain with sufficient accuracy the total isotope effects of isotopomers 1h and 1i reported in the literature for dehydrogenation with DDQ (see Table 6.3.3), which are smaller than the values for isotopomers 1a and 1c (1d), respectively. This refutes a synchronous (pericyclic) dehydrogenation of 1,4-cyclohexadiene; the measured values are consistent with a two-step mechanism.
6.5.3 Isotope Effects of the Second Step of Dehydrogenation
The isotope effects of the second rapid step of hydrogen abstraction can be estimated from the product isotope effect of isotopomer 1b. This is influenced both by the individual isotope effects of the first step, p and α, and by the analogous — though not necessarily equal — effects of the second step, p’ and α’. The influence of the second step also accounts for the dependence of the product isotope effect on the stereochemistry (ST ) of the dehydrogenation. The product isotope effect of 1b can be related to the individual isotope effects and the stereochemistry of the reaction (ST ) using the following equations.
For any stereochemistry (ST = 0.50–1.00):
\[\frac{[\text D1]}{[\text D0]}=\frac{(1-ST)/[(1-ST)+ST·α'/p'] + ST/[ST=(1-ST)]α'/p'] + 1/α}{ST/[ST+(1-ST)p'/α'] + (1-ST)/[(1-ST)+ST·p'/α'] + 1/p}\]
For exclusive cis-abstraction (ST = 1.00):
\[\frac{[\text D1]}{[\text D0]}=\frac{1 + 1/α}{1 + 1/p}\]
For illustration, the following reaction scheme shows the individual pathways by which the two reaction products, [D0]-benzene and [D1]-benzene, can be formed from 1b.
In the case of exclusive cis abstraction of the transferred hydrogens (ST = 1.00), the product isotope effect is determined solely by the individual isotope effects of the first step (see equation above). However, the effects calculated for this limiting case using the approximated values for p and α deviate by approximately 15% from the experimental values, which is outside the error limits (see Table 6.5.3). These deviations can only be explained by assuming — even for the most stereoselective OCA/1c/dioxane system — a small but, for the product isotope effect, noticeable contribution from the trans elimination of two hydrogen atoms. The product isotope effects of 1b were therefore investigated in greater detail to determine, on the one hand, the stereochemistry of the reaction and, on the other hand, the isotope effects of the second step. The experimental values were approximated based on the results of the stereochemical tracer studies, using physically reasonable values for ST , p’, and α’. The following points should be noted:
The steric purity of the reactants 1c and 1d is greater than the values obtained by pyrolysis, since the radical disproportionation that occurred as a side reaction gives the appearance of lower steric purity (see Section 4.3.2).
The stereoselectivities determined in (see Section 4.3.2) are influenced by p’ and α’. Following the preferential abstraction of an H- atom over a D-atom by a factor of p/α in the first step, the cleavage of the cis-positioned H-atom—which is inherently desired for steric reasons—is favored in 1c and inhibited in 1d due to the influence of p’ and α’. Consequently, the stereoisomers exhibit different selectivities, which are more pronounced in the reaction with DDQ than in the reaction with OCA.
For α’, a value of 1.1 is specified, based on the first reaction step.
The results of the approximation are listed in the table below which shows the influence of stereoselectivity and the primary isotope effect of the second step on the product’s apparent isotope effect of 1b. For each redox system only the probable range for ST , as determined by the tracer studies of 1c and 1d, was taken into account (OCA: 1.00–0.90; DDQ: 0.90–0.87). The measured values were determined in dioxane at 25°C.
| measured | calculated for various combinations of ST / p’ | |||||||
|---|---|---|---|---|---|---|---|---|
| 1.00/ – | 0.99/14 | 0.98/7 | 0.95/3 | 0.90/2 | 0.90/4 | 0.87/2 | ||
| DDQ | 2.1 | 1.78 | 2.1 | 2.64 | 2.17 | |||
| OCA | 2.1 | 1.78 | 2.12 | 2.12 | 2.11 | 2.13 | 2.68 | |
A comparison of the calculated product isotope effects with the stereoselectivities of 1c and 1d allows the following conclusions: The transfer of hydrogen in the second step of the reaction is associated with a higher primary isotope effect (p’ = 3) for OCA than for DDQ (p’ = 2–3). Nevertheless, due to the lower value for ST , this has a stronger effect on the stereochemistry during the dehydrogenation of 1c and 1d with DDQ than during dehydrogenation with OCA (see Table 6.4).
6.6. Reaction Mechanism
The overall kinetic isotope effects for the isotopomers 1a – 1e rule out a synchronous dehydrogenation of 1,4-cyclohexadiene by the quinones DDQ and OCA. Instead, hydrogen abstraction proceeds in two steps. In the first rate-determining step, a hydride ion is transferred to the quinone. The high primary isotope effects indicate that this transfer occurs to a significant extent via quantum mechanical tunneling. The polar transition state and the resulting ionic intermediates are better stabilized by solvents with a high dielectric constant than by more nonpolar ones, as shown by the solvent dependence of the rate constants (Paukstat et al. 1985),(Brock 1986). However, these dependencies can also be interpreted as being influenced by the quinone oxidation potentials (Clark 1960). After hydride transfer, the ions remain fixed in a coplanar arrangement as a close ion pair, from which, in a second rapid step, the cis-positioned hydrogen is preferentially abstracted as a proton. The fixation of the ions is essentially based on electrostatic interactions between the charges. In contrast, a larger π-system that promotes ionic cohesion, such as in 1,4-dihydronaphthalene, leads to stereoselectivity independent of quinone and solvent (Brock 1986). In the C6 system, the ions become increasingly solvated as the polarity of the solvent increases, so that the bonding within the ion pair is weakened and a trans elimination becomes more likely. Primary isotope effects in the second step also influence the stereochemistry of the dehydrogenation: the selectivity is greater for the cis derivative 1c than for the trans derivative 1d. The bonding between the ions in the ion pair is significantly stronger in OCA than in DDO. In contrast to OCA, the latter exhibits a clear solvent- and isotopomer-dependence of stereoselectivity, which is also lower in absolute terms than that of ortho-quinone. Based on the findings described, the following ion pairs have thus been demonstrated as intermediates in the dehydrogenation of 1,4-cyclohexadiene by the quinones DDO and OCA:
7. Temperature Dependence of the Primary Isotope Effect in the Dehydrogenation of 1,4-Cyclohexadiene by Quinones
7.1. Fundamentals of the Temperature Dependence of the Primary Isotope Effect
In organic chemistry, primary isotope effects are traditionally used as mechanistic criteria. However, the magnitudes of these effects — which are usually measured at a single temperature — have repeatedly been overinterpreted with regard to the nature of the rate-determining step (Müller 1973). Even a well established reaction mechanism involving the irreversible transfer of only one H atom in the rate-determining step does not allow conclusions to be drawn about the structure of the transition state based on the magnitude of the isotope effect. The temperature dependence of the primary isotope effect, on the other hand, represents an experimental quantity that is linked to the molecular configuration of the activated complex via the ARRHENIUS equation:\[ \frac{k_{\text{H}}}{k_{\text{D}}} = \frac{A_{\text{H}}}{A_{\text{D}}} \exp\left( -\frac{\Delta E_{A_{\text{H/D}}}}{RT} \right) \]
According to the EYRING approach, the frequency factors A contain the state sums of the reactants and the activated complex as molecular quantities and are themselves slightly temperature-dependent. For the geometries of the three-center transition state described in Section 5.1.1 and under semiclassical conditions only, AH/AD values typically fall into the range of 0.7–6, the lowest values are expected for linear symmetry. However, quantum mechanical tunneling leads to a curvature of the ARRHENIUS line. The curvature, which is directed toward higher values, increases with rising temperature and, as a consequence of the mass-dependent tunneling probability, with decreasing reduced mass (Melander and Saunders 1980) (page 144). As a result, the experimentally accessible temperature range, as secants (or tangents) to the Arrhenius curve, yields smaller ratios of the frequency factors (AH/AD) and larger differences in activation energies than in pure semiclassical behavior. KWART discussed four different categories of the transition state (Kwart 1982). Table 7.1) lists the ranges for the activation parameters specified by KWART for these categories.
| Geometry | kH/kD (25°C) | AH/AD | ΔEAH/D [kJ/mol] |
|---|---|---|---|
| linear symmetric | 6 — 8 | 0.7 — 1.4 | 4.8 |
| linear asymmetric | 2 — 5 | 1.3 — 4.2 | 4.8 |
| linear tunneling | > 9 | < 0.6 | 6.3 — 25.1 |
| nonlinear (angeled) | 1.4 | 1.4 — 6 | 0 |
7.2. Literature Findings
Over the past ten years, an increasing number of studies have been published that include the temperature dependence of the primary isotope effect as a mechanistic criterion. GREEN et al., for example, were able to demonstrate a linear transition state for intramolecular γ-H transfer in 2-hexyloxy radicals (BARTON reaction) in conjunction with stereochemical studies (kH/kD(20°C) = 6.0, AH/AD = 1.0, ΔEAH/D = 4.3 kJ/mol) (Green et al. 1986). For the reduction of organic radicals by tri-n-butylstannane, STRONG et al. (Strong et al. 1983) found increasing asymmetry of the transition state in the radical series benzyl < sec. < prim. < phenyl. The activation parameters decreased in magnitude in this order within the specified limits: AH/AD = 1.4–0.8, ΔEAH/D = 3.4–1.2 kJ/mol. VERHOEVEN et al. conducted a series of studies on the hydride transfer from NAD(H) and similar dihydropyridines to various acridinium ions (Verhoeven et al. 1986):
| Activation Parameter | mechanisms | |
|---|---|---|
| intra | inter | |
| kH/kD (25°C) | 2.7 | 4.17 |
| AH/AD | 2.7 | 0.74 |
| ΔEAH/D [kJ/mol] | 0.0 | 4.3 |
The small, temperature-independent isotope effect of the intramolecular transfer corresponds to an angled transition state, which — in contrast to the intermolecular process (where the activation parameters suggest a linear complex) — is forced by the bridging alkyl chain. HHB Note 2026: the strain in the supposed “sandwiched” cation (see Figure 13) raises significant doubts about the assumption of an intramolecular H-transfer. In an analogous intermolecular hydride transfer in a system that differed only slightly, POWELL and BRUICE (Powell and Bruice 1983) obtained activation parameters that differed significantly from those listed above:
| Reaction System | Parameter | Value |
|---|---|---|
|
kH/kD (25°C) | 5.8 |
| AH/AD | 0.2 | |
| ΔEAH/D [kJ/mol] | 7.7 | |
The authors proposed an angled transition state with tunneling contributions. As an argument for the angled structure, they cited the small semiclassical isotope effect of 2.7 (25°C) — obtained by correcting the experimental value using BELL’s equation (Bell 1974) — and X-ray structural data from analogous enzyme-catalyzed hydride transfer reactions.
FILIPPO Jr. and VITALE (Vitale and Filippo 1982) demonstrated that a temperature-independent isotope effect is not a reliable indicator of an angled transition state, using the hydrogenation of n-octyl lithium as an example, which proceeds via a polar square intermediate but nevertheless with normal activation parameters (AH/AD = 0.24; ΔEAH/D = 5.3 kJ/mol).
ANHEDE and BERGMAN (Anhede and Bergman 1984) calculated nearly identical activation parameters for an intramolecular bent and an intermolecular linear proton transfer in mono- and non-protonated methylenediamine, respectively. Based on their calculations and theoretical work by STERN and WESTON (Stern and Weston 1974), the authors generally questioned the temperature dependence of isotope effects as a distinguishing criterion between linear and bent transition states. The temperature dependence of the isotope effect in the 1.5-sigmatropic H-migration in cis-1,3-pentadiene is also a subject of controversy (kH/kD = 12.2 (25°C); AH/AD = 1.15; ΔEAH/D = 5.9 kJ/mol) (Roth and König 1966). KWART (Kwart et al. 1982) saw this as evidence for a linear transfer — in contrast to the analogous intramolecular reaction in a sterically fixed quinolizine framework (Kwart et al. 1982). He regarded the temperature-independent isotope effect (AH/AD = 5.1; ΔEAH/D = 0) as evidence for an angular transition state. Ab initio calculations by HESS and SCHAAD (Hess and Schaad 1983), however, identified an angled configuration as the most stable arrangement; for this configuration, McLENNAN and GILL (McLennan and Gill 1985) calculated a temperature dependence of the isotope effect — albeit a small one — even at an extreme angle of 90° (ΔEAH/D = 1.4 kJ/mol). One possible explanation for the high isotope effects in an angled structure of this system is the VAT (“vibrationally assisted tunneling”) proposed by DEWAR (Dewar et al. 1985) and calculated by BUCK and DORMANS (Dormans and Buck 1986), in which hydrogen tunnels out of highly excited vibrational states shared by the reactant and product.
This work aims to investigate the temperature dependence of the primary isotope effect in the dehydrogenation of 1,4-cyclohexadiene by DDQ and OCA. The mechanism of these reactions has been established (see Section 6.6) and, in conjunction with the activation parameters, is intended to contribute to the discussion regarding the evaluation of temperature dependence as a mechanistic criterion.
7.3. Determination of the Temperature Dependence
The measurements of temperature dependence were carried out in dioxane over a temperature range between 15 and 60°C. For experimental reasons (the solvent’s melting point and the volatility of CHD), this range cannot be extended further without compromising the reliability of the measured values. The primary isotope effect was determined by intermolecular competition between non-indexed 1,4-cyclohexadiene and 1,4-cyclohexadiene 1a, tetradeuterated at the 3,6-position (see Section 6.3), by dividing the measured total isotope effect (p·α) by the secondary α-effect. In this process, the value for the secondary α-effect determined at 25°C (α = 1.08 for DDO and 1.1 for OCA) was assumed to be temperature-independent, which represents a reasonable approximation given the error margins of the method and findings in the literature [65d, 95]. The ARRHENIUS lines obtained for the 1,4-CHD/DDQ and 1,4-CHD/OCA systems are shown in Figure 14.
The primary isotope effects obtained from kinetic measurements at 25°C (see Section 6.5) are plotted together with the values determined by intermolecular competition. The good agreement with the other data points largely rules out systematic errors. A critical variable in the competition experiments is the ratio of the reactants, which is identical for all measured values due to the use of a standard solution. The error bars plotted represent the statistical errors resulting from GC and mass spectrometric analysis. To calculate the fitted regression lines, all values were used with the exception of the isotope effect of the OCA/1,4-CHD system determined at 30°C. The correlations were reasonable, with coefficients of 0.997 (OCA) and 0.994 (DDQ). Table 7.3 lists the activation parameters calculated from these results. The error limits are estimated from the straight lines, which, at maximum and minimum slopes, respectively, just barely touch the limits of the error bars for the values at 15°C and 60°C.
Corrections (HHB, 2026)
The data presented in Figure 14 have to be corrected for the following points:
- The α-IE used to calculate p for DDQ has to be changed to 1.08 (cf .Section 6.5.1).
- The sizes of the error bars for the intermolecular competition and for the kinetic measurements have to be inversed (cf. Table 6.3.3).
- Linear regression should be done without omitting any data pair using a weighted method (Levenberg-Marquardt algorithm). Error bars for the temperature axis fall into the size of the data ‘points’.
The updated data are shown in Figure 15 and the resulting activation parameter are added to Table 7.3:
| System / Model | kH/kD (25°C) | AH/AD | ΔEAH/D [kJ/mol] |
|---|---|---|---|
| DDQ | 11.2 ± 0.5 | 1.1 ± 0.1 | 5.8 ± 0.3 |
| OCA | 13.6 ± 0.5 | 0.6 ± 0.1 | 7.7 ± 0.3 |
| DDQ (corrected) | 1.3 ± 0.2 | 5.5 ± 0.5 | |
| OCA (corrected) | 0.7 ± 0.2 | 7.4 ± 0.6 | |
| linear tunneling | > 9 | < 0.6 | 6.3 – 25.1 |
| angled | > 1.4 | 1.4 – 6 | 0 |
7.4. Evaluation and Discussion of the Method
The very high primary isotope effects (> 6.9 at 25°C) clearly demonstrate that a significant portion of the H-transfer occurs via quantum mechanical tunneling, which is more pronounced in OCA than in DDQ. The considerations presented below therefore need to take into account the linear case with tunneling contributions from among the transition state categories listed by KWART in Table 7.1) (Section 7.1)). The ranges for the activation parameters specified by KWART for this purpose are compared in Table 7.3 with the values determined in this work for the dehydrogenation of 1,4-cyclohexadiene by DDQ and OCA. It can be seen that the ratios of the frequency factors (1.1 and 0.6, respectively) and the differences in activation energies (5.8 and 7.7 kJ/mol, respectively) are poorly consistent — or, in the case of DDQ, not at all consistent — with linear tunneling: The values lie at the boundaries and, in case of OCA, outside the valid ranges for this scenario. A simple interpretation of the temperature dependence of the isotope effect is therefore only possible in conjunction with the other findings regarding the reaction mechanism.
Stereochemistry of the reaction (see Section 6.4)
The high cis-selectivity of hydrogen elimination can be readily explained if the primary hydride abstraction occurs from a coplanar arrangement of the reactants (Model A), as is the case, for example, in a charge-transfer complex. In this arrangement of the reactants, the H abstraction must proceed via an angled transition state. Alternatively, a linear H-transfer between two exo-positioned reactant molecules would be possible (Model B), which, during or after hydride elimination, stack on top of each other in a coplanar arrangement due to electrostatic interactions as the charge builds up.
The angled structure in Model A explains the relatively small differences in activation energies and the fairly high ratios of frequency factors observed for a high tunneling contribution in a nonlinear geometry of the transition state according to the ranges reported by KWART (Table 7.3). In contrast, Model B should yield larger differences in activation energy (ΔEAH/D) and smaller ratios of AH/AD.
Different Behavior of DDQ and OCA
The higher isotope effects of OCA compared to DDQ can be explained by a higher tunneling contribution. Since the tunneling probability increases as the width of the energy barrier decreases, this interpretation implies that the spatial proximity of the reactants is better with OCA in comparison with DDQ. At the same time, the fixation within the forming ion pair is higher due to the narrower barrier. This leads to a higher cis-selectivity with OCA which is also not, or only to a much lesser extent, solvent-dependent, as it is observed with DDQ.
Comparison with Literature Data
For the interpretation of the activation parameter with regard to an angular or linear structure of the transition state, respectively, the temperature dependence in the dehydrogenation of triphenylmethane derivatives by para-chloroanil (PCA = 2,3,5,6-tetrachloro-1,4-benzoquinone) in acetonitrile is of interest, reported by Lewis et al. (Lewis et al. 1970) (Table 7.4)
| Reaction System | Parameter | Value |
|---|---|---|
|
kH/kD (25°C) | 11.8 |
| AH/AD | 0.041 | |
| ΔEAH/D [kJ/mol] | 14.2 | |
The activation parameter of this comparable dehydrogenation, assuming the same magnitude of the isotope effect, correspond ideally to a linear H-transfer with high tunneling contribution. A reaction analogous to the 1,4-CHD/DDQ(OCA) system, originating from a coplanar arrangement with an angled C—H—O group configuration, appears to be hindered here for steric reasons as shown in Figure 17.
The results of this study show that hydride elimination from 1,4-cyclohexadiene via DDQ and OCA proceeds via an angled transition state from a coplanar arrangement with a high tunneling contribution. However, this conclusion is only valid in conjunction with the other findings, particularly those regarding stereochemistry. The temperature dependence of the isotope effect alone is not a reliable criterion for bent structures with tunnel contributions. The categories for the transition state established by KWART must therefore be expanded to include the angled geometry with tunnel contributions. However, no characteristic ranges for the activation parameter can be assigned which are distinctly different from the other transition states. In such ambiguous cases, the following options are available for a more detailed investigation:
The determination of analogous tritium isotope effects, which, due to the mass dependence of tunneling, allow conclusions to be drawn about its contribution to hydrogen transfers (Melander and Saunders 1980) (page 143).
The separate determination of the activation parameters for the labeled and unlabeled substances via kinetic measurements allows, in such ambiguous cases, the separation of the tunneling and semiclassical components of the isotope effect using the BELL equation, and thus the determination of the activation parameters for semiclassical behavior. However, these are only approximations, since a one-dimensional energy barrier is considered (Melander and Saunders 1980) (page 147).
8. Homogeneously Catalyzed Disproportionation of 1,4-Cyclohexadiene
8.1. Disproportionation on Colloidal Nickel
8.1.1. Literature Findings
While transition-metal-catalyzed hydrogenation with molecular hydrogen was the subject of numerous studies, particularly in the 1960s and 1970s (Hussey et al. 1968),[Siegel and Smith (1960)[,(House 1972), the corresponding catalytic dehydrogenation of hydroaromatics has been comparatively little studied (Linstead et al. 1942),(Linstead et al. 1937). It is, however, a reaction of considerable interest in synthetic chemistry (Fu and Harvey 1978), for which — due to its microscopic reversibility — a cis-selective stepwise mechanism (Horiuti and Polanyi 1934) analogous to that of hydrogenation was assumed (X=catalyst):
| H2 adsorption: | 2 H–X |
| Alkene adsorption: | H2CX–CXH2 |
| 1st H transfer: | H2CX–CH3 |
| 2nd H transfer: | H3C–CH3 |
In contrast, the mechanisms of platinum-catalyzed disproportionation of hydroaromatics have been investigated for only a few systems (Corson and Ipatieff 1939). For example, the palladium-catalyzed disproportionation of cyclohexene — for which BRAUDE et al. (Braude, Linstead, et al. 1954) proposed a concerted, termolecular mechanism — appears to proceed via a bimolecular H-transfer between two adsorbed cyclohexene molecules:
This was supported by second-order kinetics for cyclohexene (Carrà et al. 1964) and maximum specific activities of the catalyst for 4, 10, and 20 palladium atoms per intermediate complex, respectively (Carrà and Ragaini 1967). In contrast, the palladium-catalyzed disproportionation of 1,2-dihydronaphthalene proceeds via a two-step, stereoselective cis-dehydrogenation by the catalyst followed by rapid hydrogenation with loss of the molecular identity of the transferred hydrogen (Heesing and Müllers 1980b). The disproportionation or hydrogenation of the two isomeric cyclohexadienes using metallic Pd and Pt-Mohr (Zelinsky and Pawlow 1933), Pd (Gryaznov and Yagodovskii 1963), Rh (Freidlin and Popova 1968), and Raney-Ni catalysts (Freidlin et al. 1972) is known from the literature but has not been studied in detail. In a 1985 paper, SAKAI et al. (Sakai et al. 1985) reported on the homogeneously catalyzed disproportionation of 1,3- and 1,4-cyclohexadiene on colloidal nickel. In contrast to other (heterogeneous) metal catalysts, the reactants in this system exhibited only a negligible tendency toward isomerization despite high reactivity. In their publication, the authors limited themselves to the disproportionation of the more reactive 1,3-cyclohexadiene, for which they optimized the reaction conditions in a series of experiments and investigated the influence of catalyst poisons such as triphenylphosphane and water. They proposed the following mechanism of seven consecutive steps, which they did not, however, substantiate with experimental data:
| Step | Reaction, [C6Hn] = covalently bound molecule |
|---|---|
| adsorption equilibrium | \(\ce{Ni(0) + 2 C6H8 <=> C6H8-Ni(0)-C6H8}\) |
| oxidative addition | \(\ce{C6H8-Ni(0)-C6H8 -> C6H8-NiH[C6H7]}\) |
| insertion | \(\ce{C6H8-NiH[C6H7] -> [C6H7]Ni[C6H9]}\) |
| β-H abstraction | \(\ce{[C6H7]Ni[C6H9] -> C6H6-NiH[C6H9]}\) |
| ligand exchange | \(\ce{C6H6-NiH[C6H9] + C6H8 -> C6H8-NiH[C6H9] + C6H6}\) |
| reductive elimination | \(\ce{C6H8-NiH[C6H9] -> C6H8-Ni(0)-C6H10}\) |
| ligand exchange | \(\ce{C6H8-Ni(0)-C6H10 + C6H8 -> C6H8-Ni(0)-C6H8 + C6H10}\) |
This chapter examines the disproportionation of 1,4-cyclohexadiene catalyzed by colloidal nickel. Due to its low tendency toward isomerization combined with sufficiently high activity, this catalyst is particularly well-suited for mechanistic studies and is of potential interest for preparative purposes, especially given the easy availability of the reactants.
8.1.2. Reaction Mechanism
The catalyst solutions were prepared in situ from nickel(II)bromide by reduction with zinc in dimethylformamide, following the method described by SAKAI et al. (Sakai et al. 1985). The disproportionation was carried out at 60°C in an inert gas atmosphere. The molar ratio of reactant to catalyst was approximately 20 at a nickel concentration of 1.9·10—2 M. The only disproportionation products were cyclohexene and benzene. Small amounts of cyclohexane (< 0.5%) were formed by disproportionation and/or hydrogenation of the cyclohexene. The two products were not formed in equimolar proportions. The ratio of benzene to cyclohexene increased over the course of the reaction from 1.0 to approximately 1.1 at 90 percent conversion. The absolute differences in the percentage fractions increased from an initial value of approximately 0.2% to 4.0% toward the end of the reaction. These findings suggest that dehydrogenation and hydrogenation are two separate steps, with the catalyst’s hydrogenation activity decreasing as the conversion increases. 1,3-Cyclohexadiene could not be detected at any point during the reaction. Higher-molecular-weight compounds and polymers were not detected even after more than three days of reaction time. After longer reaction times (approx. 8–12 hours), the colloidal catalyst was precipitated by polymeric substances extracted from the septum which also occurred in the absence of 1,4-cyclohexadiene. At the same time, the total amount of C6 hydrocarbons decreased relative to the internal standard (o-xylene). This is likely due to differences in the diffusion of C6 and C8 hydrocarbons through the septum, since no side reactions occurred within the detection limits (see above). The reaction only takes place to a negligible extent on the zinc/nickel support, as demonstrated by a control experiment using precipitated colloidal nickel.
Effect of nickel bromide concentration
To investigate the influence of the catalyst on the disproportionation, batches were prepared with increasing amounts of catalyst components, and the conversion rates of the identically treated reaction solutions, determined at the same time intervals, were taken as a measure of the reaction rate. The results are shown in Figure 18:
A region can be identified in which the reaction proceeds in the first order with respect to the catalyst, assuming that the catalyst concentration is proportional to the amount of catalyst and reactants. At nickel bromide concentrations > 1.3·10—2 M, the activity increases only insignificantly. There are two possible explanations for this:
The concentration of colloidal nickel has reached its maximum possible value. Any further reduced nickel bromide is then present in a heterogeneous form as a precipitate with lower activity.
Above this limiting concentration of colloidal nickel, the particles associate to form colloids that are less catalytically active but still soluble.
HOLAH et al. (Holah et al. 1979) observed a curve similar to that in Figure 18 during alkene hydrogenation over partially hydrogenated nickel boride.
Reaction order with respect to cyclohexadiene
The reaction orders with respect to time were determined for 1,4-cyclohexadiene at 60°C and, for comparison, for 1,3-cyclohexadiene at 25°C under otherwise identical conditions. Figures 2 and 3 show the curves obtained after evaluating the kinetic data according to first- and second-order rate laws. Despite the lower temperature, the conjugated cyclohexadiene still reacts 30 times faster.
The disproportionation reaction is therefore first-order with respect to 1,3-cyclohexadiene, but second-order with respect to 1,4-cyclohexadiene and the experimental rate constants are:
| Isomer | T [°C] | k exp |
|---|---|---|
| 1,3-cyclohexadiene | 25 | (2.8 ± 0.6) × 10—2 [M—1·s—1] |
| 1,4-cyclohexadiene | 60 | (2.4 ± 0.5) × 10—1 [M—2·s—1] |
The kinetic data for both cyclohexadienes rule out isomerization of the 1,4-diene to the more reactive conjugated diene, whose steady-state concentration could lie below the detection limit of gas chromatographic product analysis. The following general reaction scheme explains the kinetic data for the reaction of both cyclohexadienes (generally abbreviated as CHD).
\[ \ce{CHD + Ni(0) <=>[k_{+}/k_{-}] CHD-Ni(0) ->[k_1] C6H7-NiH ->[k_1'] C6H6-NiH2} \]
\[ \ce{C6H6-NiH2 + CHD <=> C6H6-NiH2-C6H8 ->[k_2] C6H6-NiH-C6H9} \]
\[ \ce{C6H6-NiH-C6H9 ->[k_3] C6H6-Ni(0)-C6H10} \]
The cyclohexadiene is reversibly adsorbed onto the catalyst surface (CHD–Ni) and is then gradually dehydrogenated by the nickel via oxidative addition and γ-H abstraction, forming an intermediate dihydrido-nickel species (NiH2). The latter then reacts with a second molecule of cyclohexadiene to form the disproportionation products. The following conclusions can be drawn from the reaction scheme to explain the kinetic data:
As shown in Figure 18 for the disproportionation of 1,4-cyclohexadiene, the reaction is first-order with respect to the catalyst, since the adsorption equilibrium is shifted to the left (k+ << k—).
The disproportionation can proceed in either first or second order with respect to the two cyclohexadienes.
The reaction is first-order with respect to 1,3-cyclohexadiene (Figure 19) if the formation of the adsorbate (k+ << k1, k1‘, k2, k3) or the dehydrogenation (k1, k1’ << k+, k2, k3) is rate-determining.
A second-order reaction with respect to 1,4-cyclohexadiene (Figure 20) is possible in this reaction scheme only if the hydrogenation reaction is rate-determining (k2(k3) << k+ << k1, k1’) and the three fast pre-equilibria of adsorption and dehydrogenation occur on the reactant side (Schmid and Sapunov 1982) (page 39).
A second-order reaction involving cyclohexadiene also occurs when the transfer of one or both hydrogen atoms takes place directly between two cyclohexadiene molecules that have been adsorbed in two successive, fast pre-equilibria with small formation constants:
\[ \ce{CHD + Ni <=> CHD-Ni + CHD <=> CHD-Ni-CHD} \]
\[ \ce{CHD-Ni-CHD -> C6H10 + C6H6} \quad \text{(2 H simultaneously)} \]
\[ \ce{CHD-Ni-CHD -> C6H9-Ni-C6H7 -> C6H10 + C6H6} \quad \text{(2 H stepwise)} \]
The disproportionation of 1,3-cyclohexadiene was not investigated in further detail in the course of this work. In contrast, the mechanism of the reaction of 1,4-cyclohexadiene was subjected to a more detailed analysis using deuterated compounds.
8.1.3. Tracer Study on H/D Scrambling and Deuterium Loss
For the stereochemical investigations, it was of interest whether and to what extent isotope exchange and/or deuterium loss occurs during disproportionation. To investigate this, the tetradeuterated derivative 1a and the isotopomer 1e were reacted with colloidal nickel, and the molecular peak groups of both disproportionation products were analyzed by mass spectrometry. If intramolecular scrambling occurs in the reactants, 1a should yield higher-deuterated benzenes in addition to [1,4-D2]-benzene, and 1e should yield lower-deuterated benzenes in addition to [1,2-D2]-benzene. The scenario for the latter is exemplary shown below where a mixture of 1c/1d is formed via the isotopomer [2,3-D2]-1,4-CHD:
In both cases, intermolecular exchange processes between C6 hydrocarbons lead to higher- and lower-deuterated benzenes, whereas exchange with solvent molecules yields only lower-deuterated benzenes. The analytical results from experiments starting with 1a and 1e, respectively, are shown in Table 8.1.3. In the light of ther deuterium contents (1a: 99.6% D4, 1e: 98.6% D2, see Section 2.4) slight intramolecular scrambling has occurred with 1a, which is independent of the catalyst’s activity and thus of the reaction time. The data for isotopomer 1e show only minor deviations from the expected value which fall within the margin of error. For the exclusion of systematic errors in determining the isotopomer pattern in deuterated benzene see Section 3 and Section 4.1.1.
| 1,4-CHD | time | conversion | Benzene | ||
|---|---|---|---|---|---|
| Isotopomer | [h] | [%] | [D0] | [D1] | [D2] |
| 1a [3,3,6,6-D4] | 5 | 29 | 95.2 | 3.8 | 0.9 |
| 1a [3,3,6,6-D4] | 24 | 32 | 95.2 | 3.8 | 0.9 |
| 1e [1,2-D2] | 8 | 79 | 0.5 | 0.7 | 98.8 |
In contrast, the molecular peak groups of the reduction products (cyclohexene, cyclohexane) show a significant deviation from the expected isotopomer distribution. Due to the [M-n]+ peaks present in cyclohexene and the lack of reference samples of indexed cyclohexenes available in this study, only semi-quantitative conclusions can be drawn by comparing the molecular peak groups from different disproportionations as well as with the one of unindexed cyclohexene.
Since no findings are available regarding the mechanisms of hydrogen elimination occurring in mass spectrometers, such comparative analyses are only valid if approximately identical compositions are expected for the isotopomer mixtures under consideration, in other words if the D-contents of the isotopomers are only slightly different.
The molecular peak groups of the unlabeled cyclohexene and the reaction product from 1e are practically indistinguishable. This indicates that no H/D equilibration has occurred during the disproportionation of this isotopomer.
In contrast, in the reduction product of 1a, the [M-1]+ peak has similare intensity as the M+ peak (m/e=86). The latter is two mass units smaller than the expected value (m/e=88) for a hydrogenation step that maintains the molecular identity of hydrogens from the dehydrogenation step. The rather high intensities of other [M-n]+ peaks is expected given the high deuterium content of this isotopomer and the associated preferential formation of the [M-2]+ peaks.
Accordingly, following the dehydrogenation of cyclohexadiene, a noticeable intermolecular H/D exchange takes place between the chemisorbed deuterium from 1a and the hydrogen atoms of the solvent molecules. As the two tracer studies with 1a show, the extent of the exchange depends only insignificantly on the activity of the disproportionation catalyst and thus on the reaction time.
A second possibility, which involves deuterium loss through H/D exchange with solvent molecules, consists of a regioselective, reversible dehydrogenation of cyclohexene at the allylic position, as postulated by TSAI et al. (Tsai et al. 1982). However, this process should become more pronounced with increasing reaction time, which is only the case to a limited extent in the present study. Furthermore, in desorption experiments with mixtures of non- and perdeuterated cyclohexene and of non- and perdeuterated benzene on various nickel surfaces in ultra-high vacuum, TSAI et al. were unable to detect any isotope exchange. Therefore, it is likely that the deuterium losses occur exclusively during the dehydrogenation of 1,4-cyclohexadiene at the intermediate nickel hydride stage.
8.1.4. Tracer Studies on the Stereochemistry of the Reaction
The stereoselectively dideuterated isotopomers 1c and 1d were disproportionated under standard conditions and the reaction products were analyzed by mass spectrometry.
The dehydrogenation of 1,4-cyclohexadiene proceeds as a stereoselective cis elimination even at conversion rates of up to 80 percent, as shown by the isotope patterns of the resulting benzenes:
| 1,4-CHD | time | conversion | Benzene | ||
|---|---|---|---|---|---|
| Isotopomer | [h] | [%] | [D0] | [D1] | [D2] |
| 1c [cis-3,6-D2] | 1.5 | 43 | 28.8 | 9.1 | 62.1 |
| 1c [cis-3,6-D2] | 8.0 | 77 | 28.6 | 9.2 | 62.3 |
| 1d [trans-3,6-D2] | 3.5 | 64 | 3.3 | 84.1 | 12.6 |
| 1d [trans-3,6-D2] | 8.5 | 76 | 3.4 | 84.0 | 12.6 |
Taking into account the steric purity of the starting materials (which is higher in 1c), the cis selectivity is larger than 96 percent.
The hydrogenation of 1,4-cyclohexadienes should, while preserving the molecular identity of the transferred hydrogen, yield a mixture of approximately 29% D4, 9% D3, and 62% D2-cyclohexene in the case of 1c, and approximately 84% D3-cyclohexene in the case of 1d. This is illustrated in the figure Figure 22.
Figure 22 shows the molecular peak groups of the resulting cyclohexenes (the molecular peak group of the unlabeled cyclohexene is shown in Figure 21 for comparison).
The broad agreement between the molecular peak groups for both product mixtures demonstrates their very similar isotopomeric composition. This clearly indicates that the abstracted hydrogen atoms have reached complete equilibrium on the catalyst surface. The hydrogen is then transferred to a second molecule in accordance with the H/D ratios formed during dehydrogenation. The statistical transfer of HH, HD, and DD is influenced by isotope effects during hydrogenation and by H/D exchange with solvent molecules.
8.1.5. Inferences on the Mechanism from Total Isotope Effects
The dehydrogenation step of the disproportionation reaction will be investigated in more detail by determining and evaluating total isotope effects.
8.1.5.1. Product isotope effects for 1b and 1c
The isotopomer 1b, monodeuterated at the 3-position, was disproportionated under intramolecular competition, and the reaction products were analyzed by mass spectrometry. The analysis yielded the following product compositions: 40.2% [D0]-benzene / 59.8% [D1]-benzene. The ratio of deuterated to unlabeled benzene is the product isotope effect:
[D1]-benzene/[D0]-benzene = 1.49 ± 0.05.
The product isotope effect of [cis-3,6-D2]-1,4-cyclohexadiene 1c was determined in the tracer study on the stereochemistry of dehydrogenation (see Section 8.1.4) and is given by the ratio:
[1,4-D2]-benzene/[D0]-benzene = 2.20 ± 0.05.
The two product isotope effects can be related to each other via the individual isotope effects (see Section 6.5). For a synchronous or a two-step, cis-selective dehydrogenation, the following relationships apply, in which the product isotope effects depend only on the primary (p) and secondary-α isotope effects (α) of the first dehydrogenation step:
| Isotopomer | Benzene Ratio | Synchronous | Two-Step |
|---|---|---|---|
| 1b [3-D1] | [D1]/[D0] | \(p/α\) | \((1+α^{-1}/(1+p^{-1})\) |
| 1c [cis-3,6-D2] | [1,4-D2]/[D0] | \((p/α)^{2}\) | \(p/α\) |
Accordingly, in a synchronous dehydrogenation, the product isotope effect of 1b must correspond to the square root of the product isotope effect of 1c. This is indeed the case (1.49 = [2.20]1/2). If a secondary α effect of 1.1 is specified, the relationships for a synchronous process yield a primary effect of 1.64 or 1.63, respectively.
To verify this, however, it was necessary to check whether a similarly good agreement could also be achieved with a two-step mechanism. To this end, the product isotope effects of 1b and 1c can be approximated by using the relationships listed above and assuming 100 percent cis abstraction (ST = 1.00). Congruency between the two experimental values is provided with values for the individual isotope effects: p = 1.43 and α = 0.65.
However, an inverse α effect is theoretically not plausible for the reaction under consideration, in which the C atom undergoes hybridization from sp3 to sp2 (see Section 5.1.2)). Furthermore, the value lies outside the range described in the literature for inverse α effects (Shiner et al. 1968),(Llewellyn et al. 1960),(Halevi 1963).
The approximation was therefore performed for a two-step reaction in which an incomplete stereoselectivity (ST < 1.00) influences the measured values. For this purpose, a cis-selectivity between 95 and 100 percent (ST = 0.95–1.00) was assumed (see Section 8.1.4), and the product isotope effect of 1b was calculated using constant values for the secondary α effects (α = α’ = 1.1).
For primary isotope effects of the first stage (p) between p = 2–4, the calculations were performed by varying the primary effect of the second step (p’ = 2–4) for different ST values. The formula used for the approximation is described in Section 6.5.3.
Table 8.1.5.1b shows exemplary the results for a primary isotope effect of the first dehydrogenation step of p = 2.4, which is the one derived from the relationship for a cis-selective two-step mechanism (ST = 1.00) from the product isotope effect of 1c (vide supra).
| p’ | cis-stereoselectivity ST | |||||
|---|---|---|---|---|---|---|
| 0.95 | 0.96 | 0.97 | 0.98 | 0.99 | 1,00 | |
| 2.0 | 1,45 | 1.43 | 1.41 | 1.39 | 1.37 | – |
| 3.0 | – | 1.50 | 1.46 | – | – | – |
| 4.0 | 1.62 | 1.57 | 1.51 | 1.46 | 1.40 | – |
| p = 2.4* | – | – | – | – | – | 1.35 |
| *: for comparison derived from 1c dehydrogenation: [1,4-D2]/[D0] = p/α = 2.20 | ||||||
The approximation shows that the product isotope effects are also consistent with a two-step mechanism if the hydrogen abstraction proceeds with 96 to 98 percent cis-selectivity and a primary isotope effect in the second step in the range p’ = 2–4. Even when p is varied within the range specified above, good agreement is obtained with the measured value for 1b.
Based on the product isotope effects of 1b and 1c, the two-step mechanism involving 100% cis elimination of the hydrogens can be ruled out.
However, the results cannot be used to distinguish between a synchronous and a two-step dehydrogenation that is not entirely stereoselective.
8.1.5.2. Total kinetic isotope effects of 1a and 1b
Further insights should be provided by the total kinetic isotope effects of 1a and 1b. Therefore, these isotopomers were each reacted in intermolecular competition with the unlabeled 1,4-cyclohexadiene (for methodology, see see Section 5.2.1). Using the BIGELEISEN equation the total kinetic isotope effects were derived from the measured values and listed in Table 8.1.5.2 as total kinetic isotope effects.
The experimental values are compared with calculated isotope effects for the two conceivable mechanisms in the dehydrogenation step: synchronous versus two-step hydrogen transfer, the latter assuming 100% cis-selectivity. The calculations are relying on the primary and the secondary alpha isotope effect, p and α, respectively, which can be derived from the product isotope effects of 1b and 1c (assuming that α = 1.1). However, as outlined in Section 8.1.5.1 for the stereoselective, two-step mechanism, the inconsistency of the product isotope effects between 1b or 1c leads to different values for the primary isotope effect.
| Experimental Isotope Effects | Calculated Isotope Effects | ||||
|---|---|---|---|---|---|
| CHD | total kinetic IE | synchronous | 2-step (ST =1.00) | ||
| from 1b | from 1c | ||||
| 1/1a | kHH/kDD | 2.53 ± 0.01 | 3.2 | 3.9 | 2.6 |
| 1/1b | kHH/kHD | 0.71 ± 0.01 | 1.3 | 1.3 | 1.2 |
The two total kinetic isotope effects can be interpreted as follows, taking into account the second-order reaction on 1,4-cyclohexadiene determined in Section 8.1.2, according to which the hydrogenation reaction must be rate-determining.
Intermolecular Competition Between 1 and 1b
The isotopomer distribution in the dehydrogenation product is approximately determined by the reactant ratio (1/1b = 0.83) and the product isotope effect of 1b (1.49): 67.4% D0 / 32.6%D1 (experiment: 68.4% D0, 31.6% D1). Thus, within the margin of error, no overall kinetic isotope effect is observed for this step. Assuming an approximately equal complexation of the isotopomers by nickel, this suggests, that dehydrogenation does not occur in the rate-determining step. However, it should be noted that the total isotope effect expected for rate-determining dehydrogenation can only assume small values anyway due to the intramolecular competition between one deuterium atom and three hydrogen atoms in the allylic positions.
Intermolecular Competition Between 1 and 1a
In this case dehydrogenation is also not rate-determining. Since no intramolecular competition is possible in isotopomer 1a, the individual isotope effects during disproportionation have a stronger effect here than in the case of intermolecular competition with 1b, so that a “normal” total isotope effect is observed (IE > 1). However, no further conclusions can be drawn from the magnitude of this isotope effect for the following reasons:
- The isotope effects in the rate-determining hydrogenation reaction are unknown.
- Due to the high indexing in 1a, this isotopomer could be hydrogenated at a different rate compared to the non-indexed cyclohexadiene.
- The tetradeuterated derivative could be preferentially complexed and thus dehydrated compared to 1 due to the more electropositive and sterically less demanding C–D bonds.
The measured total isotope effect is therefore influenced not only by the equilibrium state during adsorption but also by both the hydrogenation reaction and the dehydration reaction, and is thus of no value for further mechanistic considerations.
8.1.6. Mechanism of Disproportionation
The findings presented in Sections 8.1.2 through 8.1.5 can be explained by the following reaction scheme, which illustrates the main reaction pathway occurring during the disproportionation of 1,4-cyclohexadiene using the isotopomer 1d as an example:
The 1,4-cyclohexadiene is adsorbed onto the catalyst surface in a rapid pre-equilibrium state that is far on the reactant side. Dehydrogenation proceeds stepwise via oxidative addition and γ-H abstraction in two equally rapid equilibrium reactions, forming a dihydrido-nickel derivative with the π-bonded dehydrogenation product benzene. In a further rapid pre-equilibrium step, another molecule of 1d is then adsorbed upon desorption of the benzene, followed by the hydrogenation of the adsorbed 1d to cyclohexene as the rate-determining reaction. A stepwise course analogous to heterogeneous, metal-catalyzed hydrogenations is likely for this step (Horiuti and Polanyi 1934). Finally, cyclohexene is desorbed from Ni(0) before the next cycle can start at that site of the nickel surface.
The three fast pre-equilibria with small equilibrium constants (preferred reverse reaction), together with the slow, rate-determining hydrogenation reaction, explain the disproportionation kinetics determined in Section 8.1.2, which proceeds as a second-order reaction with respect to 1,4-cyclohexadiene. The high cis-selectivity of > 96% in the two-step dehydrogenation is due to the steric fixation of the partially dehydrogenated cyclohexadiene on the catalyst surface as a result of the σ-allylic bond and π-interactions of the double bonds.
Due to the reversibility of the dehydrogenation, the abstraction of trans-positioned hydrogen atoms occurs as side reaction in approximately 5% cases. This is consistent with the product isotope effects of 1b and 1c, which require high but not complete cis-selectivity (see Section 8.1.5.1). However, the effects are also consistent with a single-step dehydrogenation (see below), so they do not constitute unequivocal proof of the described mechanism.
The loss of the molecular identity of the abstracted hydrogen — reflected in the largely identical compositions of the reduction products of the stereo-isotopomers 1c and 1d — occurs through complete H/D equilibration at the dihydrido intermediate stage.
The absence of a total kinetic isotope effect for 1b also fits into a kinetic model in which the hydrogenation reaction is rate-limiting (see Section 8.1.5.2).
In contrast, the total kinetic isotope effect of 1a cannot be explained by this reaction mechanism without further assumptions, since it exhibits a “normal” value (> 1.0), as would be expected for a dehydrogenation reaction occurring in the rate-determining step. However, if one takes into account the molecular properties — which differ significantly from those of the non-indexed 1,4-cyclohexadiene due to the high indexing and which were already evident during the gas chromatographic separation of these isotopomers (see Section 4.2) — this overall kinetic isotope effect is put into perspective:
On the one hand, contributions from equilibrium isotope effects must be taken into account; on the other hand, a conceivable difference in the reactivity of the isotopomers during hydrogenation must be considered.
Furthermore, the intermediate formation of the dihydrido-nickel derivative explains the slightly preferential formation of benzene over cyclohexene. A portion of the desorbed hydrogen in equilibrium with the nickel hydride may escape from the reaction vessel and is not available for hydrogenation. The loss of deuterium during the disproportionation of tetradeutero-cyclohexadiene 1a may also occur at the nickel hydride stage through an H/D exchange with the solvent.
The intramolecular H/D scrambling detected during the dehydrogenation of 1a — which occurred as a side reaction at approximately 5% — suggests that, in addition to a σ-allyl-bound cyclohexadiene derivative, a π-allyl-bound cyclohexadiene derivative may also exist as an intermediate. This species can be coordinated either via a π3- or a π5-bond.
In addition to the described reaction pathway, the findings could also be in parts explained by the two following alternative mechanisms.
Direct H-Transfer between Donor and Acceptor
The second-order reaction on 1,4-cyclohexadiene (see Section 8.1.2) would be in accordance with a direct hydrogen transfer between two adsorbed cyclohexadiene molecules. However, this is contradicted by the absence of a total kinetic isotope effect for 1b, as well as the first-order reaction with respect to the reactant observed for 1,3-cyclohexadiene. There is no further evidence supporting such an alternative mechanism. Literature findings on the hydrogenation of both cyclohexadienes catalyzed by Raney nickel and cobalt (Freidlin and Popova 1968),(Freidlin et al. 1972) as well as rhodium (Gryaznov and Yagodovskii 1963) support the existence of rapid preliminary equilibria, as depicted in the above reaction scheme: under hydrogenation conditions, 1,4-cyclohexadiene disproportionated only at reduced hydrogen pressure, whereas 1,3-cyclohexadiene was dehydrated even at elevated pressures. These findings suggest that the coordination of 1,4-cyclohexadiene by nickel is largely suppressed and/or that hydrogen shifts the equilibrium of the dehydrogenation reactions toward the reactants.
Synchronous Dehydrogenation
The high cis-selectivity may be the result of a synchronous dehydrogenation occurring in a reaction pathway otherwise identical to the scheme above.
The product isotope effects of 1b and 1c are also in accordance with this mechanism. Since this course of events cannot account for the intramolecular H/D exchange, this mechanism requires that a stepwise dehydrogenation (see above) occurs simultaneously as a side reaction.
8.2. Disproportionation on the Ziegler Catalyst (Ni(acac)2/AlEt3 System)
8.2.1. Literature Findings
Ziegler catalysts have been known since the 1960s as particularly efficient hydrogenation catalysts (Sloan et al. 1963) and are of technical interest as reactive homogeneous systems (Pelimenshchikov and Zhidomirov 1983). UV spectroscopic and magnetic measurements by SCHMIDT et al. (Schmidt et al. 1983) as well as 1H-NMR and IR spectroscopic studies by van OMMEN et al. (Ommen et al. 1977),(Ommen et al. 1979),(Ommen et al. 1981) show that the transition metal acetylacetonates reduced with trialkylaluminum exist as M(0) complexes with aluminum alkylenes as stabilizing ligands. The formation pathway and the intermediates involved depend not only on the metal but also, to a decisive extent, on the ratio of aluminum alkyl to transition metal compound and on the solvent (Schmidt et al. 1983). Contrary to earlier descriptions of these hydrogenation catalysts as colloidal metals (Ziegler et al. 1960), the metal hydrides formed with hydrogen are discussed as the active species (Kroll 1969), although they have been isolated only for a few systems (e.g., a cobalt hydrido complex (Petit et al. 1980),(Yamamoto et al. 1971)). In contrast to hydrogenation, the dehydrogenation of hydrocarbons catalyzed by ZIEGLER catalysts remains a reaction that is still largely unknown today. HANSON (Hanson 1980) reported in 1980 that ZIEGLER complexes of cobalt and nickel catalyze the disproportionation of 1,3- and 1,4-cyclohexadiene. In 1982, SAKAI et al. (Sakai et al. 1982) were able to detect benzene as a disproportionation product during the selective hydrogenation of 1,4-cyclohexadiene with a nickel complex. The disproportionation of 1,3-cyclohexadiene and cyclohexene, as well as the aromatization of tetralin, was reported by COSTA et al. in 1984 (Costa et al. 1984). They used cobalt and nickel acetylacetonates reduced with aluminum triisobutyl, whose activity could be completely inhibited by triphenylphosphane. To date, no mechanistic studies on dehydrogenation or disproportionation catalyzed by ZIEGLER catalysts have been reported in the literature. Concurrently with this work, BROCK’s (Brock 1986) own research group investigated the disproportionation of 1,2-dihydronaphthalene catalyzed by a defined cobalt hydrido complex.
8.2.2. Selection of the Catalyst
The present work aims to investigate the disproportionation of 1,4-cyclohexadiene catalyzed by ZIEGLER complexes. For mechanistic studies, it is necessary to find a catalytic system that allows for mild conditions and also exhibits a consistent reaction course. In homogeneous catalyzed reactions of 1,4-cyclohexadiene, the main interfering side reactions to be expected are isomerization to 1,3-cyclohexadiene (Lyons 1969),(Green and Kuc 1972),(Pertici et al. 1980) and olefin polymerization of the reactant and/or cyclohexene (Schuchardt and Santos Diaz 1985),(Kobayashi et al. 1983) 2U. Based on literature findings regarding the hydrogenation (Sloan et al. 1963) and disproportionation activity (Hanson 1980) of various ZIEGLER systems, screening experiments were conducted to test the catalytic activity for the disproportionation and isomerization of 1,4-cyclohexadiene using various transition metals and reaction conditions (Table 8.2.2). The activities are satisfactorily high only for the nickel and cobalt complexes and, to a limited extent, for the chromium catalyst. Furthermore, as the rate of disproportionation increases, the tendency toward isomerization decreases. For these reasons, the nickel system is by far the most suitable for mechanistic studies on 1,4-cyclohexadiene.
| Catalyst System | Reaction Conditions | Products [%] | |||
|---|---|---|---|---|---|
| M(acac)n | AlR3/M | T [°C] | t [h] | C6H6 & C6H10 | 1,3-CHD |
| Mn(acac)3 | 6 | 50 | 5 | 3.1 | 1.7 |
| 24 | 4.3 | 1.8 | |||
| Mo(acac)2 | 3 | 50 | 24 | 4.9 | 4.5 |
| 6 | 5 | 5.5 | 5.9 | ||
| Cr(acac)3 | 6 | 50 | 5 | 32.2 | 15.7 |
| 22 | 84.9 | 5.7 | |||
| Co(acac)2 | 3 | 50 | 2 | 6.8 | 1.6 |
| 4 | 69.1 | < 0.5 | |||
| 25 | 7 | 7.3 | 0.8 | ||
| 20 | 100 | — | |||
| Ni(acac)2 | 3 | 25 | 0.25 | 19.0 | — |
| 0.5 | 68 | — | |||
| 20 | 100.0 | — | |||
8.2.3. Reaction Conditions and Product Analysis
The reactions of 1,4-cyclohexadiene and, for comparison, 1,3-cyclohexadiene were carried out at 25°C under a nitrogen atmosphere in toluene and dioxane. The ratio of nickel(II) acetylacetonate to triethylaluminum was 1:5, with a nickel concentration of 4·10-3 M in dioxane and 2·10-3 M in toluene. The dihydroaromatic was added in a 50-fold excess relative to the nickel complex. The only reaction products were cyclohexene, benzene, and small amounts of cyclohexane (< 2.5%), formed by disproportionation and/or hydrogenation of the cyclohexene. The ratio of benzene to cyclohexene increased from 1.0 to 1.1 toward the end of the reaction as disproportionation proceeded, corresponding to an absolute difference of four percent in the percentage product yields. While the conjugated diene could not be detected at any point during the disproportionation of 1,4-cyclohexadiene, the reaction of 1,3-cyclohexadiene resulted in the formation of the 1,4-isomer to a small extent (approximately 1% of the C6 hydrocarbons). The total amount of C6 hydrocarbons remained constant within the margin of error relative to the internal standard. Furthermore, no higher-condensed compounds were detected by gas chromatography.
8.2.4. Reaction Kinetics
The kinetics of the disproportionation of 1,4- and 1,3-cyclohexadiene were monitored by gas chromatography.
Disproportionation of 1,4-cyclohexadiene
The decrease in the concentration of 1,4-cyclohexadiene cannot be explained by a simple kinetic model (see Figure 24).
The curves are typical of an autocatalytic reaction (Schmid and Sapunov 1982) (page 54). Qualitatively identical curves were obtained for the reaction of 1,4-cyclohexadiene in toluene and in dioxane, in catalyst solutions that were two and twelve hours old,respectively, at Al:Ni ratios ranging from 3 to 6, and in the presence and absence of each of the two disproportionation products. The latter rules out autocatalysis by the reaction products and suggests the occurrence of an induction period during which the catalytically active complex is formed. This behavior has been described in the literature for the Ziegler-catalyzed exchange reaction of olefins with chiral aluminum alkyls (Giacomelli and Lardicci 1971),(Matkovskii et al. 1982).
\[ \ce{AlR2-CH2CRR'H + CH2=CHR'' <=>[Ni(mes)2] AlR2-CH2CH2R'' + CH2=CRR'} \] mes = N-methylsalicylaldimine ; note the chirality on the reactant carbon
The authors explained their findings using a kinetic model in which the catalytically active species is formed only as the reaction proceeds. Furthermore, their kinetic data on racemization were best approximated by a reaction scheme in which there is also a dynamic equilibrium between the catalytically active species and an inactive complex formed by complexation with excess aluminum alkyl. Deactivation of ZIEGLER catalysts by aluminum alkyls was also reported by KROLL for olefin hydrogenation (Kroll 1969). However, they were able to completely counteract the inhibitory effect by using equivalent amounts of a Lewis base such as dioxane. A precise analysis of the kinetics after the induction period was not feasible due to the high reaction rate and poor reproducibility.
Disproportionation of 1,3-cyclohexadiene
The disproportionation of 1,3-cyclohexadiene proceeds significantly slower than that of the 1,4-isomer, particularly in dioxane (see Figure 25). Straight lines are obtained from the rate equations. With respect to the reactant the reaction proceeds according to a zero-order law in dioxane and a first-order law in toluene. Clearly, different mechanisms are at work depending on the solvent.
No induction period was observed in either solvent. The active catalyst species is therefore formed so rapidly during the disproportionation of 1,3-cyclohexadiene that the induction period could not be detected analytically due to the slower subsequent steps involving the conjugated diene.
Competitive reaction between 1,3- and 1,4-cyclohexadiene
In a further experiment, the two isomeric cyclohexadienes were reacted in competition with each other in dioxane. Figure 26 shows the change in concentration over time of the reactants and the sum of the products. The two cyclohexadienes do not react independently of one another. The diagram shows that, despite its higher reactivity, the disproportionation of 1,4-cyclohexadiene does not become noticeable before the 1,3-cyclohexadiene has almost completely reacted. This finding can be explained by better coordination of the conjugated diene in the nickel complex, which blocks the reactive species for the 1,4-cyclohexadiene, although it is inherently faster reacting.
8.2.5. Mechanism of Disproportionation
The following schematic diagram explains the findings regarding the kinetics of disproportionation and the isomerization of the two cyclohexadienes:
The reduction of nickel(II) acetylacetonate with triethylaluminum yields a catalytically inactive initial complex, which, according to the literature (Schmidt et al. 1983), is a Ni(0) complex with aluminum alkylene as stabilizing ligands.
From this, an active complex (L(x+1)Ni) is formed through the action of cyclohexadiene in a slow reaction, as indicated by the induction period. No further conclusions can be drawn regarding its structure. Since the formation of ZIEGLER systems is discussed in the literature as radical reactions (Schmidt et al. 1983), this is also conceivable for the formation of this active complex.
In a fast equilibrium favoring the reactants, the cyclohexadiene then displaces a ligand L from the active species to form the adsorbate (CHD->NiLx), from which the nickel hydrido complex (C6H7-NiHLx) is formed by oxidative addition of the cyclohexadiene. The partial isomerization of 1,3-cyclohexadiene requires that the cyclohexadienyl group be present, at least in a secondary equilibrium, as a π3- or π5-bonded π-ligand (shown in the scheme only as a π3-ligand).
In rapid subsequent steps, a second cyclohexadiene molecule is then coordinated and hydrogenated. Based on the mechanisms of homogeneous catalytic hydrogenation described in the literature (Johnstone et al. 1985);(Gessner and Heesing 1985) and the findings obtained in our own research group regarding the analogous disproportionation of 1,2-dihydronaphthalene on a cobalt hydrido catalyst (Brock 1986), a stepwise process is likely.
The occurrence of the induction period during the disproportionation of 1,4-cyclohexadiene presumes that the rate constant k1 is small. In addition, the dehydrogenation rate constant k2 must be large enough to compensate for the small equilibrium constant K. As a result, the formation of the active complex — which is initially rate-limiting — has virtually no effect on the rapid dehydrogenation as the conversion increases.
No induction period has been observed for 1,3-cyclohexadiene. Accordingly, it is not k1 but k2 that is rate-determining, so that the induction period at the start of the disproportionation was not detected analytically. This is consistent with the reaction rate of 1,4-cyclohexadiene, which is two orders of magnitude higher than that of the conjugated diene after the induction period.
The scheme also explains the first-order reaction of 1,3-cyclohexadiene in toluene. In contrast, the disproportionation in dioxane proceeds much more slowly and follows a zero-order reaction with respect to 1,3-cyclohexadiene. The concentration of the diene is therefore not significant for the rate-determining step in this solvent. This finding can be reconciled with the reaction mechanism by assuming a much smaller constant k1 for the formation of the active complex in this medium compared to the other reactions, such that catalyst formation remains the rate-determining step throughout the entire observed time period.
The preferential disproportionation of the conjugated diene in the competitive experiment with the more reactive 1,4-cyclohexadiene can be explained by the different equilibrium positions in the complexation of the cyclohexadienes. For 1,3-cyclohexadiene, the equilibrium lies on the associate side, whereas for the 1,4-isomer, due to the weaker π-interactions, it lies further on the reactant side, similar to what was observed with colloidal nickel (Section 8.1.6). As a result, the active complex for the more reactive 1,4-cyclohexadiene is largely blocked, so that the less reactive, conjugated diene reacts first.
The different equilibrium positions and the long residence time of the 1,3-diene at the catalyst explain its isomerization to the thermodynamically less stable and — in the disproportionation under consideration — more reactive 1,4-cyclohexadiene, whereas no isomerization could be detected for the latter during its reaction.
Further conclusions regarding the mechanism are not possible based on the available findings. Given the complex nature of the reaction, which has not yet been exhaustively analyzed, the use of labeled substances—which could, in particular, elucidate the nature of the rate-determining step and the stereochemistry of the reaction—was not employed at this stage of the investigation.
9. Summary
Hydrogen transfers are of great importance in preparative organic chemistry as selective and mild oxidation and reduction reactions. For the applicability and understanding of these reactions, the most comprehensive knowledge of the mechanisms is essential. In addition to synchronous hydrogen transfers, stepwise processes involving ions and radicals as well as one-electron transitions are discussed. The subject of this work is mechanistic studies of two types of hydrogen transfer:
thermal transfer to quinones,
homogeneous-catalyzed disproportionation.
Dihydroaromatics are of particular interest as hydrogen donors because they are readily available and highly reactive. From a mechanistic perspective, 1,4-dihydroaromatics are particularly interesting, as there is ongoing debate whether their aromatization involves synchronous or two-step ionic dehydrogenation. In this work, 1,4-cyclohexadiene was therefore used as hydrogen donor.
Syntheses of Labeled 1,4-Cyclohexadienes
Five regio- and stereoselectively deuterated 1,4-cyclohexadienes were synthesized with the high isotopomeric purity required for kinetic measurements:
Their syntheses proceeded via a Diels–Alder reaction of labeled 1,3-butadienes with trans-β-chloroacrylic acid, followed by intramolecular β-elimination:
This key reaction yielded benzene-free products without H/D exchange. The butadiene precursors for 1a and 1e were obtained by base-catalyzed H/D exchange in non-deuterated or perdeuterated 2,5-dihydrothiophene-1,1-dioxide with D2O or H2O, respectively, e.g.:
[(Z)-1—D1]— and [(Z,Z)-1,4-D2]-1,3-butadiene were obtained by stereoselective reduction of the correspondingly configured chlorine derivatives with zinc/copper and D2O, e.g.:
[(E,Z)-1,4-D2]-1,3-butadiene, on the other hand, was prepared according to the multi-step synthesis by FLEMING [31]:
All syntheses of the indexed butadienes had to be optimized in terms of yield, degree of indexation, and stereochemical purity. This applies in particular to the isotopomers 1c and 1d, whose syntheses had proven to be barely reproducible and yielded purities that were, in some cases, unusable (Müller et al. 1984),(Hagemann et al. 1985). Further attempts to synthesize 1c and 1d via nucleophilic and electrophilic substitution from trans- and cis-3,6-bistrimethylsilyl-1,4-cyclohexadiene, respectively, were stereo-unselective. The steric arrangement of the deuterium atoms in the isotopomers 1c and 1d was confirmed, on the one hand, by 1H-NMR spectroscopy in the butadiene precursors and, on the other hand, in the cyclohexadienes based on their IR and Raman spectra as well as their stereoselective pyrolysis. The steric purity was > 90% (1c) and > 85% (1d), respectively.
Isotope Analysis
Mass spectrometric analysis of the C6 hydrocarbons was performed in part under low-voltage conditions (14–18 eV) to suppress [M-n]+ peaks. A new GC/MS method was developed for screening the benzene product mixtures: The measurements were performed at 80 eV. The [M-n]+ peaks could be accounted for arithmetically, since the molecular peak groups of the benzene isotopomers synthesized for this purpose were reproducible. Gas chromatographic separation of isotopomers yielded satisfactory results only for the cyclohexadienes. Partial separation of benzene isotopomers was only observed when the indexing differed significantly (C6H6/C6D6).
Mechanistic studies on thermally induced H-transfer to quinones
The two reactive quinones 2,3-dichloro-5,6-dicyano-1,4-benzoquinone (DDQ) and 3,4,5,6-tetrachloro-1,2-benzoquinone (OCA) — which are frequently used in preparative chemistry — were employed as hydrogen acceptors:
The reactions were carried out at 25°C in dioxane. They proceeded without side reactions according to a first-order kinetics law for quinone and cyclohexadiene. First, the total kinetic isotope effects for 1a — 1e and the product isotope effect of 1b were determined through kinetic measurements and by analyzing inter- and intramolecular competition. Analysis of the experiments revealed a consistent picture only for a two-step dehydrogenation: H-transfer to both quinones occurs via primary hydride transfer followed by proton abstraction. The high primary isotope effects of 13.6 (OCA) and 11.5 (DDQ) indicate a tunneling component in the hydride transfer. Tracer studies with 1c and 1d showed that dehydrogenation proceeds with high cis-selectivity. This selectivity is higher for OCA (> 95%) than for DDQ (91%). In contrast to DDQ, it remains consistently high with OCA even in highly polar solvents (acetonitrile, N-methylformamide). These findings demonstrate a mechanism involving a narrow intermediate ion pair, in which steric fixation is significantly more pronounced with OCA as the acceptor than with DDQ. The geometry of the transition state of the hydride abstraction was determined by measuring the temperature dependence of the primary isotope effect. The activation parameters obtained in this way suggest an angled transition state for the hydride transfer, which occurs with high tunneling contributions. At the same time, the limitations of this method as a mechanistic criterion became apparent: In the presence of tunneling contributions, the activation parameters allow reliable conclusions about the geometry of the transition state only when numerous additional experimental findings are available, as was the case for the reaction investigated in this work.
Homogeneous-catalyzed disproportionations of 1,4-cyclohexadiene
Homogeneous-catalyzed hydrogen transfer reactions have recently been increasingly used as an alternative to catalytic hydrogenation in preparative chemistry, as they often proceed regioselectively and stereoselectively under mild conditions. They are also attracting growing interest in industrial chemistry, such as coal processing. The mechanisms of dehydrogenation have so far been investigated for only a few systems. Alternative systems to the expensive transition metals (Pd, Pt, etc.) or their complexes (e.g., the Wilkinson catalyst) are attracting growing interest as catalysts.
Disproportionation on colloidal nickel
Colloidal nickel, given the easy availability of the catalyst components, serves as an active catalyst for the disproportionation of 1,4-cyclohexadiene without isomerization to the more stable 1,3-cyclohexadiene occurring. At 60°C in dimethylformamide, the reaction proceeds without the formation of byproducts, following a first-order kinetics law with respect to the catalyst and a second-order kinetics law with respect to the reactant. Tracer studies with 1a and 1e for H/D exchange, as well as with 1c and 1d for stereochemistry, and the determination of total isotope effects for 1a and 1b, as well as product isotope effects from 1b and 1c, provided further information that can be explained by the following reaction scheme:
The dehydrogenation is preceded by the adsorption of 1,4-cyclohexadiene as a fast equilibrium that favors the reactants. Dehydrogenation proceeds stepwise via oxidative addition and γ-H abstraction in two fast equilibria, both of which are shifted to the left. The cis-selectivity of the dehydrogenation is slightly less than 100 percent due to reversibility. Minor intramolecular H/D scrambling requires the presence of π-bonded ligands in addition to σ-bonded cyclohexadienyl groups. Hydrogenation is the rate-limiting step. It proceeds with loss of the molecular identity of the abstracted hydrogen, as is evident from tracer experiments with 1c and 1d.
Disproportionation on a ZIEGLER catalyst
The disproportionation of organic molecules on ZIEGLER complexes — which are known as active hydrogenation catalysts — is a reaction that has hardly been studied to date. Among the various ZIEGLER systems tested for the disproportionation of 1,4-cyclohexadiene, the nickel(II) acetylacetonate/triethylaluminum system proved to be the most active compared to Co, Cr, Mo, and Mn based systems. Furthermore, the reaction proceeds at 25°C in dioxane and toluene without isomerization or the formation of byproducts. The reaction kinetics suggest a complex reaction mechanism. The induction period observed with 1,4-cyclohexadiene, the different reaction orders observed for 1,3-cyclohexadiene in dioxane and toluene (dioxane: 0th order, toluene: first-order), as well as the preferred conversion of the less reactive 1,3-cyclohexadiene in the competitive reaction between the two cyclohexadienes, can be explained by the following simplified reaction scheme:
The catalyst components form an initial complex, which is modified under the influence of cyclohexadiene into the active complex for disproportionation. From this complex, in a rapid, left-handed equilibrium, a ligand is displaced by a cyclohexadiene molecule, which is gradually dehydrated in the rate-determining step via oxidative addition followed by β- or γ-H abstraction. The subsequent rapid steps correspond to the hydrogenation mechanism observed with other transition-metal complexes.
Experimental Section
1. General Methods
1.1. UV Spectroscopy
The kinetic measurements were performed using a CARY 219 spectrophotometer from VARIAN. The cuvette holder was connected to an NB 22 thermostat from HAAKE. The temperature was maintained at ± 0.2°C.
1.2. IR Spectroscopy
The spectra were recorded using a PERKIN ELMER Gräting Spectrometer, Model 421, as a thin film between NaCl plates.
1.3. Raman Spectroscopy
A CODERG T 800 spectrometer was used to record the Raman spectra. Excitation was provided by an argon laser at 514 nm (Schnöckel 198z). The spectra were recorded at room temperature. Approximately 5 μl of the substance was fused in glass capillaries intended for determining melting points.
1.4. 1H-NMR Spectroscopy
The 1H-NMR spectra were recorded using the PFT WM 300 spectrometer from BRUKER. Deuterium- decoupled 1H-NMR spectra were obtained using a BRUKER WM 200 (Bandmann 198y). TMS was used as the standard was TMS. Unless otherwise specified, the solvent was deuterated chloroform.
1.5. 13C-NMR Spectroscopy
The spectra were recorded on a BRUKER WM 300 NMR instrument using deuterated chloroform as the solvent and TMS as the internal standard. Both 1H broadband and off-resonance-decoupled spectra were recorded.
1.6. Melting Points
The melting points were measured using a KOFLER hot-stage microscope from REICHERT and have been corrected.
1.7. Refracting Indices
The refractive indices were determined using an Abbé refractometer from ZEISS.
1.8. Gas Chromatography
1.8.1. Analaytical Methods
Analytical determinations were performed using the F22 instrument from PERKIN-ELMER and the Aerograph Series 2700 from VARIAN. Retention times and peak areas were determined using the 3390A integrator from HEWLETT PACKARD and the Autolab Minigrator from SPECTRA PHYSICS. The relative error of the integration was ± 5%. The standard conditions used for the analysis are described in the following chapters. Any deviations from these conditions are listed separately.
1.8.1.1. Separation of C4 Hydrocarbons & Isotopomers
The analysis of indexed and substituted C4 hydrocarbons, as well as the separation of the isotopomers of 1,4-cyclohexadiene and benzene, respectively, was performed using a 130m long Duranglas capillary (0.2 mm inner diameter), which was prepared from a 1.50m long capillary tube (6 mm outer diameter, 2.2 mm inner diameter; rinsed with distilled water, acetone, and air-dried) using a HEWLETT/PACKARD HP 1045A capillary drawing machine at a drawing ratio of 20. The capillary, deactivated with octamethylcyclosiloxane (350°C, 70 h), was statically packed with GE-SE-52 (0.2% in n-pentane) at room temperature.
A 1-pl (picoliter) sample of a 1% solution in n-pentane was injected at a split ratio of 1:100. The following conditions were applied:
| Substances | Column | Injector | Detector | Pressure |
|---|---|---|---|---|
| 1,3-Butadiene | 25°C | 50°C | 200°C | 0.3 bar |
| 1,4-Dichloro-1,3-butadiene | 70°C | 150°C | 200°C | 1.1 bar |
| C6 Hydrocarbons | 25°C | 150°C | 200°C | 0.2 bar |
1.8.1.2. Non-Indexed C6 and C8 Hydrocarbons
The analysis of non-indexed C6 and C8 hydrocarbons was performed on a 2.5 m long glass column (2 mm inner diameter), deactivated with dimethyldichlorosilane (5% in toluene, 25°C, 24 h), on Chromosorb WAW DMCS 100/120 packed with 20% FRAKTONITRIL III (1,2,3-tris(2-cyanoethoxy)propane). The C6(8) hydrocarbons were separated isothermally at 75°C with a nitrogen flow rate of 25 ml/min. After four minutes, the temperature was raised at a rate of 12.5°C/min to 90°C to separate higher hydrocarbons. The injector temperature was 170°C, and the detector temperature was 200°C.
1.8.1. Preparative Methods
A VARIAN Aerograph Type 940 was used for preparative separations. A 4 m long glass column (8 mm inner diameter) was used, packed with Chromosorb W 45/60 AW, coated with 20% FRAKTONITRIL II (1,2-bis(2-cyanoethoxy)propane). The separation was performed isothermally at 80°C with a nitrogen flow rate of 30 ml/min. The injector temperature was 165°C, the detector temperature was 290°C, and the preparative outlet was 135°C. In each case, 0.5 mL of a 1.0–2.5% solution of C6(8) hydrocarbons in n-pentane was injected. The fractions were collected in U-tubes (8 mm inner diameter), which were filled in the front leg with approximately 2 mL of Chromosorb P 60/80 and immersed in a cooling trap at -78°C (see figure). After isolation, the section of the glass tube filled with the adsorbent was cut out using an ampoule cutter and subjected directly to mass spectrometric analysis without intermediate storage. If the preparative separation was performed to purify 1,4-cyclohexadienes intended for 1H-NMR spectroscopic analysis, these were condensed on quartz wool in the front leg of the U-tube and subsequently eluted directly with deuterated chloroform. If 1,4-cyclohexadiene was to be purified prior to pyrolysis, condensation could only take place on the glass wall. The purity of the collected fractions was determined by gas chromatography and was > 99.5%.
1.9. Mass Spectrometric Isotope Ratio Measurements
1.9.1. GC/MS Coupling
The GC/MS isotope ratio measurements were performed on the CH7 mass spectrometer from VARIAN. The C6 hydrocarbons were separated on a 50 m long quartz capillary (inner diameter 0.25 mm) with a chemically bonded FFAP phase from MACHEREY-NAGEL. The carrier gas was helium at a pre- pressure of 0.2 bar. The column temperature was 60°C, and the injector temperature was 170°C. Approximately 0.2 μl of an approximately 1% solution in n-pentane was injected. The ionization energy was 70 eV. For each GC peak, approximately four to five spectra were recorded in the mass range of m/e = 70–90. The signals from each spectrum were stored using the CAT (Computer Averaged Transients) method and then averaged across all spectra. This method largely compensated for systematic errors, such as the extreme change in ion current at the flanks of the GC peak or isotopomer separation on the column. Since the GC/MS analyses exhibit a relatively high statistical error, each sample had to be measured at least ten — or better yet, twenty — times. The concentration of an isotopomer could then be determined with an accuracy of +/- 1.0–2.0 % accuracy. The high ionization energy led to pronounced (M-n)+ peaks for 1,4-cyclohexadiene (> 50% relative to M+), whereas for benzene they were less than 20%. For quantitative analyses, these were taken into account by measuring the molecular peak groups of the relevant benzene isotopomers separately. They proved to be highly reproducible and yielded the following patterns:
| m/e | [D0]-Benzol | [D1]-Benzol | [1,4-D2]-Benzol |
|---|---|---|---|
| 74 | 4.5 +/- 0.3 | ||
| 75 | 1.6 +/- 0.1 | 3.0 +/- 0.1 | |
| 76 | 8.3 +/- 0.3 | 2.1 +/- 0.1 | 2.1 +/- 0.2 |
| 77 | 16.0 +/- 0.7 | 9.1 +/- 0.2 | 1.9 +/- 0.2 |
| 78 | 100.0 | 18.3 +/- 0.8 | 8.3 +/- 0.5 |
| 79 | 6.8 +/- 0.2 | 100.0 | 11.9 +/- 0.7 |
| 80 | 6.7 +/- 0.1 | 100.0 | |
| 81 | 6.7 +/- 0.3 |
1.9.2. Push-Rod Measurements
The analysis of the indexed 1,4-cyclohexadienes, as well as some of the reaction products isolated by preparative gas chromatography, was performed via direct injection through the push rod, also on the VARIAN CH7 mass spectrometer. The sample vials (Section 1.8.2) were connected to the push rod via a coolable device designed by U. GESSNER and R. PAUKSTAT (Gessner 1984) — and further developed for the measurements carried out in this work — to the push rod. To achieve the vacuum of < 10-6 Torr required for the measurement, the sample holder had to be cooled with liquid nitrogen. Leaks occurring at these temperatures in the Teflon seal between the push rod and the sample holder were prevented by fabricating the latter with double the length. This also ensured a constant vacuum of <10-6 Torr at a sample temperature of –178°C even during longer measurement periods. By carefully heating the sample (the immersion depth of the soldered copper strips in the nitrogen was slightly reduced), an ion current suitable for the measurement could then be established. The molecular peak group was scanned at least ten times at an ionization energy of 14 to 16 eV. The time required to record one scan was approximately two seconds. The indexing was calculated from the peak heights using established methods. Comparative measurements on the non-indexed substances allowed us to rule out systematic errors (e.g., peak discrimination), which would have manifested as 13C isotope peaks that were either too small or too large. In addition, the [M—1]+ peaks occurring in cyclohexadiene, which accounted for approximately 4% of the intensity of the M+ peak, were taken into account. The statistical error of a series of measurements was approximately 0.5–1.0% for the percentage fraction of an isotopomer (absolute error).
2. Synthesis of 1,4-Cyclohexadiene
Following the general procedure described below, all 1,4-cyclohexadienes used as starting materials were prepared from the corresponding 1,3-butadienes.
2.1. 1,3-Butadiene
By pyrolysis of 9.4 g (80 mmol) of 2,5-dihydrothiophene-1,1-dioxide (Cope et al. 1961) at 150°C. The butadiene was freed from SO2 by washing with 10% NaOH solution and condensed at -78°C.
Yield: approx. 6 mL (> 90%) (Ref. (Cope et al. 1961) 88%)
2.2. 2-Chloro-4-cyclohexene-1-carboxylic acid
By the Diels-Alder reaction of 1,3-butadiene with E-3-chloropropenoic acid, analogous to the procedure described by W.P. NORRIS (Norris 1968): The 1,3-butadiene obtained according to 2.1 was condensed at -78°C into a glass ampoule (length: 150 mm, inner diameter: 11 mm, wall thickness: 4 mm) and mixed with 10.7 g (100 mmol) of E-3- chloropropenoic acid. The ampoule, which had been melted down under 0.1 Torr, was heated to 100°C in an autoclave for 120 hours, with an external nitrogen pressure of 25–50 bar to prevent the ampoule from bursting. After the reaction was complete, the non-volatile products were dissolved in a solution of 9.9 g (100 mmol) KHCO3 in 35 mL water and separated from polymeric byproducts by filtration and extraction with n-pentane. The product obtained by precipitation with semi-concentrated hydrochloric acid was contaminated with approximately 30% of the starting material and was recrystallized from n-hexane.
Yield: 5.1 g (approx. 45%) (based on butadiene; Ref. (Norris 1968) 68%)
Melting point: 110°C (from n-hexane) (Ref. (Norris 1968) 110–112°C)
Increasing the temperature and extending the reaction time did not increase the yield; they merely led to increased polymer formation.
2.3. 1,4-Cyclohexadiene
4.8 g (30 mmol) of the acid obtained in 2.2 were converted to the potassium salt with 3.0 g (30 mmol) of KHCO3 in 20 mL of water. The water was removed on a rotary evaporator at 17 Torr and 15–20°C, and the residue was heated to 60°C with 4.5 g (30 mmol) of NaI in 50 mL of hexamethylphosphoric triamide for four hours at 0.1 Torr. The product was condensed at –78°C, distilled into a second cold trap at 1 Torr and 0°C for purification, and dried over Na2SO4.
Yield: 1.4 g (60%) (Ref. (Norris 1968) 70%)
GC analysis: > 99.5% (Ref. (Norris 1968) 98%)
1H-NMR: δ = 2.67 (t; J2,3 = 1.1 Hz, 4 H, 3- and 6-H), 5.69 (t; 4 H, 1-,2-,4- und 5-H)
3. Synthesis of Labeled Compounds
3.1. [3,3,6,6-D4]-1,4-Cyclohexadiene
3.1.1. [2,2,5,5-D4]-2,5-Dihydrothiophene-1,1-dioxide
9.4 g (80 mmol) of 2,5-dihydrothiophene-1,1-dioxide was stirred in 5 mL of absolute dioxane with 9.4 g (470 mmol) of deuterium oxide and 0.1 g of K2CO3 for 48 hours at room temperature (Cope et al. 1961). After removing the solvent by freeze-drying, the process was repeated. The degree of deuteration was determined by 1H-NMR spectroscopy after each H/D exchange. The allyl hydrogen content was < 1% after five exchange cycles. The degree of indexation was additionally determined using a low-voltage push-rod measurement. The sample holder needed only to be slightly warmed with a hair dryer to generate a suitable ion current. The molecular peak group of the butadiene forming in the mass spectrometer was measured. The [M-n]+ peaks amounted to approximately 2% of the M+ peak for non-indexed butadiene. The degree of indexation was thus determined to be: 99.3% D4, 0.7% D3 (Ref. (Cope et al. 1961) 95.5% D4, 4.5% D3).
3.1.2. [3,3,6,6-D4]-1,4-Cyclohexadiene
The synthesis was carried out in the same manner as described in Section 2).
Yield: 1.2 g (59%)
GC analysis: > 99.5%
1H-NMR: δ = 5.69 (s; 1-, 2-, 4-, and 5-H); the signal from the allyl hydrogens is completely absent. 13C-NMR: δ = 24.93 (quint; C-3 and C-6, J(¹³C-2H) = 19.5 Hz), 124.27 (d; C-1, -2, -4, and -5)
Isotope ratio measurement: 99.6% D4, 0.4% D3 (Ref. (Norris 1968) 96% D4, Ref. (Müller et al. 1984) 90–95%)
3.2. [3-D1]-1,4-Cyclohexadiene
3.2.1. (Z)-1-Chloro-1,3-butadiene
By reacting 31.3 g (250 mmol) of (E)-1,4-dichlorobut-2-ene (88% E-isomer) with a molten mixture of 42.0 g (750 mmol) of KOH and 7.5 g (417 mmol) of water at 120–130°C (Bartlett et al. 1967). The product was distilled off during the reaction, dried over CaCl2, and redistilled.
Yield: 12.0 g (54%) (Ref. (Bartlett et al. 1967) 53%)
Boiling point: 68°C (Ref. (Bartlett et al. 1967) 66°C)
GC analysis (4 m steel column, SE 30 4%, 30°C, 20 mL/min): >99.5%
3.2.2. [(Z)-1-D1]-1,3-Butadiene
The synthesis was carried out by analogy with the procedure described by STEPHENSON (Stephenson et al. 1977) under strict exclusion of oxygen: 39.2 g (600 mmol) of zinc powder was suspended in 60 mL of water and, under nitrogen, 130 mL of 0.15 M CuCl2 solution containing 0.5% HCl was added dropwise. The zinc/copper alloy was filtered through a vacuum frit, washed to remove salts, washed successively with acetone, deuterium oxide, acetone, and ether, and dried under vacuum. The reaction with 9.0 g (100 mmol) of the chlorobutadiene obtained in 3.2.1 and 24.0 g (1.2 mol) of deuterium oxide was carried out in 180 mL of absolute tetrahydrofuran under reflux and a slow stream of nitrogen over a period of two hours. The deuterated butadiene was condensed at -78°C into a glass ampoule for further reaction.
Yield: approx. 7 mL (90%)
GC analysis (see Section 3.2.1.): 97% 1,3-butadiene, 3% tetrahydrofuran
3.2.3. [3-D1]-1,4-Cyclohexadiene
The synthesis was carried out as described in Section 2).
Yield: 1.1 g (64%)
GC analysis: >99.5%
1H-NMR: δ = 2.65 (m; 3.08 H, 3- and 6-H), 5.69 (m; 4 H, 1-, 2-, 4-, and 5-H)
13C-NMR: δ = 25.34 (t; C-3, J(¹³C-2H) = 19.5 Hz), 25.72 (t; C-6), 124.27 (d; C-1, -2, -4, and -5)
Isotope ratio measurement: 99.3% D1, 0.7% D0 (Ref. (Stephenson et al. 1977) 95–96% D1)
3.3. [cis-3,6-D2]-1,4-Cyclohexadiene
3.3.1. 1,3,4,4-Tetrachlorobut-1-ene (E/Z isomer mixture)
100 mL (1.30 mol) of (E)-1,2-dichloroethylene was reacted with 3.2 g (13 mmol) of dibenzoyl peroxide (75%) in a glass ampoule (length: 250 mm, inner diameter: 26 mm, wall thickness: 2 mm) in an autoclave at 70°C for 70 hours (Frank and Blackham 1950). The contents of the ampoule were dissolved in 100 mL CH2Cl2, excess peroxide was removed with an aqueous FeSO4 solution, and the cis/trans mixture of tetrachlorobutene was obtained by fractional distillation.
Yield: 42.8 g (24%) (Ref. (Frank and Blackham 1950) 30%)
Boiling point: 78–81°C/17 Torr (Ref. (Frank and Blackham 1950) not specified)
3.3.2. (Z,Z)-1,4-Dichloro-1,3-butadiene
42.0 g (210 mmol) of the isomer mixture obtained in 3.3.1 was slowly added dropwise, while cooling with ice, to a suspension of 112 g (1.70 mol) of zinc in 200 mL of methanol; the reaction solution was stirred for 24 hours at room temperature, filtered, and extracted with n-pentane (Braye 1963). The solvent was largely removed at 30°C under normal pressure using a rotary evaporator.
Yield: approx. 80% (GC analysis (Ref. (Braye 1963) 77.5%) Relative stereoisomer ratio (GC analysis):
32% (Z,Z)-1,4-dichloro-1,3-butadiene
50% (E,Z)-1,4-dichloro-1,3-butadiene
18% (E,E)-1,4-dichloro-1,3-butadiene
(Ref. (Bartlett and Wallbillich 1969) 30% (Z,Z)-, 50% (E,Z)-, 20% (E,E)-isomers)
The (Z,Z)-1,4-dichloro-1,3-butadiene was separated from the other two isomers by distillation. The separation was carried out in a FISCHER MMS 200 split-column distillation apparatus (25 theoretical plates, 0.5 mL working volume) at 100 Torr. The reflux ratio was adjusted so that approximately 1–2 mL of product was obtained within 12 hours. The temperature gradient on the column was approximately 5°C. With only a slight increase in distillation rate, the isomeric purity of the distillate dropped to < 90%.
Boiling point: 68°C/100 Torr (Ref. (Porri and Aglietto 1976) 64°C/96 Torr)
GC analysis: 97.5% (Z,Z)-1,4-dichloro-1,3-butadiene, 2.5% (E,Z)-1,4-dichloro-1,3-butadiene
n20D = 1.5210 (Ref. (Porri and Aglietto 1976) 1.5212)
1H-NMR: δ = 6.21 (ddd; J1,2 = 4.5 Hz, J1,3 = 1–5 Hz, J1,4 = 0.3 Hz, 2 H, 1- and 4-H), 6.74 (ddd; 2 H, 2- and 3-H)
3.3.3. [(Z,Z)-1,4-D2]-1,3-Butadiene
12.0 g (100 mmol) of (Z,Z)-1,4-dichloro-1,3-butadiene was prepared according to the method described by STEPHENSON (Stephenson et al. 1977) with 20 g (1.00 mol) of deuterium oxide and a zinc/copper alloy (prepared from 60 g (920 mmol) of zinc) for four hours in 200 mL of absolute dioxane under a gentle stream of nitrogen while refluxing. The resulting butadiene was passed through a cooling trap (0°C) and a CaCl2 tube and condensed at -78°C.
Yield: approx. 7.5 mL (> 90%) (Ref. (Stephenson et al. 1977) 70–90%)
GC analysis: 92% 1,3-butadiene 1H-NMR; δ = 4.97 (d; JE1,2 = 8.8 Hz, 1.90 H, E1- and E4-H), 5.08 (d; JZ1,2 = 16.3 Hz, 0.10 H, Z1- and Z4-H), 6.23–6.28 (m; 2 H, 2- and 3-H)
3.3.4. [cis—3,6—D2]—1,4—Cyclohexadiene
The synthesis was carried out as described in Section 2).
Yield: 1.3 g (60%)
GC analysis: >99.5%
1H-NMR: δ = 2.65 (m; 2.02 H, 3- and 6-H), 5.694 (d; J2,3 = 0.8 Hz, 4 H 1-, 2-, 4-, and 5-H)
13HC-NMR: δ = 25.40 (t; C-3 and -5, J(13C-2H) = 19.5 Hz), 124.40 (d; C-1, -2, -4, and -5)
Isotope ratio measurement: 96.0% D2, 4.0% D1 (Ref. (Müller et al. 1984) 84–86% D2)
IR: τCHD: 835 m, 870 vw, γCHD: 895 m, 910 vs, δCHD: 1270 s, 1280 s, νC=C: 1635 s, 1645 shoulder, 1653 m, 1662 m cm-1
Raman: τCHD: 835 s, γCHD: 910 vw, δCHD: 1257 vs, 1276 s; νC=C: 1677 s cm-1
Steric purity: see Section 4.
3.4. [trans-3,6-D2]-1,4-Cyclohexadiene
3.4.1. (2R,5S)-Tricyclo[4.2.2.0]deca-3,7,9-triene-7,8-dicarboxylic acid dimethyl ester
31.5 g (300 mmol) of freshly distilled cyclooctatetraene and 39.1 g (280 mmol) of bisdimethoxycarbonylethene were heated under reflux for 8 hours at a bath temperature of 150–153°C (Avram et al. 1957). The product was distilled off via a microbridge (max. bath temperature 140°C). Under these conditions, significant amounts of dimethyl phthalate were already formed as a decomposition product (approx. 40 molar percent). No separation was performed. The yield was determined by 1H-NMR spectroscopy.
Yield (mixture): 27.0 g (43%) as a mixture (Ref. (Avram et al. 1957) 50%)
Yield (product): 17.2 g (25.4%) product (1H-NMR)
Boiling point: 70–90°C/0.05 Torr (Ref. (Bartlett and Wallbillich 1969) 140–150°C/1 Torr)
1H-NMR: δ = 2.72 (s; 2 H, 7- and 10-H), 3.78 (s; 6 H, methyl-H of the tricycle), 3.85 (m; 2 H, 1- and 4-H), 3.91 (s; 6 H, methyl-H of the dimethyl phthalate), 6.07 (s; 2 H, 8- and 9-H), 6.13 (m; 2 H, 5- and 6-H), 7.5–7.75 (m; 4 H, aromatic H of dimethyl phthalate)
3.4.2. (2R,3S,4R,5S)-[3,4-D2]-Tricyclo[4.2.2.0]deca-7,9-diene-7,8-dicarboxylic acid dimethyl ester
26.7 g (70 mmol) of the mixture obtained in 3.4.1. was deuterated according to the method of COPE et al. (Cope et al. 1952) with 0.6 g of palladium-activated carbon (MERCK, 10% Pd) in 60 mL of absolute methanol at room temperature under normal pressure until 1500 cmsup>3 of deuterium (0.97 mole equivalents) had been incorporated.
Yield of crude product: 26.1 g (97%)
1H-NMR: δ = 1.10 (m; 2 H, 8- and 9-H), 2.31 (m; 2 H, 7- and 10-H), 3.67 (s; 6 H, methyl-H) 3.92 (m; 2 H, 1- and 4-H), 6.43 (m; 2 H, 5- and 6-H)
In addition, the phthalic acid ester and starting material peaks appeared. The product still contained approximately 15% of the starting material. The corrected yield is 85%.
3.4.3. cis-[3,4-D2]-1-Cyclobutene
By pyrolysis of the tricycle synthesized in 3.4.2. according to the method described by COPE et al. (Cope et al. 1952) at 200–210°C and 100 Torr in a weak nitrogen stream. The reaction was complete after three hours.
Yield: approx. 4 mL (approx. 90%) (Ref. (Cope et al. 1952) 95%)
GC analysis: 70% cyclobutene, 30% 1,3-butadiene
3.4.4. [(E,Z)-1,4-D2]-1,3-Butadiene
By pyrolysis of the cyclobutene obtained in 3.4.3 at 285 ± 10 °C and 100 Torr in a weak nitrogen stream using the apparatus outlined below. After three hours, all of the cyclobutene had been distilled off.
Yield: approx. 4 mL (approx. 100%)
GC analysis: >99.5% 1,3-butadiene
1H-NMR: δ = 4.97 (d; JE1,2 = 8.8 Hz, 1 H, E1- and E4-H), 5.08 (d; JZ1,2 = 16.3 Hz, 1 H, Z1- and Z4-H), 6.23–6.28 (m; 2 H, 2- and 3-H)
3.4.5. [trans-3,6-D2]-1,4-Cyclohexadiene
The synthesis was carried out as described in Section 2).
Yield: 0.41 g (61%)
GC analysis: > 99.5%
1H-NMR: δ = 2.65 (m; 2.08 H, 3- and 6-H), 5.69 (d; J2,3 = 1.2 Hz, 4 H, 1-, 2-, 4-, and 5-H)
13C-NMR: δ = 25.40 (t; C-3 and C-6 CHD, J(13C-2H) = 19.5 Hz), 25.77 (t; C-3 and C-6 CHH), 124.4 (d; C-1, -2, -4, and -5)
Isotope ratio measurement: 91.6% D2, 8.4% D1 (Ref. (Fleming and Wildsmith 1970) 96% D2, Ref. (Müller et al. 1984), (Hagemann et al. 1985) 84–86% D2)
IR: τCHD: 840 vvw, 870 s, γChd: 910 s, δCHD: 1275 vs, νC=C: 1635 s, 1645 s cm-1
Raman: τCHD: 822 s, 837 shoulder, 877 vvw, δCHD: 1205 vs, 1276 m; νC=C: 1673 vs; cm-1
Stereochemical purity: see Section 4.
3.5. [1,2-D2]-1,4-Cyclohexadiene
3.5.1. [1,1,2,3,4,4-D6]-1,3-Butadiene
The reduction of hexachloro-1,3-butadiene was carried out according to a method by CRAIG and FOWLER (Craig and Fowler 1961). 52.0 g (200 mmol) of perchlorobutadiene was added over a period of two hours under a gentle stream of nitrogen to a boiling suspension of zinc-copper (prepared in situ from 98.0 g (1.50 mol) of zinc, 4.8 g (50 mmol) of CuCl, and 0.8 g (10 mmol) of NaI) in 120 mL of absolute dioxane and 30.0 g (1.5 mol) of deuterium oxide. The butadiene, passed through a cooling trap (0°C) and a CaCl2 tube, was condensed at -78°C into a glass ampoule (length: 200 mm, inner diameter: 17 mm, wall thickness: 2 mm).
Yield: approx. 10 mL (approx. 60%) (Ref. (Charlton and Agagnier 1973) 44%, Ref. (Craig and Fowler 1961) 59%)
3.5.2. [2,2,3,4,5,5-D6]-2,5-Dihydrothiophene-1,1-dioxide
To prepare hexadeuterobutadiene, 12 mL of SO2 was condensed into the ampoule and fused under vacuum together with 0.10 g of hydroquinone. The ampoule was heated to 100°C in an autoclave (25– 50 bar back pressure) for 12 hours. The crude product was dissolved in water, purified by filtration to remove polymers, and recrystallized from methanol.
Yield: 13.7 g (92%) (Ref. (Charlton and Agagnier 1973) 84%)
Melting point: 63°C (Ref. (Charlton and Agagnier 1973) 63.8°C)
3.5.3. [3,4-D2]-2,5-Dihydrothiophene-1,1-dioxide
The deuterated sulfolene was treated with water and K2CO3 as described in Section 3.1.1. The hydrogen incorporation was monitored by 1H-NMR spectroscopy using anisol as an internal standard. This process was repeated eleven times.
Yield: 11.1 g (84%) (Ref. (Charlton and Agagnier 1973) 87%)
Isotope ratio measurement (Section 3.1.1): 98.4% D2, 1.6% D1 (Ref. (Charlton and Agagnier 1973) 86.0% D2, 9.5% D1, 4.2% D0, determined in [2,3-D2]-anthracene as a byproduct of the synthesis)
3.5.4. [1,2-D2]-1,4-Cyclohexadiene
The synthesis was carried out as described in Section 2).
Yield: 1.2 g (59%)
GC analysis: >99.5%
1H-NMR: δ = 2.65 (d; J3,4 = 1.1 Hz, 4 H, 3- and 6-H), 5.69 (d; 2.03 H, 4- and 5-H)
13C-NMR: δ = 25.59 (t; C-3, C-6), 123.91 (t; C-1, C-2, J(13C-2H) = 24.0 Hz), 124.36 (d; C-4, C-5)
Isotope ratio measurement: 98.6% D2, 1.4% D1
3.6. Further Experiments on the Synthesis of [trans-3,6-D2]-1,4-Cyclohexadiene
3.6.1. trans-3,6-Bistrimethylsilyl-1,4-cyclohexadiene
The synthesis and isomer separation were carried out according to a procedure by KEIL and EFFENBERGER (Keil and Effenberger 1982). 9.5 g (120 mmol) of benzene was slowly added under nitrogen to a solution of 49.7 g (460 mmol) of trimethylchlorosilane and 2.5 g (360 mmol) of lithium in 100 mL of absolute tetrahydrofuran. After the addition was complete, the mixture was stirred for 36 hours at room temperature and filtered under a nitrogen atmosphere. The remaining LiCl was precipitated with n-pentane, the organic phase filtered, concentrated to dryness on a rotary evaporator, and the remaining solvent was removed under vacuum.
Yield: 23.9 g (89%) (Ref. (Keil and Effenberger 1982) 94%)
GC analysis (4 m steel column, OV 22 10%, 130°C/4 min; 10°C/min; 190°C/8 min): trans/cis isomer ratio = 3.2, with 2.5% 1,4-bistrimethylsilylbenzene present.
The mixture of isomers was dissolved in 90 mL of absolute ethanol, the solution was cooled to ‒5°C, and the trans product was collected after 24 hours in a vacuum flask cooled to 0°C, washed with a small amount of cold ethanol, and dried under vacuum.
Yield: 8.0 g (11%)
GC analysis: > 99.5% trans-3,6-bistrimethylsilyl-1,4-CHD
The second crystal fraction (1.7 g) contained 1% of the cis isomer.
3.6.2. [cis,trans-3,6-D2]-1,4-Cyclohexadiene
3.6.2.1. Basic Hydrolysis
Based on a synthesis by DUNOGUSS et al. (Dunogues et al. 1972), a solution of 4.0 g (20 mmol) of trans-3,6-bistrimethylsilyl-1,4-cyclohexadiene in 10 mL of hexamethylphosphoric acid triamide (HMPTA) was added dropwise over 30 minutes at 10 Torr to an ice-cooled solution of 1.54 g (30 mmol) of KOD in 10 mL of deuterium oxide. The mixture was stirred for another three hours at room temperature. The volatile products, condensed at -78°C, were washed with water, dried over Na2SO4, and distilled into a second cold trap.
Yield: 1.5 g (30%) (Ref. (Dunogues et al. 1972) 90%)
GC analysis: 56% 1,4-cyclohexadiene, 44% hexamethyldisiloxane
The cyclohexadiene was purified via preparative gas chromatography for further use.
Isotope ratio measurement: 91.2% D2, 7.5% D1, 1.3% D0 Stereochemical purity: see Section 4.
3.6.2.2. Reaction with Trifluoro[OD]acetic Acid
6.0 g (27 mmol) of trans-3,6-bistrimethylsilyl-1,4-cyclohexadiene in 30 mL of n-pentane was added dropwise over 30 minutes under a nitrogen atmosphere at 0°C to a solution of 15 g (130 mmol) of trifluoro[OD]acetic acid in 30 mL n-pentane. The reaction mixture was stirred for heating to room temperature, quenched with 30 mL of deuterium oxide, and the remaining trifluoroacetic acid was extracted with water. The reaction was monitored by gas chromatography, as more highly condensed derivatives of the initially formed isomeric cyclohexadienes were increasingly produced as the reaction progressed. The reactant was thereby converted quantitatively.
GC analysis (4 m steel column, OV 22 10%, 70°C/4 min; 10°C/min; 190°C/16 min): 32% C6 hydrocarbons, 29% hexamethyldisiloxane, 19% starting material, 20% higher-condensed byproducts. The dried pentane phase was distilled under normal pressure using a microcolumn.
Boiling point: 92°C
Yield: 3.4 mL
GC analysis (2.5 m glass column, FRAKTONITRIL III, (see Section 1.8.1.2)): 59% hexamethyldisiloxane, 15% 1,3-cyclohexadiene, 21% 1,4-cyclohexadiene, 1% benzene, 4% higher- condensed byproducts. The 1,4-cyclohexadiene was isolated via preparative GC for further characterization.
GC analysis (glass column, see above): 0.1% hexamethyldisiloxane, 0.1% 1,3-cyclohexadiene, 96.9% 1,4-cyclohexadiene, 0.9% benzene, 2.0% higher-condensed compounds.
1H-NMR: δ = 2.65 (m; 1.94 H, 3- and 6-H of the CHD group), 2.68 (t; J2,3 = 1.0 Hz, 0.20 H, 3- and 6-H of the CH2 group), 5.69 (d; 4 H, 1-, 2-, 4-, and 5-H)
13C-NMR: δ = 25.39 (t; C-3 and -6 of the CHD group, J(13C-2H) = 19.4 Hz), 25.77 (t; C-3 and -6 of the CH₂ group), 124.40 (d; C-1, C-2, C-4, and C-5)
Isotopic composition: > 95% D2 (determined by 1H-NMR) Stereochemical purity: see Section 4.
3.6.3. [cis,trans-1,2,3,4,5,6—D6]—1,4-Cyclohexadiene
The synthesis was carried out as described in Sections Section 3.6.1 and Section 3.6.2.1, starting from [D6]-benzene.
Isotope ratio measurement: 96.4% D6, 3.6% D5
Steric purity: see main text, Stereochemical purity: see Section 4.3, Figure 2.
3.7. [1-D1]-Benzene
Prepared by reacting the Grignard compound derived from bromobenzene and magnesium with deuterium oxide. Purification was carried out by distillation followed by preparative gas chromatography.
Yield: 84%
Isotope ratio measurement: 95.1% D1, 4.9% D0 (Ref. (Todd 1973) 91.5% D1, Ref. (Koppang et al. 1986) 90–95% D1, Ref. (Asomaning et al. 1973) > 99% D1, Ref. (Mueller et al. 1976) > 98% D1)
3.8. [1,4-D2]-Benzene
Obtained by quantitative dehydrogenation of [3,3,6,6-D4]-1,4-cyclohexadiene with 2,3-dichloro-5,6-dicyanobenzoquinone (DDQ). The isotopomeric purity was derived from that of the cyclohexadiene (see Section 3.1.2).
Ref. (Angelini et al. 1980): 88% D2, 8% D1, 4% D0
3.9. [1,2-D2]-Benzene
Analogous to Section 3.8, prepared from [1,2-D2]-1,4-cyclohexadiene.
4. Pyrolysis of Labeled 1,4-Cyclohexadienes
Pyrolysis was carried out in Duranglas ampoules (length: 150 mm, inner diameter: 11 mm, wall thickness: 4 mm, volume: approx. 14 mL). 20 mg (0.2 mmol) of the cyclohexadiene was melted at 0.01 Torr and pyrolyzed for one hour at 340 ± 10 °C. The volatile products were analyzed by gas chromatography and had the following typical composition (product analysis of the pyrolysis of [trans- 3,6-D2]-1,4-cyclohexadiene):
1.2% cyclohexane, 6.2% cyclohexene, 5.2% 1,4-cyclohexadiene, 87.4% benzene. The benzene was isolated via preparative gas chromatography for isotope ratio measurement, and the mixture was analyzed using the push-rod method or GC/MS coupling, respectively:
| 1,4-Cyclohexadiene | Method | Benzene [%] | ||
|---|---|---|---|---|
| [3-D1] | GC/MS | 32.5 ± 0.9 | 67.5 ± 0.9 | — |
| [cis-3,6-D2] | 16.8 ± 0.2 | 10.7 ± 0.2 | 71.5 ± 0.3 | |
| [cis-3,6-D2] | GC/MS | 18.5 ± 0.5 | 9.8 ± 0.8 | 71.9 ± 0.9 |
| [trans-3,6-D2] | 5.7 ± 0.1 | 83.7 ± 0.5 | 10.6 ± 0.6 | |
| [trans-3,6-D2]1 | 4.1 ± 0.5 | 85.0 ± 0.5 | 10.9 ± 0.5 | |
| [cis,trans-3,6-D2]2 | 12.1 ± 0.8 | 49.7 ± 0.8 | ||
| [cis,trans-3,6-D2]3 | GC/MS | 15.9 ± 0.5 | 45.5 ± 0.4 | 38.6 ± 0.8 |
1: second synthesis batch
2: from Section 3.6.2.1
3: from Section 3.6.2.2
5. Hydrogen Transfer to Quinones
5.1. General Experimental Conditions and Product Analysis
The reactions were carried out in dioxane as the solvent, which was absolute-dried over LiAlH4 and distilled. The solvents acetonitrile and N-methylformamide used to investigate the stereochemistry (MERCK, > 99%) were dried over a 4 Å molecular sieve without further purification and used as such. The commercially available quinones (DDQ: MERCK 98%, OCA: FLUKA purum) were recrystallized from toluene and dried under vacuum to remove residual solvent. All experiments were conducted at 25.0 ± 0.2 °C. To test for byproducts, 1 mL of cyclohexadiene stock solution (0.200 M in 1,4-cyclohexadiene, 0.055 M in n-hexane) was mixed with 1 mL of quinone stock solution (0.050 M in DDQ or OCA) in screw-cap vials from MÜLLER & KREMPEL (40 × 12.75 mm) and maintained at 25 °C for 24 hours in a HAAKE NB 22 thermostatic bath. The reactions proceeded uniformly without the formation of byproducts. GC analysis of the reaction solutions yielded area ratios identical to those of the blank solutions with n-hexane as the internal standard, within the margin of error. Furthermore, more highly condensed compounds could not be detected by gas chromatography (2 m steel column, SE 30 4%, 280°C).
5.2. Determination of Total Isotope Effects
5.2.1. Kinetic Measurements
To determine the rate constants, the reaction solutions were prepared in the cuvettes as follows: 2 mL of dioxane was pipetted into the cuvettes, which were then sealed with a GC septum. Cyclohexadiene was injected using a gas-tight GC syringe, and the injected amount (12–25 mg) was determined by differential weighing. 500 ± 2 pl of temperature-controlled quinone stock solution was injected into the thermostated cuvette. The absorbance of the quinones was measured at specific time intervals at λ = 390 nm (DDQ) and λ = 425 nm (OCA). During the intervals between measurements, the cuvettes were kept out of the light path to prevent decomposition of the quinone and heating of the solution due to light exposure. The concentration of the quinone stock solution was chosen such that, at a conversion of approximately 80%, the measured absorbances fell within the range of 1.3 to 0.4. Depending on the reactivity of the system, cyclohexadiene was present in a 90- to 200-fold excess. Each system was measured four times; the statistical error was ± 2%. In contrast to cyclohexadiene and benzene, which exhibit negligible absorbance in the wavelength range of 300 to 500 nm, the absorbance of hydroquinone became a significant source of error at higher conversions. The measured absorbances were therefore corrected as follows:
\(E_{korr} = (E-[Q]_0·ε_{HQ})·ε_{Q}/(ε_{Q} - ε_{HQ})\)
\(E\) = measured absorbance
\([Q]_0\) = initial concentration of quinone
\(ε_{Q}\) = molar extinction coefficient of quinone
\(ε_{HQ}\) = molar extinction coefficient of hydroquinone
The following values were determined for the extinction coefficients [M-1·cm-1]:
| \(ε_{Q}\) | \(ε_{HQ}\) | |
|---|---|---|
| DDQ (390 nm) | 1720 | 80 |
| OCA (425 nm) | 1990 | 20 |
The pseudo-first-order straight lines, obtained by plotting \(E_{korr}\) against time \(t\) on a logarithmic scale, exhibited correlation coefficients of > 0.999 in all cases. From the pseudo-first-order rate constants and the respective concentrations of cyclohexadiene, the following second-order rate constants with respect to 1,4-cyclohexadiene were obtained [M-1·cm-1]·102:
| 1,4-Cyclohexadiene | DDQ | OCA | |
|---|---|---|---|
| [D0] | kHH | 1.32 ± 0.03 | 0.492 ± 0.009 |
| [3-D1] | kHD | 0.98 ± 0.02 | 0.365 ± 0.007 |
| [cis-3,6-D2] | kcis | 0.67 ± 0.01 | 0.250 ± 0.005 |
| [trans-3,6-D2] | ktrans | 0.67 ± 0.01 | 0.240 ± 0.005 |
| [3,3,6,6-D4] | kDD | 0.106 ± 0.002 | 0.0337 ± 0.0007 |
| [1,2-D2] | kβ | 1.38 ± 0.03 | 0.558 ± 0.009 |
The total kinetic isotope effects are:
| Total IE | DDQ | OCA |
|---|---|---|
| kHH/kHD | 1.35 ± 0.03 | 1.35 ± 0.04 |
| kHH/kcis | 1.97 ± 0.06 | 1.97 ± 0.07 |
| kHH/ktrans | 1.98 ± 0.07 | 2.00 ± 0.09 |
| kHH/kDD | 12.4 ± 0.5 | 14.6 ± 0.6 |
| kHH/kβ | 0.95 ± 0.03 | 0.88 ± 0.04 |
5.2.2. Intermolecular Competition
Threaded vials from MÜLLER & KREMPEL (40 × 12.75 mm) were used as reaction vessels for the competition experiments. The experiments were conducted in a HAAKE NB 22 thermostatic bath (+/- 0.2°C). The initial volumes were 2 mL of dioxane and 0.5 mL of quinone stock solution (5.77·10-2 M). Using a precision syringe (HAMILTON 1750), 0.500 ml of pre-tempered cyclohexadiene stock solution (approx. 21.8·10-2 M) was injected. The ratio of the cyclohexadienes used in competition with each other was determined from the initial weights of the stock solutions. The reaction solutions contained 4.4 × 10-2 M cyclohexadiene (sum of both isotopomers) and 1.1 × 10-2 M quinone. After 48 hours, the reaction solution was processed as follows: The solutions were drawn into a 10-mL disposable syringe filled with 5 mL of 10% Na-2SO-3 solution, extracted with 1 mL of n-pentane, and the pentane phase was washed twice with 5 mL of water. The conversion of cyclohexadiene was determined by gas chromatography. The ratio of the isomeric benzenes to each other was determined using GC/MS coupling. The isotope effects were calculated from the measured values using the equation by BIGELEISEN [65a], which accounts for the systematic error caused by higher conversion rates as well as arbitrary ratios of the competing reactants.
5.2.2.1. Competition Between 1,4-Cyclohexadiene and [3,3,6,6-D4]-1,4-Cyclohexadiene
The correction equation for the simple case of competition occurring exclusively at the intermolecular level (Melander and Saunders 1980) (page 95) is:
\(k_1/k_2 = log(1 - F_1)/log(1 - F_1·R_p/R_0)\)
\(F_1 = (1 + R_0)·X/(1 + R_p)\)
\(F_1\) = conversion of unlabeled 1,4-cyclohexadiene
\(k_1 = k_{HH}\)
\(k_2 = k_{DD}\)
\(X =\) sum of conversion of unlabeled and of [3,3,6,6-D4]-1,4-cyclohexadiene
\(R_0 =\) ratio of [3,3,6,6-D4]-1,4-cyclohexadiene to 1,4-cyclohexadiene
\(R_p =\) ratio of [1,4-D2]-benzene to benzene
Data and isotope effect for the dehydrogenation with DDQ:
\(R_0 = 1.98 ± 0.06\)
\(R_p = 0.23 ± 0.01\)
\(X = 27.1 ± 0.5%\)
\(F_1 = 65.7%\)
\(k_{HH}/k_{DD} = 13.5 ± 1.5\)
5.2.2.2. Competition Between 1,4-Cyclohexadiene and [3-D1]-1,4-Cyclohexadiene
In this case, a different correction equation must be applied, since intramolecular competition occurs in [3-D1]-1,4-cyclohexadiene in addition to intermolecular competition (Melander and Saunders 1980) (page 91, 102):
\(k_1/k_2 = log(1 - F·R_p/R_0)/log(1 - F)\)
\(F = (1 + R_p)·U/(1 + R_0)\)
\(R_p = (X·Y-1)/(X+1)\)
\(k_1 = k_{HH}\)
\(k_2 = k_{HD}\)
\(F\) = conversion of [3-D1]-]-1,4-cyclohexadiene
\(U\) = sum of conversion of unlabeled and of [3-D1]-1,4-cyclohexadiene
\(R_0 =\) ratio of [3-D1]-1,4-cyclohexadiene to 1,4-cyclohexadiene
\(X =\) ratio of [D1]-benzene to benzene (from intramolecular comp. Section 5.2.3)
\(Y =\) ratio of [D1]-benzene to benzene (from this experiment, intermolecular)
Data and isotope effect for the dehydrogenation with DDQ:
\(R_0 = 0.97 ± 0.05\)
\(R_p = 1.215\)
\(U = 29.0 ± 0.5%\)
\(X = 2.10 ± 0.06%\) (from Section 5.2.3)
\(Y = 2.27 ± 0.1%\)
\(F = 32.6%\)
\(k_{HH}/k_{HD} = 1.33 ± 0.09\)
5.2.3. Intramolecular Competition in [3-D1]-1,4-Cyclohexadiene
The experiment was conducted in the same manner as the method described for intermolecular competition. Here, quantitative dehydrogenating of 1,4-cycloheaxadiene (9.25·10-2 M solution of [3-D1]- 1,4-cyclohexadiene) was achieved with a 1.4-fold excess of quinone.
Mass spectrometric analysis was performed using both the GC/MS and the push-rod method. Both methods yielded identical values within the margin of error:
[D1]-benzene / benzene: 2.10 ± 0.06 (DDQ) and 2.10 ± 0.08 (OCA).
5.3. Intramolecular Competition in [cis-3,6-D2]- and [trans-3,6-D2]-1,4-Cyclohexadiene
The procedure was carried out in analogy to the method described in Section 5.2.3 and Section 5.2.2 by quantitative dehydrogenation with DDQ and OCA in the solvents dioxane, acetonitrile, and N-methylformamide. Mass spectrometric analysis was performed via GC/MS coupling. The values given have a mean standard deviation of ±1.5% absolute.
5.3.1. [cis-3,6-D2]-1,4-Cyclohexadiene
| Quinone / Solvent | Benzene [%] | ||
|---|---|---|---|
| D0 | D1 | D2 | |
| DDQ / Dioxane | 7.4 | 17.9 | 73.3 |
| DDQ / Acetonitrile | 7.2 | 20.1 | 72.7 |
| DDQ / N-Methylformamide | 8.5 | 28.9 | 62.3 |
| OCA / Dioxane | 7.9 | 10.3 | 81.8 |
| OCA / Acetonitrile | 8.0 | 8.9 | 83.2 |
| OCA / N-Methylformamide | 7.9 | 12.9 | 78.9 |
5.3.2. [trans-3,6-D2]-1,4-Cyclohexadiene
| Quinone / Solvent | Benzene [%] | ||
|---|---|---|---|
| D0 | D1 | D2 | |
| DDQ / Dioxane | 6.6 | 71.4 | 22.0 |
| DDQ / Acetonitrile | 5.6 | 59.7 | 34.5 |
| DDQ / N-Methylformamide | 5.8 | 55.8 | 38.4 |
| OCA / Dioxane | 5.4 | 88.0 | 6.8 |
| OCA / Acetonitrile | 5.5 | 88.2 | 6.3 |
| OCA / N-Methylformamide | 6.5 | 83.7 | 9.2 |
5.4. Determination of the Temperature Dependence of the Primary Isotope Effect
The determination was performed by measuring intermolecular competition between 1,4- cyclohexadiene and [3,3,6,6–D4]–1,4–cyclohexadiene using the method described in Section 5.2.2 over a temperature range of 15 to 60 °C. To prevent evaporation losses and interfering water vapor from the heating bath, the cyclohexadiene solution was injected through a GC septum that sealed the threaded neck of the vial. Before and after injection, the vials were additionally sealed with the standard screw caps. Mass spectrometric analysis was performed under low-voltage conditions via the push rod.
The GC/MS coupling proved unsuitable for intermolecular competition since the separation of benzene from excess cyclohexadiene on the capillary was poorly reproducible due to capillary overload. The benzene peak was often masked by noticeable tailing of the cyclohexadiene peak. When less substance was injected, the separation improved at the expense of measurement accuracy due to the lower ion current. However, since the temperature dependence of the isotope effect required the high-precision measurement the push-rod method was preferred. The ratio of the two isotopomers was:
\(R_0 = 2.04 ± 0.04\)
Formula for the calculation of \(p\): see Section 5.2.2.1
The primary isotope effect was calculated from \(k_{HH}/k_{DD}\) and the secondary α-isotope effect, determined in Section 5.2.1, which was 1.08 for both DDQ and OCA. The secondary isotope effects were assumed to be temperature-independent. The \(ln(p)/T^{-1}\) regression lines exhibited correlation coefficients R of 0.995 for DDQ and 0.997 for OCA, respectively.
[Annotation, HHB, 2026: In the original version a value of 1.10 for DDQ was given. However, the individual isotope effects were derived from approximations using equations for two different reaction mechanisms and found to be equal for DDQ and OCA, see Main Part, Section 6.5.1.]
5.4.1. Dehydrogenation with DDQ
| \(T\) [°C] | 15 | 30 | 45 | 60 |
|---|---|---|---|---|
| \(1/T\) [K-1∙103] | 3.470 | 3.299 | 3.143 | 3.002 |
| \(X\) [%] | 26.2 | 25.5 | 24.2 | 26.6 |
| \(R_p\) | 0.230 | 0.257 | 0.268 | 0.307 |
| \(F_1\) | 64.8 | 61.8 | 58.1 | 61.0 |
| \(k_{HH}/k_{DD}\) | 13.8 | 11.9 | 11.1 | 9.8 |
| \(p\)* | 12.5 | 10.8 | 10.0 | 8.9 |
| \(lnp\)* | 2.53 | 2.38 | 2.30 | 2.19 |
| \(p\)# | 12.8 | 11.0 | 10.3 | 9.1 |
| \(lnp\)# | 2.55 | 2.4 | 2.33 | 2.21 |
*: calculated with α = 1.1
#: calculated with α = 1.08 [see annotation in text]
5.4.2. Dehydrogenation with OCA
| \(T\) [°C] | 15 | 30 | 45 | 60 |
|---|---|---|---|---|
| \(1/T\) [K-1∙103] | 3.470 | 3.299 | 3.143 | 3.002 |
| \(X\) [%] | 26.4 | 22.7 | 27.2 | 27.6 |
| \(R_p\) | 0.195 | 0.219 | 0.260 | 0.291 |
| \(F_1\) | 67.3 | 56.7 | 65.7 | 65.1 |
| \(k_{HH}/k_{DD}\) | 16.8 | 13.4 | 12.3 | 10.8 |
| \(p\) | 15.6 | 12.4 | 11.4 | 10.0 |
| \(lnp\) | 2.75 | 2.52 | 2.43 | 2.30 |
6. Catalytic Disproportionation of 1,4-Cyclohexadiene
6.1. Disproportionation on Colloidal Nickel
6.1.1. General Experimental Conditions
6.1.1.1. Preparation of Colloidal Nickel
The preparation of the catalytically active colloidal nickel was carried out in accordance with the procedure described by M. SAKAI et al. (Sakai et al. 1985). Duranglass flasks (V = 25mL) were used as reaction vessels; these were placed inside a fused glass jacket and thermostated externally using an NB 22 thermostat from HAAKE. Up to four flasks could be connected in series without the temperature gradient exceeding one degree at 60°C. The reaction solutions were mixed using a magnetic stirrer and sealed gas-tight by closing the flasks with septa (butyl rubber, NS 14.5). 44 mg (0.2 mmol) of anhydrous nickel(II)bromide (JANSSEN, 99%) was suspended in 10 mL of dimethylformamide (FLUKA, > 99.5%, puriss., p.a.). An inert gas atmosphere was established by purging the suspension and the overlying gas space for one hour with a stream of nitrogen (approx. 10 mL/min) which was dried by bubbling subsequently through concentrated sulfuric acid and potassium hydroxide pellets. The catalyst solution was stirred overnight at room temperature under a slight nitrogen overpressure. The formation of colloidal nickel was evident from the solution’s increasing darkening. The rate of catalyst formation proved to be extremely dependent on the stirring speed: through optimization, the reduction of nickel bromide could be largely achieved within just one hour. If stirring was absent or only slightly too fast, the mixtures remained unchanged for days; even in these cases, the catalyst was reliably formed after subsequent correction of the stirring speed.
6.1.1.2. Disproportionation of 1,4-Cyclohexadiene on Colloidal Nickel
The catalyst solutions prepared as described in Section 6.1.1.1 were first maintained at 60 ± 1 °C for one hour before 0.4 mL (4.2 mmol) of nitrogen-saturated 1,4-cyclohexadiene was injected using a gas-tight precision syringe. For analysis, 0.15 mL of the reaction solution was withdrawn using a nitrogen-flushed 1 mL disposable syringe, quenched with 0.4 mL of 2 N hydrochloric acid, and the aqueous mixture was extracted with 0.15 mL of toluene. The toluene extract was washed with 0.4 mL each of 2 N hydrochloric acid and water and analyzed with gas chromatography.
6.1.1.3. Deactivation of the Catalyst
The colloidal state of the nickel is unstable under the reaction conditions; after prolonged (reaction) times, a clear solution was obtained above a black precipitate. This also occurred in the absence of 1,4-cyclohexadiene. The precipitated nickel exhibited low catalytic activity for the disproportionation reaction: the conversion was 11% after 48 hours.
6.1.2. Product Analysis
The disproportionation of 1,4-cyclohexadiene was carried out as described in Section 6.1.1.2, using o-xylene as an internal standard. The ratio of reactant to internal standard was 1.20. Samples were taken from the reaction solution at various times and analyzed by gas chromatography:
| Time | Composition | C6 Balance | ||||
|---|---|---|---|---|---|---|
| [h] | C6H12 | C6H10 | 1,3-C6H8 | 1,4-C6H8 | C6H6 | C6/C8# |
| 1 | — | 35.5 | — | 24.3 | 40.1 | 0.77 |
| 2 | 1.0 | 42.0 | — | 13.7 | 44.3 | 0.76 |
| 4.5 | 1.2 | 43.6 | — | 9.1 | 47.3 | 0.75 |
| 72 | 1.4 | 40.3 | — | 5.3 | 54.4 | 0.63 |
#: ratio C6-hydrocarbons to o-xylene
More highly condensed compounds could not be detected by gas chromatography (2.5 m glass column, SE 30 4%, 150°C/2 min, 10°C/min, 300°C/60 min).
6.1.3. Formation of Polymeric Byproducts
The precipitation of the catalyst, the depletion of C6 hydrocarbons during long reaction times, and the simultaneous increase in the ratio of benzene to cyclohexene may be attributable to the formation of cyclohexene polymers. Therefore, a reaction solution (0.02 M NiBr2, 0.1 M Zn, 0.4 M 1,4- cyclohexadiene) and a blank sample (no NiBr2, 0.1 M Zn, 0.4 M 1,4-cyclohexadiene) were prepared and heated for 120 hours at 60°C. GC analysis of both solutions subsequently revealed the following compositions (%):
| Sample | C6H12 | C6H10 | 1,3-C6H8 | 1,4-C6H8 | C6H6 |
|---|---|---|---|---|---|
| Reaction | 0.7 | 42.0 | — | 0.3 | 57.1 |
| Blind test | — | 1.0 | — | 97.9 | 1.1 |
Both solutions were then prepared for exclusion chromatography (gel permeation chromatography, GPC) by mixing each with 15 mL of concentrated hydrochloric acid, extraction with 70 mL of toluene, the organic phases being washed with water, dried over sodium sulfate, and concentrated to dryness. GPC separation of the residues (approx. 3 mg) was performed on a WATERS M45 HPLC system using a combination of SHODEX A801 and A802 columns in tetrahydrofuran at a flow rate of 0.5 mL/min. The exclusion limit was set at a molecular weight of 5,000 g/mol. Detection was performed using a WATERS Lambda Max Model 481 (λmax = 250 nm) and a KNAUER differential refractometer. Quantitative analysis was carried out using a SHIMADSU C-R3A integrator. The residues of both solutions exhibited identical chromatograms in terms of retention times and relative area ratios. In particular, no additional peak was detected in the reaction sample. The catalyst is therefore precipitated with polymers extracted from the septum. For a discussion of the hydrocarbon balances, see Section 8.1.2 of the main part.
6.1.4. Effect of Nickel Bromide Concentration
Reaction solutions were prepared with the same concentration of 1,4-cyclohexadiene (0.4 M) and increasing amounts of catalyst components, as described in Section 6.1.1. The molar ratio of zinc to nickel bromide was 5.0 ± 0.1 in all cases. After two hours, the conversion was determined by gas chromatography. For each measurement point, a triple determination was performed. The values shown are those from the most active reaction mixture in each case. The deviations from the other two reaction mixtures were < 15% (relative) in all cases.
| [NiBr2] M | C6H12 | C6H10 | 1,4-C6H8 | C6H6 | Conversion [%] |
|---|---|---|---|---|---|
| 2.8∙10-3 | — | 6.2 | 87.3 | 6.5 | 12.7 |
| 5.3∙10-3 | — | 12.6 | 74.8 | 12.6 | 25.2 |
| 8.7∙10-3 | — | 20.3 | 59.2 | 20.6 | 40.8 |
| 1.1∙10-2 | — | 27.7 | 44.5 | 27.9 | 54.5 |
| 1.3∙10-2 | — | 31.5 | 36.5 | 31.9 | 63.5 |
| 2.0∙10-2 | 0.06 | 33.4 | 32.1 | 34.4 | 67.9 |
| 2.6∙10-2 | 0.1 | 38.1 | 23.0 | 38.8 | 77.0 |
6.1.5. Reaction Order with Respect to 1,3- and 1,4-Cyclohexadiene
The reaction orders as a function of time were determined for 1,3- and 1,4-cyclohexadiene. The isomeric reactants were disproportionated under identical conditions as described in Section 6.1.1, and the change in concentration of the C6 hydrocarbons over time was monitored by gas chromatography. The reaction solutions contained 0.2 M cyclohexadiene, 0.01 M NiBr2, and 0.05 M zinc. For practical reasons, the disproportionation of the more reactive 1,3-cyclohexadiene was carried out at 25°C. The data obtained were evaluated for both systems according to first- and second-order rate laws for cyclohexadiene. The integrated rate equations are:
First-order: \(ln([\text{CHD}]_t/[\text{CHD}]_0) = k_{exp}·t\)
Second-order: \([\text{CHD}]_t^{⁻¹} = k_{exp}·t + [\text{CHD}]_0\)
Disproportionation of 1,3-Cyclohexadiene (25°C)
| t [min] | C6H12 | C6H10 | 1,3-C6H8 | C6H6 | \(ln \frac{[\text{CHD}]_t}{[\text{CHD}]_0}\) | \([\text{CHD}]_t^{⁻¹}\) |
|---|---|---|---|---|---|---|
| 10 | — | 10.7 | 78.5 | 10.9 | -0.242 | 6.15 |
| 20 | — | 17.1 | 65.7 | 17.2 | -0.420 | 7.35 |
| 30 | 0.09 | 22.1 | 55.1 | 22.7 | -0.596 | 8.77 |
| 40 | 0.1 | 26.3 | 47.4 | 26.4 | -0.747 | 10.19 |
| 50 | 0.2 | 28.0 | 39.8 | 32.0 | -0.921 | 12.14 |
| 70 | 0.3 | 34.4 | 28.0 | 37.4 | -1.273 | 17.25 |
| 120 | 0.3 | 43.3 | 9.5 | 46.9 | -2.354 | 50.86 |
Linear regression 1st order: R = -0.9998
Disproportionation of 1,4-Cyclohexadiene (60°C)
| t [min] | C6H12 | C6H10 | 1,4-C6H8 | C6H6 | \(ln \frac{[\text{CHD}]_t}{[\text{CHD}]_0}\) | \([\text{CHD}]_t^{⁻¹}\) |
|---|---|---|---|---|---|---|
| 30 | — | 22.2 | 54.3 | 23.5 | -0.611 | 8.90 |
| 60 | 0.2 | 29.4 | 37.6 | 32.8 | -0.978 | 12.85 |
| 90 | 0.22 | 34.5 | 28.0 | 37.3 | -1.273 | 17.25 |
| 120 | 0.23 | 36.4 | 22.5 | 40.9 | -1.492 | 21.47 |
| 150 | 0.24 | 38.8 | 18.6 | 42.4 | -1.682 | 25.97 |
| 180 | 0.25 | 39.9 | 15.8 | 43.9 | -1,845 | 30.58 |
| 270 | 0.26 | 42.0 | 11.4 | 46.3 | -2.172 | 42.38 |
Linear regression 2nd order: R = -0.9997
The rate constants are calculated as follows:
1,3-Cyclohexadiene (1st order): \(k = 2.8 ± 0.5 · 10^{-2}\) [M-1s-1] (25°C)
1,4-Cyclohexadiene (2nd order): \(k = 2.4 ± 0.5 · 10^{-1}\) [M-2s-1] (60°C)
6.1.6. Intramolecular Competition in Labeled 1,4-Cyclohexadienes
The disproportionation of the indexed 1,4-cyclohexadienes was carried out as described in Section 6.1.1 at 60°C. The reaction solutions contained 0.2 M CHD, 0.01 M NiBr2₂, and 0.05 M zinc. Samples of 1.0 mL were taken in each case; 0.15 mL of each sample was analyzed for hydrocarbon compositions by gas chromatography. The remainder was quenched with 2 mL of 2 N hydrochloric acid; the aqueous mixture was extracted with 0.5 mL of n-pentane, and the organic phase, after washing with 2 mL of 2 N hydrochloric acid and drying over sodium sulfate, was subjected to mass spectrometric analysis via GC/MS coupling. In addition to the molecular peak group of benzene (see Section 1.9.1), the isotope patterns of cyclohexene were recorded. The molecular peak group of unlabeled cyclohexene was measured as a reference spectrum:
[D0]-Cyclohexene isotope pattern of the cyclohexene molecular peak group
| m/e | 77 | 78 | 79 | 80 | 81 | 82 | 83 | 84 |
|---|---|---|---|---|---|---|---|---|
| % | 13 | 4 | 17 | 2 | 28 | 100 | 7 | — |
Since no reference spectra were available for the other isotopomers, only semi-quantitative conclusions can be drawn. Therefore, only the intensities normalized to the base peak are reported. Isotopomer distributionIsotope pattern of the cyclohexene molecular peak group in benzene.
[cis-3,6-D2]- and [trans-3,6-D2]-1,4-Cyclohexadiene
| Benzene [%] | C6H10 M-Peak Group (m/e), [%] | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| D0 | D1 | D2 | 80 | 81 | 82 | 83 | 84 | 85 | 86 | 87 |
| [cis-3,6-D2]-1,4-CHD / reaction time: 1.5 h / conversion: 43% | ||||||||||
| 28.8 | 9.1 | 62.1 | 10 | 14 | 13 | 38 | 100 | 56 | 15 | 1 |
| [cis-3,6-D2]-1,4-CHD / reaction time: 8 h / conversion: 77% | ||||||||||
| 28.6 | 9.2 | 62.3 | 10 | 13 | 13 | 39 | 100 | 67 | 20 | 2 |
| [trans-3,6-D2]-1,4-CHD / reaction time: 3.5 h / conversion: 64% | ||||||||||
| 3.3 | 84.1 | 12.6 | 11 | 16 | 14 | 36 | 100 | 81 | 24 | 3 |
| [trans-3,6-D2]-1,4-CHD / reaction time: 8.5 h / conversion: 76% | ||||||||||
| 3.4 | 84.0 | 12.6 | 11 | 15 | 13 | 34 | 100 | 83 | 26 | 4 |
[3-D1]-1,4-Cyclohexadiene
| Benzene [%] | C6H10 M-Peak Group (m/e), [%] | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| D0 | D1 | D2 | 78 | 79 | 80 | 81 | 82 | 83 | 84 | 85 |
| [3-D1]-1,4-CHD / reaction time: 8.5 h / conversion: 79% | ||||||||||
| 40.2 | 59.8 | — | 9 | 8 | 15 | 9 | 31 | 100 | 40 | 8 |
The error limits for the isotopomer distribution in benzene are < 1.5% absolute (see Section 1.9.1). Due to its sharpness, only 2–3 spectra could be recorded per GC peak for cyclohexene. The error limits for the intensities are therefore greater than those for benzene. For low-intensity peaks (up to approximately 20% of the baseline peak), they range from 1–5% absolute and increase to up to 10% absolute for the more intense peaks (>20%).
6.1.7. Tracer Study with [1,2-D2]- and [3,3,6,6-D4]-1,4-Cyclohexadiene
The procedure, sample collection and preparation, as well as the analysis, were carried out as described in Section 6.1.1. The molecular peak groups of both disproportionation products were measured and reported as intensities normalized to the base peak.
Isotope Patterns in Benzene
For comparison, samples from quantitative dehydrogenation with quinones (see Section 3.8 and Section 3.9, respectively) were recorded on the same measurement date.
| Benzene Isotopomers, M-Peak Group (m/e) [%] | |||||
|---|---|---|---|---|---|
| 77 | 78 | 79 | 80 | 81 | 82 |
| [1,2-D2]-1,4-CHD / reaction time: 8 h / conversion: 79% | |||||
| 2.3 ± 0.1 | 8.2 ± 0.5 | 17.6 ± 0.7 | 100 | 6.9 ± 0.4 | — |
| [1,2-D2]-1,4-CHD / from quinone dehydrogenation | |||||
| 2.2 ± 0.2 | 7.7 ± 0.5 | 16.9 ± 0.6 | 100 | 6.8 ± 0.4 | — |
| [3,3,6,6-D4]-1,4-CHD / reaction time: 5 h / conversion: 29% | |||||
| 2.0 ± 0.2 | 6.9 ± 0.6 | 16.0 ± 0.5 | 100 | 11.0 ± 0.6 | 1.9 ± 0.2 |
| [3,3,6,6-D4]-1,4-CHD / reaction time: 24 h / conversion: 32% | |||||
| 2.0 ± 0.2 | 7.1 ± 0.5 | 15.9 ± 0.6 | 100 | 10.8 ± 0.6 | 1.7 ± 0.2 |
| [3,3,6,6-D4]-1,4-CHD / from quinone dehydrogenation | |||||
| 2.0 ± 0.2 | 7.3 ± 0.5 | 15.5 ± 0.6 | 100 | 6.7 ± 0.2 | — |
Isotope Patterns in Cyclohexene
For comparison, unlabeled cyclohexene was also measured.
| Cycohexene Isotopomers, M-Peak Group (m/e) [%] | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| 77 | 78 | 79 | 80 | 81 | 82 | 83 | 84 | 85 | 86 | 87 |
| [1,2-D2]-1,4-CHD / reaction time: 8 h / conversion: 79% | ||||||||||
| — | — | 8 | 8 | 11 | 6 | 26 | 100 | 7 | — | — |
| [D0]-Cyclohexane | ||||||||||
| 13 | 4 | 17 | 3 | 27 | 100 | 7 | — | — | — | — |
| [3,3,6,6-D4]-1,4-CHD / reaction time: 5 h / conversion: 29% | ||||||||||
| — | — | — | 10 | 13 | 13 | 24 | 67 | 100 | 98 | 9 |
| [3,3,6,6-D4]-1,4-CHD / reaction time: 24 h / conversion: 32% | ||||||||||
| — | — | — | 11 | 15 | 15 | 33 | 78 | 100 | 86 | 7 |
6.1.8. Total Kinetic Isotope Effects in Intermolecular Competition
The intermolecular competition experiments were conducted under identical conditions as described in Section 6.1.1. Only the molecular peak groups of benzene were measured. For a derivation of the formulas and the meaning of the symbols and abbreviations, see Section 5.2.2.
1,4-Cyclohexadiene and [3,3,6,6-D2]-1,4-Cyclohexadiene
\(R_0 = [\text{3,3,6,6-D4]-1,4-CHD} / \text{1,4-CHD} = 1.66 ± 0.06\)
\(X = ([\text{CHD}]_0 - [\text{CHD}]_t)/[\text{CHD}]_0 · 100 = 24.7 ± 0.5\)
Reaction time: 65 min
| Composition | C6H12 | C6H10 | C6H8 | C6H6 |
|---|---|---|---|---|
| [%] | 0.25 | 12.2 | 75.3 | 12.3 |
\(R_p = [\text{1,4-D2]-benzene} / \text{[D0]-benzene} = 0.75 ± 0.03\)
\(F1 = \text{Conversion of 1,4-CHD} = 37.6%\)
\(k_{HH}/k_{DD} = 2.53 ± 0.1\)
1,4-Cyclohexadiene and [3-D1]-1,4-Cyclohexadiene
\(R_0 = [\text{3-D1]-1,4-CHD} / \text{1,4-CHD} = 1.2 ± 0.06\)
\(U = ([\text{CHD}]_0 - [\text{CHD}]_t)/[\text{CHD}]_0 · 100 = 37.8 ± 0.5\)
Reaction time: 65 min
| Composition | C6H12 | C6H10 | C6H8 | C6H6 |
|---|---|---|---|---|
| [%] | — | 18.5 | 62.2 | 19.3 |
\(X = \text{[1-D1]-benzene/[D0]-benzene} = 1.49 ± 0.05\)# \(Y= \text{[D0]-benzene/[1-D1]-benzene} = 2.17 ± 0.05\)##
\(R_p = 0.897\)
\(F = \text{Conversion of [3-D1]-1,4-CHD} = 32.6%\)
\(k_{HH}/k_{HD} = 0.71 ± 0.1\)
#: Measured value from intramolecular competition in 1b (Section 6.1.6)
##: Measured value from intermolecular competition between 1 and 1b (this section)
6.2. Disproportionation on the Ziegler Catalyst
6.2.1. General Experimental Conditions
6.2.1.1. Preparation of the Catalyst
The active catalyst species was prepared in situ in the reaction vessels described in Section 6.1.1 at 25°C. 10 mL of a solution containing 29.5 mg (0.11 mmol) of anhydrous nickel(II)acetylacetonate (MERCK, 98%) in 50 mL of absolute toluene, or 59.0 mg (0.22 mmol) in 50 mL of dioxane freshly distilled over LiAlH4. After purging the reaction vessels with dried nitrogen for one hour, 1 ml of a stock solution of 260 mg (2.3 mmol) of triethylaluminum (SCHERING, technical grade) in 20 ml of absolute toluene, or 500 mg (4.4 mmol) in 20 ml of absolute dioxane. The stock solution was prepared by injecting the triethylaluminum using a disposable syringe under strict exclusion of oxygen into a flask purged with nitrogen and containing the solvent. The triethylaluminum content was determined by differential weighing. Both stock solutions were freshly prepared for each batch. Catalyst formation (Al/Ni = 5; in dioxane: 4.2·10-3 M Ni; in toluene: 2.1·10-3 M Ni) began immediately after the injection of the aluminum alkyl, resulting in a deep black, homogeneous solution. This process occurred more rapidly in toluene than in dioxane. One hour after the addition of the reducing agent, further changes in the solution were observed in this solvent neither.
6.2.1.2. Disproportionation of 1,4-Cyclohexadiene
The addition of 0.1 mL (1.06 mmol; if in toluene) or 0.2 mL (2.12 mmol; if in dioxane) of 1,4-cyclohexadiene to the two-hour-old catalyst solutions was carried out as described in Section 6.1.1.2. The reaction temperature was 25°C in all cases. The cyclohexadiene-to-nickel ratio was 48 ± 2; the solutions were 0.1 M (toluene) and 0.2 M (dioxane) in cyclohexadiene, unless otherwise specified.
6.2.2. Product Analysis
Product analysis was performed as described in Section 6.1.2, using o-xylene as an internal standard in toluene. Gas chromatographic product analysis could not be performed in dioxane because dioxane exhibited the same retention time as o-xylene on the available columns and could never be completely eluted. 0.5 mL of a stock solution of 1,4-cyclohexadiene (2.0 M) and o-xylene (1.4 M) was injected. Samples were taken from the reaction mixture at various times and analyzed by gas chromatography. The table below shows a result representative of all experiments.
| Time | Composition | C6 Balance | ||||
|---|---|---|---|---|---|---|
| [min] | C6H12 | C6H10 | 1,3-C6H8 | 1,4-C6H8 | C6H6 | C6/C8# |
| 10 | 0.3 | 8.0 | — | 83.7 | 8.0 | 0.99 |
| 20 | 0.6 | 22.5 | — | 53.3 | 23.6 | 0.96 |
| 30 | 0.8 | 37.0 | — | 23.7 | 38.5 | 0.96 |
| 40 | 1.2 | 46.2 | — | 3.8 | 48.8 | 0.99 |
| 50 | 1.6 | 47.7 | — | — | 50.7 | 0.99 |
| 3•103 | 2.5 | 45.7 | — | — | 51.8 | 0.95 |
#: ratio C6-hydrocarbons to o-xylene
Apart from C6 and C8 hydrocarbons, as well as the C2 and C4 hydrocarbons already formed during catalyst preparation (approx. 1.5%), no other compounds were detected by gas chromatography. The catalyst solution is unstable under the reaction conditions: after disproportionation, a dark brown precipitate settled out, forming a clear supernatant. The time of phase separation proved to be non-reproducible; however, in some cases it occurred as early as the end of the reaction. The precipitated sediment exhibited lower activity, though still high compared to the precipitated colloidal nickel (see Section 6.1.1.3). After 60 minutes, the conversion of 1,4-cyclohexadiene was 40 percent [Annotation, HHB, 2026: contradicts the data in the table; that conversion percent may apply to 1,3-cyclohexadiene].
6.2.3. Kinetics of the Disproportionation of Isomeric Cyclohexadienes
The kinetic studies of the disproportionation of both cyclohexadienes were conducted in toluene and dioxane at 25°C under identical conditions according to the procedure described in Section 6.1.2. The following tables show the results of product analyses over reaction time.
1,4-Cyclohexadiene (Toluene, [CHD]0 = 0.09 M)
| t [min] | C6H12 | C6H10 | 1,3-C6H8 | 1,4-C6H8 | C6H6 | \(ln \frac{[\text{CHD}]_t}{[\text{CHD}]_0}\) |
|---|---|---|---|---|---|---|
| 10 | — | 7.5 | — | 84.5 | 8.0 | -0.169 |
| 20 | 0.3 | 22.1 | — | 55.2 | 22.4 | -0.595 |
| 30 | 0.8 | 36.1 | — | 25.5 | 37.6 | -1.367 |
| 40 | 1.2 | 44.9 | — | 6.6 | 47.4 | -2.718 |
| 50 | 1.7 | 47.8 | — | — | 50.5 | — |
1,4-Cyclohexadiene (Dioxane, [CHD]0 = 0.2 M)
| t [min] | C6H12 | C6H10 | 1,3-C6H8 | 1,4-C6H8 | C6H6 | \(ln \frac{[\text{CHD}]_t}{[\text{CHD}]_0}\) |
|---|---|---|---|---|---|---|
| 5 | — | 5.0 | — | 90.0 | 5.0 | -0.105 |
| 10 | 0.4 | 18.2 | — | 62.5 | 18.9 | -0.470 |
| 15 | 0.7 | 31.5 | — | 35.4 | 32.4 | -1.039 |
| 20 | 1.1 | 41.2 | — | 14.5 | 43.2 | -1.931 |
| 30 | 1.7 | 45.6 | — | 4.8 | 48.2 | -3.037 |
Linear regression 2nd order: R = -0.9997
1,3-Cyclohexadiene (Toluene, [CHD]0 = 0.08 M)
| t [min] | C6H10 | 1,3-C6H8 | 1,4-C6H8 | C6H6 | \(ln \frac{[\text{CHD}]_t}{[\text{CHD}]_0}\) | C6C/8 | conversion [%] |
|---|---|---|---|---|---|---|---|
| 10 | 13.4 | 73.3 | — | 13.4 | -0.311 | 0.88 | 26.7 |
| 20 | 20.6 | 58.7 | — | 20.7 | -0.531 | 0.92 | 41.3 |
| 30 | 25.8 | 47.7 | 0.5 | 26.0 | -0.740 | 0.91 | 52.3 |
| 40 | 26.4 | 45.8 | 0.7 | 27.8 | -0.781 | 0.93 | 54.2 |
| 50 | 31.0 | 36.0 | 0.9 | 32.1 | -1.022 | 0.92 | 64.0 |
| 60 | 34.9 | 27.8 | 1.0 | 36.3 | -1.280 | 0.93 | 72.2 |
| 70 | 37.6 | 21.7 | 1.1 | 39.7 | -1.528 | 0.93 | 78.3 |
Linear regression 1st order: R = -0.995
1,3-Cyclohexadiene (Dioxane, [CHD]0 = 0.2 M)
| t [min] | C6H10 | 1,3-C6H8 | 1,4-C6H8 | C6H6 | \(ln \frac{[\text{CHD}]_t}{[\text{CHD}]_0}\) | C6C/8 | conversion [%] |
|---|---|---|---|---|---|---|---|
| 20 | 4.0 | 92.1 | — | 4.0 | -0.082 | n.d. | 7.9 |
| 40 | 6.7 | 86.4 | — | 6.8 | -0.151 | n.d. | 13.6 |
| 60 | 9.9 | 79.8 | — | 10.2 | -0.236 | n.d. | 20.2 |
| 90 | 15.1 | 68.8 | — | 16.1 | -0.393 | n.d. | 31.2 |
| 120 | 20.9 | 56.7 | — | 22.4 | -0.580 | n.d. | 43.3 |
| 180 | 32.3 | 34.4 | 0.2 | 33.1 | -1.079 | n.d. | 65.6 |
| 240 | 41.2 | 15.7 | 0.4 | 42.8 | -1.900 | n.d. | 84.3 |
Linear regression 2nd order: R = -0.999
Competition Between 1,3- and 1,4-Cyclohexadienes (Dioxane, [1,3-CHD]0 = [1,4-CHD]0 = 0.1 M)
| t [min] | C6H12 | C6H10 | 1,3-C6H8 | 1,4-C6H8 | C6H6 |
|---|---|---|---|---|---|
| 30 | — | 3.6 | 43.7 | 49.1 | 3.6 |
| 90 | — | 12.4 | 29.5 | 45.3 | 12.9 |
| 150 | — | 25.2 | 9.8 | 38.3 | 26.6 |
| 194 | 0.5 | 37.5 | 0.2 | 22.0 | 39.8 |
| 205 | 1.2 | 43.6 | — | 9.0 | 46.2 |
Disproportionation in the Presence of Cyclohexene (Dioxane, [CHD]0 = 0.2, [C6H10]0 = 0.09 M
| t [min] | C6H12 | C6H10 | 1,4-C6H8 | C6H6 |
|---|---|---|---|---|
| 2.5 | — | 35.1 | 60.1 | 4.22 |
| 5.0 | 0.2 | 38.1 | 50.6 | 10.9 |
| 7.5 | 0.5 | 44.5 | 38.1 | 16.9 |
| 10.0 | 0.6 | 49.7 | 27.3 | 22.2 |
| 15.0 | 1.0 | 57.9 | 8.9 | 32.1 |
| 20.0 | 1.5 | 60.2 | 0.5 | 37.8 |
Disproportionation in the Presence of Benzene (Dioxane, [CHD]0 = 0.2, [C6H6]0 = 0.1 M
| t [min] | C6H12 | C6H10 | 1,4-C6H8 | C6H6 |
|---|---|---|---|---|
| 2.5 | — | 0.9 | 66.0 | 33.1 |
| 5.0 | — | 1.5 | 65.7 | 32.8 |
| 7.5 | 0.1 | 2.7 | 57.5 | 39.7 |
| 10.0 | 0.2 | 5.7 | 58.1 | 36.0 |
| 15.0 | 0.7 | 13.7 | 43.9 | 41.5 |
| 30.0 | 0.8 | 25.4 | 9.0 | 64.8 |
Acknowledgments
1987
My special thanks go to my esteemed advisor Prof. Dr. Albert Heesing (🕆 2015) for assigning this topic, as well as for his assistance in scientific questions, countless suggestions and – most important – critical analysis of my inferences throughout the course of this work.
Special thanks are also going to:
Dr. Heinrich Luftmann for his assistance and dedication in the development of mass spectrometric analytical methods and in the gas chromatographic separation of isotopomers.
Ms. Birgit Wibbeling for her collaboration on numerous experiments and her constant willingness to help.
Dr. H. Schnöckel, Priv.-Doz., from the Institute of Inorganic Chemistry at the University of Münster for recording the Raman spectra.
Dipl.-Ing. G. Bandmann of the Institute of Organic Chemistry at the University of Essen for recording the deuterium-decoupled 13C satellite 1H-NMR spectra.
The members of the working group for their constant willingness to engage in scientific discussion.
Finally, I would like to thank all the staff and employees of the institute who contributed to the success of this work.
2026
This refurbishment was achieved with reasonable effort thanks to the efficiency of artificial intelligence, in particular for:
- OCR in the photographed original version.
- Translation into English.
- LLM support for coding.