I CONTENTS ACKNOWLEDGEMENTS VII SUMMARY VIII CHAPTER 1 INTRODUCTION 1.1GENERAL 11.2 USES OF COUPLED PHENOLIC S 4 1.2.1 Antioxidants 41. 2. 2 Other Uses 61.3 METHODS OF PREPARATION OF COUPLED PHENOLICS 6 1.3.1 General Types of Coupling Reaction Mechanisms 81.3.2 Chemical and Electrochemical Methods for Oxidatively Coupling Phenolics 151.3.2.1 Chemical oxidative coupling 151.3.2.1.1Vanadium(IV) and vanadium(V) 161.3.2.1.2 A (nitrosonaphtholato)metal complex 181.3.2.1.3 Activated manganese dioxide 21 1.3.2.1.4 Cupric salts 23 1.3.2.2 Electrochemical oxidative coupling 251.3.2. 2.1 Direct electrochemical oxidations 25 1.3.2. 2.2 Indirect electrochemi cal oxidations 291.4 OBJECTIVES AND MOTIVATION FOR THIS STUDY 32
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
The oxidative coupling of 2,6-di-t -butylphenol under mild reaction conditions is well
documented and the subject of many patents. However, the coupling of other mono-
and di- substituted phenols is not as well documented and thus there is scope for
further investigation for providing a convenient, environmentally friendly and
economically viable method for the oxidative coupling of these phenols.
In this study, the oxidative coupling of a variety of alkylated phenolic substrates, 2-t -
butylphenol, 2,6-di-t -butylphenol, 2,4-di-t -butylphenol and 2,4-dimethylphenol, using arange of different oxidizing agents, were investigated by means of experimental
and/or theoretical means. The dibutylated aromatics provided the highest selectivities
to their respective coupled products, with results obtained with the dimethyl analogue
being only satisfactory, and that for 2-t -butylphenol being totally inefficient.
PM3 Molecular orbital (MO) calculations were used to predict the possible modes of
coupling for the substrates 2,6-di-t -butylphenol and 2,4-di-t -butylphenol, and these
results were then compared with those obtained experimentally in the laboratory.
Preliminarily, the coupling of unsubstituted phenolics was also assessed by means of
MO calculations.
Much emphasis was placed on Ce(IV) as the oxidant, and the reaction conditions
under which it was used and the results that were obtained have not been reported
before and are therefore novel. The oxidation of 2,4-di-t -butylphenol using Ce(IV) in
the presence of methanesulphonic acid was optimized to afford high yields andselectivities to the desired ortho C-ortho C coupled product under mild reaction
conditions. Various reaction parameters were also investigated in this case, such as
varying the MeSO3H concentration, the solvent, the reaction temperature, the reaction
time, the substrate loading, the rate of oxidant addition and the substrate to oxidant
ratio. Ce(IV) also gave a high selectivity to the para C-para C coupled product when
using 2,6-di-t -butylphenol as the substrate. However, it was not as effective with 2,4-
dimethylphenol, and even less so with 2- t -butylphenol.
The oxidation reactions of 2-t -butylphenol and 2,4-dimethylphenol with various
coupling agents were also investigated with the intention of obtaining high selectivities
to the respective desired coupled products. In these studies, 2-t -butylphenol afforded
a large number of products, irrespective of the oxidant used. The dimethyl analogue
was more selective, but results were not optimal. It was clear that the number of
substituents on the phenol ring, their nature and their position with regards to the
hydroxyl moiety was of great importance and made a significant impact on thepreferred coupling mode of the substrate. It was observed that steric effects also
played a major role in the outcome of these reactions: 2,6-di-t -butylphenol never
afforded any C-O coupled products whereas 2-t -butylphenol, 2,4-di-t -butylphenol and
2,4-dimethylphenol all appeared to undergo some C-O coupling.
Finally, reaction mechanisms were provided for both the K3Fe(CN)6 and Ce(IV) work,
these reacting in basic and acidic media, respectively. It was proposed that both of
these mechanisms operate through the initial formation of the phenoxyl radical.
dihydroxybiphenyls, the focus of this investigation. All of these products have
considerable economic importance because they are used to manufacture
thermosets, insulating foams, adhesives, laminates, impregnating resins, and serveas raw materials for varnishes, herbicides and insecticides.
1.2 USES OF COUPLED PHENOLICS
1.2.1 Antioxidants
One of the more important uses of many phenolic materials is their ability to serve asantioxidants. Antioxidants are merely compounds that are added to, or occur in,
various materials, both living organisms and synthetic organic materials –
antioxidants then readily react with free radicals that would otherwise damage the
materials prematurely. The free radicals are normally the result of autoxidation, a
process that occurs spontaneously all around us all the time due to the oxygen in the
air.
In human blood plasma, α-tocopherol, well known as a component of vitamin E, has
proved to be the most efficient phenol derivative to date to trap damaging phenoxyl
radicals (ROO•),8,9 caused by autoxidation, and therefore acts as an efficient
antioxidant. Uninhibited free radical peroxidation in vivo is implicated in a wide variety
of degenerate diseases such as cancer, heart disease, inflammation and even aging.
Thus, over the last two decades, there has been a tremendous increase in the
research of phenols as antioxidants.10,11
Phenols owe their efficient antioxidant activity to their ability to scavenge radicals by
hydrogen or electron transfer in processes that are much faster than radical attacks
on the substrate. The antioxidant property can be related to the readily abstractable
phenolic hydrogen as a consequence of the relatively low bond dissociation enthalpy
of the phenolic O-H group. Thus phenols and dihydroxybiphenyls are an extremely
Dihydroxybiphenyls are used in toner resins to increase surface additive adhesionand to optimize cohesion between the toner particles.14 It also acts as a binder resin,
thus eliminating the need for gels to be present in the toner, and enabling the
magnetic brush development system to achieve consistent, high quality copy
images.15
They are also used as inexpensive and simple starting materials for producing
polycarbonate resins,
16
which are used to reinforce rubber vulcanizates.
17
Dihydroxybiphenyls are extensively used in coating agents,18 glass moulding19 and
infrared-reflecting colourants,20 and they are reacted with acid catalysts to form
polymers which are used as a polymer scale deposition preventative agent.21
1.3 METHODS OF PREPARATION OF COUPLED PHENOLICS
The diversity of phenol oxidation products offers interesting synthetic possibilities for
the preparation of simple and polymeric molecules containing phenolic and/or quinoid
structural elements; these can be formed from both like and unlike radical
species.13,22 The successful synthesis of various natural products from phenols has
been well documented from the 1950’s to the present.23-28
Biogenetic oxidative coupling routes were first investigated in 1957,29,30 and the
prevalence of the overall coupling process in the biosynthesis of natural products was
authenticated. Thus the oxidative coupling step has been found to be extremelyimportant in the natural formation of compounds as diverse as lignins,31 lignans,32
tannins,33 plant pigments,22 and an estimated 10% of all known alkaloids.23 (Lignin is
a complex biopolymer that accounts for 20-30% of the dry weight of wood. It is
formed by the free radical polymerization of substituted phenylpropane units to yield
polymers which have a number of functional groups such as aryl ethers, phenols and
The major difficulty with oxidative coupling reactions of phenols is that a large variety
of potential products are possible from a single substrate when carried out in the
presence of various chemical or biological oxidants. This is because the phenolicmolecules are able to undergo both carbon-carbon (Scheme 4 shows para-para
coupling, though ortho-para coupling may also occur) and carbon-oxygen (Scheme 5)
To understand the effect that both the nature of the reactant and oxidant has on the
type of products that are formed, one must have an understanding of the various
reaction pathways that are possible, from a mechanistic point of view. A summary of literature reports dealing with the various mechanisms is now briefly discussed.
1.3.1 General Types of Coupling Reaction Mechanisms
The reaction pathway for the oxidative coupling of phenols has been extensively
investigated.38,39 There are two main modes of coupling that may be highlighted.
These are an external and an internal oxidation process. In the former, electrons aretransferred from the phenolic compound to an external oxidizing agent, whilst the
internal oxidation process involves an internal oxidation-reduction reaction in which
one substrate molecule is oxidized whilst another is simultaneously reduced. Since
there is no change in the net overall oxidation state, this process may be termed a
“non-oxidative coupling (NOC)” reaction.
In our investigations, only the external oxidative coupling process was studied. For
this reason, literature reports dealing only with this mode are summarized here.
External oxidative coupling reactions may be grouped into two separate classes,
those involving free radical intermediates, and those that are non-radical in nature.
These may further be subdivided into several general mechanistic types.
a) Mechanisms involving free radical intermediates
i) Direct coupling of two phenoxyl radicals (FR1)
ii) Homolytic aromatic substitution (FR2)iii) Heterolytic coupling preceded by two successive one-electron oxidation
steps (FR3)
b) Mechanisms which are non-radical in character
i) Heterolytic coupling preceded by a single two-electron transfer (NR1)
ii) Concerted coupling and electron transfer (NR2)
It has previously been widely accepted that, in the field of phenol oxidations, the FR1
mechanism is the most viable (without discounting the FR2 mechanism). Most
reviewers have included the FR3 mechanism in their discussions but have attachedlittle importance to it. Until recently, no one has considered the NR1 and NR2
mechanisms as significant enough to warrant a discussion of them in this context.
The para-para (C-C) coupling of a simple 2,6-disubstituted phenol is used to illustrate
the five general types of processes (FR1, FR2, FR3, NR1 and NR2) as listed above.
In all cases, the oxidized phenolic species is written as the neutral phenol molecule,
and only intermediates are shown as unprotonated. The following scheme (Scheme6) highlights the FR1, FR2 and FR3 mechanisms.
in free radical aromatic substitutions,45 have not yet been observed in oxidative
coupling reactions. This may perhaps be due to the fact that the conversion of (4)
to (3) is a facile one since (3) has enhanced stability due to its aromaticity.
• Thirdly, the phenoxyl radical may be further oxidized by removal of an electron, to
yield a phenoxyl cation, according to mechanism FR3. The initial substrate (1),
with concomitant hydroxyl proton loss, may then heterolytically couple with the
cation to afford (2).
Examples of the NR1 and NR2 non-radical processes are shown in Schemes (8) and
(9), respectively. In both illustrations, the oxidant is represented as a tripositive metalion (M3+), which forms an initial metal-phenolate complex with (1).
were not investigated successfully because they were not well understood; it was
always assumed that coupling occurred through the bonding of two phenoxyl radicals
(FR1) to form the coupled biphenol. However, it has since become clear that thetypes of mechanisms involved are extremely dependent on the nature of the oxidant
and/or catalyst used. Some of these, including vanadium (IV) and (V), a
(nitrosonaphtholato)metal complex, activated manganese dioxide, and cupric salts,
and the reaction pathways they are involved in, will now be discussed further.
1.3.2.1.1 Vanadium(IV) and vanadium(V)
Vanadium(V) oxytrichloride (VOCl3) and vanadium(IV) tetrachloride (VCl4) have beenused to oxidatively couple phenols in aprotic solvents.61 When phenol (5) was used
as the substrate in the presence of VCl4, a dark insoluble residue was initially formed
which was accompanied by the vigorous evolution of HCl gas. This residue was
shown to be a form of vanadium-phenolate species, but when analyzed, the
elemental composition was not consistent with any simple structure. Acid hydrolysis
thereof afforded high yields of the para-para coupled product, identified as 4,4’-
diphenol (6). Also observed were the para-ortho and ortho-ortho coupled products,
identified as 2,4’-diphenol (7) and 2,2’-diphenol (8), as shown in the following scheme
(Scheme 10).
Scheme 10: The oxidative coupling of phenol using VCl4 as oxidant
When the hydrolysis step was carried out in the presence of deuterium oxide, no
carbon-deuterium bonds were formed, indicating that the vanadium is bonded to the
phenolic oxygen. Furthermore, it was found that phenol (5) itself could not becoupled oxidatively using vanadium(V) oxytrichloride but ra ther only those substituted
phenolics, such as the naphthols, that have oxidation potentials lower than (5).
A simple mechanism involving the formation of a vanadium phenolate compound has
previously been proposed, but does not provide explanations for all observations
made. In this proposal, the vanadium-phenolate decomposes to form the phenoxyl
radical and a lower valence vanadium species, whereafter the coupling/dimerizationstep occurs to afford the biphenol. It has been suggested by Carrick61 that phenolic
coupling occurs by a rearrangement of electrons in a complex containing at least two
phenoxide residues and one metal center. Whether vanadium(V) or vanadium(IV)
acts as one or two electron oxidizing agents here is not clear and, furthermore, the
course of the phenolic coupling itself is also not clear. However, the existence of
metal-phenolate compounds has been established, enhancing the possibility that a
non-radical (two electron oxidation) pathway may be involved. The NR2 mechanism
can be used to explain the existence of a metal-phenolate derivative (Scheme 11).
Complex (11) has manganese in the 3+ oxidation state since this metal ion was found
to be electrochemically stable. It was suggested that complex (11) then abstracts a
hydroxyl hydrogen from (9) to yield the peroxymanganese (12) and the phenoxylradical (13). Complex (12) immediately decomposes to afford phosphine oxide and a
hydroxyl radical. Radical (13) then reacts with (9) to yield the coupled product (14)
which tautomerizes to (15). Thereafter, after a similar oxidation cycle, radical (15)
affords the diphenyl diol (16), which is oxidized by the same catalytic cycle to give
(17). The latter compound is ultimately transformed to the diphenoquinone (10).
1.3.2.1.3 Activated manganese dioxide
37
When activated manganese dioxide was reacted with 2,6-xylenol (18), the analysis of
the product mixture showed the presence of a polyphenylene ether (19), 3,3’,5,5’-
tetramethyl-p ,p ’-biphenol (20) and 3,3’,5,5’-tetramethyldiphenoquinone (21) [Scheme
14].
Scheme 14: Oxidation of 2,6-xylenol using MnO2
The molecular weight of polymer (19) varied substantially, depending on the reactant
ratio and the reaction solvent used, ranging from 2 000 to 20 000. The polymer was
the major product, with diol (20) and diphenoquinone (21) being formed in much
smaller amounts, when a molar ratio of 3:1 (oxide:xylenol) was used. However, when
(18) was used in molar excess, (21) was the principle product, with a low molecular weight oligomer also being formed. Products (20) and (21) are formed by carbon-
carbon coupling, whilst (19) is formed exclusively by carbon-oxygen coupling
(Scheme 15).
Scheme 15: Reactions showing C-C and C-O coupling using MnO2 as oxidant
Scheme 16: Oxidation of 2,4-disubstituted phenols using cupric salts
In these reactions, the phenolic compound was in excess and also served as the
solvent. The cupric salt was regenerated by bubbling air through the solution
(Equation 2) and, as a result, could be used in catalytic amounts, with oxygen serving
as the principal oxidizing agent. In the above scheme, when R = H or CH3, it was
found that larger amounts of resinous materials were produced in the presence of
oxygen. Phenol itself gave polymeric products exclusively but, in the absence of oxygen, a light amber oil was produced which consisted mainly of the coupled dimer
(26) [Scheme 17]. Small amounts of other coupled products were also formed such
as the para-para (6) and ortho-para (7) coupled products (Scheme 17).
• Electrochemical methods can be used to synthesize a wide variety of organic
chemicals: any oxidation that can be carried out using conventional chemical
oxidizing agents can theoretically be carried out in an electrochemical cell.
• Electrochemical syntheses often have a lower environmental impact than
conventional oxidation methods since electrolytic routes often replace toxic
reagents and hazardous process conditions.68
Generally, the phenolic substrate forms an electrogenerated radical species, the
dimerization of which (to afford the desired product) is in competition with a further
one electron oxidation that results in the corresponding cation. In the case of phenolitself, electropolymerization is known to occur at the anode surface resulting in the
formation of a passivating film on the electrode surface.56,57 In addition to polymeric
products, both p -benzoquinone and 4,4’-diphenoquinone are also produced as minor
products (in 20 and 10% yields, respectively) as shown in the following scheme
phenolic substrates bearing bulkier alkyl substituents afford radicals that are expected
to have enhanced stability. For example, the radical species of 2,6-di-sec -
butylphenol was detected using multiple internal reflection Fourier transform infraredspectroscopy (MIRFTIRS), thus confirming the radical mechanism during the anodic
oxidation of this substrate.
When 2,6-di-tert -butylphenol (9) was reacted under constant current electrolysis
conditions (1.0 mA.cm-2; 2.5 F.mol-1) in MeOH-CH2Cl2, using a divided cell, it was
converted to the corresponding 4,4’-diphenoquinone (10) in 84.7 % yield. A
subsequent electroreduction, achieved by merely altering the current direction,resulted in the formation of biphenol (16) in 92.5 % yield (Scheme 19).53
Scheme 19: Direct electrochemical oxidation of 2,6-di-t -butylphenol
p -Cresol (29) was also electrolyzed at a controlled potential (+0.25 V vs SCE) in a
basic medium to afford Pummerer’s ketone (30) in 74 % yield, as seen in thefollowing scheme (Scheme 20).69
• The in-cell method: The reaction between the organic substrate and the redox
reagent, in its active oxidation state, occurs within the cell.
• The ex-cell method: The reaction is carried out in a reactor separate to the cell.
This approach has advantages over the in-cell method in that the conditions for the
electrode reaction and the chemical step may be optimized separately and,
furthermore, that the electrolyte may be purified between the reactor and the cell.
One redox catalyst that has been used successfully for oxidative coupling is the
Ce+3/Ce+4 couple.69 The cerium(IV) ions were generated from cerium(III) in the
presence of perchloric acid. Using 2,6-dimethylphenol as the substrate in aqueous or aqueous-acetonitrile solutions of perchloric acid (0.5 - 1.0 M) at room temperature,
the corresponding 4,4’-diphenoquinone and 1,4-benzoquinone were obtained as the
main products.70
Under similar conditions, the oxidation of 2,6-diisopropylphenol (31a), 2-tert -butyl-6-
methylphenol (31b), 2,6-diphenylphenol (31c) and 2,6-dichlorophenol (31d) resulted
in the formation of the corresponding 4,4’-diphenoquinones (32a-d) in addition to
oligomeric poly(1,4-phenylene) oxides (33a-d) [Scheme 21]. In the case of 2,6-
diphenylphenol (31c), the quantity of carbon-oxygen coupled product was low due to
the steric hindrance associated with the large phenyl groups adjacent to the oxygen
atom.
When 2,6-diisopropylphenol (31a) and 2- tert -butyl-6-methylphenol (31b) were
oxidized by cerium(IV) in a two phase system, namely, in CCl4 and an aqueous
acetonitrile solution of perchloric acid, at a high concentration of perchloric acid (4.0M) in the reacting phase, this afforded the corresponding 1,4-benzoquinones in good
yields. This was not observed with 2,6-diphenylphenol and 2,6-dichlorophenol.
The results obtained from these reactions are summarized in Table 1.2.
Dihydroxybiphenyls, as mentioned previously, have many important uses aschemicals in their own right, but also as intermediates in the manufacture of other
materials. Dihydroxybiphenyls are often prepared by means of oxidative coupling
procedures. However, the reaction is only efficient for disubstituted phenols such as
2,6-di-t -butyphenol and 2,4-dimethylphenol. The literature contains many reports on
the successful coupling of these substrate types, but is, however, virtually devoid of
studies carried out on monosubstituted phenols such as 2-t -butylphenol. The reasons
for this are clear: the C-C coupling of 2,4- or 2,6- disubstituted phenols is reallypossible only in the 6- and 4- positions, respectively, leading to reactions that afford
high yields of the desired coupled product. In contrast, a monosubstituted phenol
such as 2-t -butylphenol has two positions available through which C-C coupling may
occur, the 4- and 6- positions.
Hence, in the latter case, complex oxidative coupling reaction mixtures are obtained.
These often contain significant proportions of polymeric materials, and thus low
selectivities to the desired product are a result. This in turn implies tedious and time-
consuming purification steps. There thus appears to be a need to study these
reactions more closely with the view to developing a better understanding as to the
mechanisms at work so that the knowledge base for this reaction type may be
enhanced, and ultimately a better process may be devised.
In addition, it must be mentioned that another factor that has fuelled our interest in
this investigation is the ready availability of the starting materials that are to becoupled. SASOL produces phenol during its petroleum cracking process, and
alkylated phenols may readily be prepared from it. These alkylated phenols serve as
substrates in our coupling reactions.
This study is therefore concerned with the oxidative coupling of various mono- and di-
substituted phenols using chemical and indirect electrochemical oxidation methods.
2.2.1.1 Preparation of 3,3’-di-t -butyl-4,4’-dihydroxybiphenyl71
To a mixture of 4,4’-dihydroxybiphenyl (0.3701 g, 1.987 mmol), SiO2 (1.013 g), and
Na2CO3 (1.935 g, 18.25 mmol) in CCl4 (7 mL) was added t -butyl bromide (0.7867 g,
5.741 mmol), and the reaction mixture was stirred vigorously for 24 h at 70°C. The
SiO2 was filtered off and washed with ethyl acetate. The ethyl acetate washings andfiltrate were then combined and the solvent was removed under vacuum. The product
was isolated using thin-layer chromatography with hexane:ethyl acetate (90:10) as
the developing solvent system. The desired product, 3,3’-di-t -butyl-4,4’-
dihydroxybiphenyl, was thus obtained, and had m.p. 182-183°C (lit.71, m.p. 181-
2.2.1.2 Preparation of 3,3’,5,5’-tetra-t -butyldiphenoquinone
2,6-Di-t -butylphenol (0.222 g, 1.075 mmol) was added to silver oxide (0.5147 g, 2.222
mmol) in methanol (25 mL), after which the reaction mixture was stirred for 1 h. The
solids were removed by filtration and washed with hot toluene, the toluene then being
combined with the filtrate. This solution was then concentrated down on the rotary
evaporator to afford crude 3,3’,5,5’-tetra-t -butyldiphenoquinone (99.00 %) as theprimary product, which was further purified by recrystallization using ethyl
2.2.1.3 Preparation of 3,3’,5,5’-tetra-t -butyl-4,4’-dihydroxybiphenyl
To a suspension of 3,3’,5,5’-tetra-t -butyldiphenoquinone (1.169 g, 2.861 mmol) inether (50 mL) was added a solution of sodium hydrosulphite (8.030 g, 46.12 mmol) in
aqueous NaOH (1.0 M, 100 mL). After stirring the reaction mixture for 1 h, the
aqueous layer was acidified with concentrated HCl (15 mL). The organic layer was
separated, dried (MgSO4) and concentrated to give 3,3’,5,5’-tetra-t -butyl-4,4’-
dihydroxybiphenyl (99.00 %), which was further purified by recrystallization using
2.2.1.5 Preparation of 3,3’,5,5’-tetramethyl-2,2’-dihydroxybiphenyl
A Ce4+ solution (20 mL, 2.057 mmol) was added to 2,4-dimethylphenol (0.1366 g,1.118 mmol) in a 50 mL round-bottomed flask and stirred vigorously at 750 rpm for 1
h. The reaction mixture was then extracted using ethyl acetate (3 x 25 mL), and the
organic layer washed with water (3 x 25 mL) and dried (MgSO4). The solvent was
removed under vacuum, and the product was isolated using column chromatography,
with hexane:ethyl acetate (90:10) as the developing solvent system. The desired
product, 3,3’,5,5’-tetramethyl-2,2’-dihydroxybiphenyl, had m.p. 130-134°C (lit.4, m.p.
Celite was first purified by successively washing with methanol containing 10%
concentrated HCl, and distilled water, until neutral. It was then dried at 120°C for 12
h. This purified celite (30.00 g) was then added to a mechanically stirred solution of
silver nitrate (34.00 g, 200.1 mmol) in distilled water (200 mL). A solution of
Na2CO3·10H2O (30.00 g, 104.9 mmol) in distilled water (300 mL) was then added
slowly to the resulting homogeneous solution. When the addition was complete,
stirring was continued for a further 10 min. The yellow-green precipitate that formed
was filtered off and finally dried on a rotary evaporator over a period of several hours.Every 0.57 g of this silver carbonate/celite reagent contained 1.00 mmol of Ag2CO3.
Potassium hydroxide (5.675 g, 101.1 mmol) was thoroughly mixed with manganese
dioxide (4.345 g, 54.62 mmol) and left in an oven at 350°C for 3 h. The fused green
potassium manganate that so formed was filtered and then used for the preparation
of barium manganate.
Preparation of barium manganate
To a 500 mL flask containing distilled water (100 mL) was added barium hydroxide
(7.698 g, 24.40 mmol), and the pH was adjusted to 7 with dilute hydrochloric acid. To
the resulting warm solution was added potassium manganate (8.236 g, 41.78 mmol)
with stirring. The colour of the reaction mixture immediately changed to dark purple.
The reaction mixture was filtered with suction and the so-obtained dark blue crystals
were washed several times with distilled water, and placed in an oven at 100°C for 24
h to afford active barium manganate.
2.2.2.3 Preparation of a (nitrosonaphtholato)metal complex (Mn II(1-
nnap)2)27
Preparation of 1-nitroso-2-naphthol
After 2-naphthol (14.68 g, 101.8 mmol) was dissolved in hot NaOH (0.6 M, 340 mL),the solution was cooled to 0ºC. NaNO2 (7.054 g, 102.2 mmol) was added, and 6 M
H2SO4 (16 mL) was carefully dropped into the resulting solution during 1.5 h with
stirring. The mixture was stirred for a further 1 h. The brown solid that formed was
filtered, washed with water (250 mL) and dried in a desiccator. The crude material
was recrystallized from petroleum ether (b.p. 60-80ºC) to afford 1-nitroso-2-naphthol
as reddish brown needles; m.p. 107-109ºC (lit.75, m.p. 106-108ºC).
1-Nitroso-2-naphthol (4.015 g, 21.28 mmol) was dissolved in a 10 M NaOH solution(50 mL) at 0ºC during 2 h, and the mixture was stirred at room temperature overnight.
The green solid that formed was filtered, washed with 2 M NaOH solution, and dried
in a desiccator to afford the corresponding sodium salt (3.613 g, 17.11 mmol,
76.25%).
Preparation of MnII(1-nnap)2
Nitrosonaphthol sodium salt (3.182 g, 14.29 mmol) was dissolved in water (200 mL),
and MnCl2 (1.880 g, 9.520 mmol) was added. After stirring for 2 h, the solid that
formed was filtered, thoroughly washed with water and dried in a desiccator. The
solid was recrystallized from CH2Cl2-hexane to give dark brown crystals of MnII(1-
nnap)2 with a m.p. > 300°C (lit.27, m.p. >300°C).
2.2.2.4 Electrochemical preparation of cerium(IV) from cerium(III) using a
divided cell
The required amount of methanesulphonic acid was added to both the anode and
cathode compartments to approximately the same level in each, after which the
required amount of cerium carbonate was slowly added to the anode compartment.
Figure 2.1: Experimental setup for the electrochemical generation of Ce(IV)
Both the anode and cathode compartments were heated (60°C) and stirred (500 rpm)
for the designated time period. After completion of this time period, the reactionmixture from the anode compartment was filtered, and a 5 mL sample of the filtrate
was titrated against a ferrous sulphate solution with ferroin as indicator. This was
done in order to determine the Ce4+ concentration. The results obtained for the
oxidation of Ce3+ to Ce4+ in various methanesulphonic acid solutions of varying
concentrations may be observed in Table 2.3, where the data from triplicate titrations
was stirred for 1 h, after which it was filtered. The so-recovered solid was washed
repeatedly with water (200 mL) to result in a brown solid of Ag2O, which was dried in
a vacuum desiccator for 48 h.
2.3 EXPERIMENTAL PROCEDURES
2.3.1 Oxidative Coupling Reactions
2.3.1.1 Oxidation of alkylphenols using silver carbonate/celite73
General procedure
Before use, the silver carbonate/celite reagent (0.2 mmol of Ag2CO3) was freed from
residual water azeotropically by distillation with toluene. The alkylphenol (0.1 mmol)
was then added to the silver carbonate/celite reagent and the reaction mixture was
stirred in toluene (200 mL) for various reaction times. The reaction mixture was then
filtered to remove the solid phase, the solvent evaporated with a rotary evaporator,
and the resulting mixture analyzed by HPLC and GC-MS.
2.3.1.2 Oxidation of alkylphenols using copper complexes of dicarboxylic
acids76
General procedure
Into a 250 mL reaction vessel, which was fitted with a gas addition tube, a condenser,
a thermometer, and a stirrer capable of operating at speeds ranging from
approximately 800 rpm to 10 000 rpm, was added sodium lauryl sulphate (0.10 g,0.35 mmol), deionised water (75 mL) and the alkylphenol (approximately 65 mmol).
To the resulting slurry (which was stirred between 800 and 10 000 rpm depending on
the experiment), was added a mixture of cupric acetate (1.0-50.0 mmol) and a
dicarboxylic acid (1.0-50.0 mmol) in deionised water (50 mL). The resulting mixture
was stirred for 5 min while heating to temperatures ranging from 60 to 80 °C. Sodium
hydroxide (0.4 M, 100 mL) was added during the course of the reaction to maintain
the pH of the reaction mixture at 9. The mixture was stirred under oxygen or nitrogen
depending on the experiment. The flow of gas was initially rapid to flush the system.
After approximately 30 min, the gas flow was reduced and maintained at a levelsufficient to cause slow bubbling. The reaction mixture was stirred and maintained
under oxygen or nitrogen for time periods varying from 6 to 30 h. The reaction
mixture was then cooled to room temperature and then acidified to pH 3 with HCl (3
M). The reaction mixture was extracted using ethyl acetate (3 x 50 mL), and the
organic layer washed with water (3 x 50 mL) and dried (MgSO4). The organic layer
was then concentrated on a rotary evaporator and analyzed by HPLC and GC-MS.
2.3.1.3 Oxidation of alkylphenols using manganese(III) acetate26
General procedure
The alkylphenol (7.00 mmol) was added to a mixture containing glacial acetic acid
(130 mL) and manganese (III) acetate (3.753 g, 14.00 mmol). The reaction mixture
was then heated to 100°C for 1 h after which it was cooled down, extracted with
chloroform (3 x 50 mL), and the organic layer washed with water (3 x 50 mL) and
dried (MgSO4). The organic layer was concentrated on a rotary evaporator and
analyzed by HPLC and GC-MS.
2.3.1.4 Oxidation of alkylphenols using barium manganate73
General procedure
The alkylphenol (10.00 mmol) in toluene (50 mL) was added to barium manganate
(12.81 g, 50.00 mmol) in a 100 mL round-bottomed flask. The reaction mixture wasthen stirred at room temperature for 1 h, and then vacuum filtered. The solid was
washed repeatedly with ethyl acetate (total volume of 150 mL), and the combined
organic washings concentrated on a rotary evaporator and analyzed by HPLC and
2.3.1.5 Oxidation of alkylphenols using a (nitrosonaphtholato)metal
complex27
General procedure
A mixture of the alkylphenol (1.00 mmol), the (nitrosonaphtholato)manganate
complex (0.0399 g, 0.100 mmol) and triphenylphosphine (0.2885 g, 1.100 mmol), in
dry CHCl3 (30 mL), was stirred for 5 h at 23°C under an oxygen atmosphere (1 atm).
The reaction mixture was then quenched with 2 M HCl (50 mL). The aqueous mixture
was extracted with CHCl3 (3 x 25 mL), and the organic layer washed with water (3 x
25 mL) and dried (MgSO4). The organic layer was then concentrated on a rotaryevaporator and analyzed by HPLC and GC-MS.
2.3.1.6 Oxidation of alkylphenols using FeCl3 in an organic solvent77
General procedure
A mixture of the alkylphenol (7.0 mmol) and FeCl3 (2.271 g, 14.00 mmol), in an
appropriate solvent (20 mL), was stirred in a round-bottomed flask at 50°C for 2 h.
The reaction mixture was then decomposed with dilute HCl (50 mL), and the organic
layer washed with water (3 x 20 mL) and dried (MgSO4). The organic layer was then
concentrated on a rotary evaporator and analyzed by HPLC and GC-MS.
2.3.1.7 Oxidation of alkylphenols using FeCl3 without solvent77
General procedure
The alkylphenol (7.0 mmol) and FeCl3·6H2O (2.271 g, 14.00 mmol) were mixedtogether without any solvent, and the mixture then placed in a test tube and kept at
50°C for 2 h. The reaction mixture was then decomposed with dilute HCl (50 mL),
and the aqueous layer extracted with ethyl acetate (3 x 25 mL). The organic layer
was washed with water (3 x 25 mL), dried (MgSO4), concentrated on a rotary
General procedureThe alkylphenol (10.00 mmol) was added to silver oxide (4.635 g, 20.00 mmol) in 150
mL of methanol, after which the reaction mixture was stirred for 1 h at room
temperature. The solids were removed by filtration and washed with hot toluene, the
toluene then being combined with the filtrate. The resulting organic solution was
concentrated on a rotary evaporator and the reaction mixture analyzed by GC-MS
and HPLC.
2.3.1.9 Oxidation of alkylphenols using lead tetra-acetate79
General procedure
The alkylphenol (7.00 mmol) was dissolved in toluene (100 mL) and stirred while lead
tetra-acetate (6.2058 g, 13.997 mmol) was slowly added to the reaction mixture over
1 h. The reaction mixture was then washed with water (3 x 75 mL) and the organic
layer dried (MgSO4) and concentrated on a rotary evaporator. Analysis was carried
out using GC-MS and HPLC.
2.3.1.10 Oxidation of alkylphenols using Ce4+
General procedure
The required amount of Ce4+ solution and the alkylphenol were added together in a
round-bottomed flask in the required solvent and stirred vigorously at 750 rpm for the
required time period. The reaction mixture was then extracted using ethyl acetate (3x 25 mL), and the organic layer washed with water (3 x 25 mL) and dried (MgSO4).
The organic layer was then concentrated on a rotary evaporator and analyzed by GC-
There are several mechanisms that may be proposed by which phenolic substrates
may oxidatively couple with one another to form dimers. Naturally, the nature of the
phenolic substrate plays a crucial role in the mode by which it ultimately combines
with another substrate molecule. Of significant importance in this regard is the nature
of substitution of the phenolic ring, not only encompassing the number and type of
substituents, but also the positions they occupy relative to the hydroxyl moiety and
each other.
When one considers, for simplicity sake, the coupling of an unsubstituted phenol with
another such substrate, six modes of coupling may be identified. In five of these, (A-
E) shown in Scheme 22, the immediate precursor to the coupled product is shown as
the phenoxyl radical (which is resonance stabilized, see previous Scheme 7), and theimmediate product upon coupling is the dienone form of the dimer, which then
Scheme 23 is an illustration of the sixth possible mode of coupling (F) in which two
phenoxyl radicals combine through oxygen, resulting in the peroxide as shown.
Scheme 23: Peroxide formation from the coupling of phenoxyl radicals
Thus when two phenoxyl radicals couple with another, they may do so in one of the
following ways:
• Ortho C-ortho C coupling (A): A resonance form of the phenoxyl radical in
which the radical is centered at the ortho position couples with another
identical species;
• Para C-para C coupling (B): A resonance form of the phenoxyl radical in which
the radical is centered at the para position couples with another identical
species;
• Ortho C-para C coupling (C): A resonance form of the phenoxyl radical in
which the radical is centered at the ortho position couples with another
resonance form of the phenoxyl radical in which the radical is centered at the
para position;
• Ortho C-O coupling (D): A resonance form of the phenoxyl radical in which theradical is centered at the ortho position couples with the oxygen-centered
radical of another phenoxyl species;
• Para C-O coupling (E): A resonance form of the phenoxyl radical in which the
radical is centered at the para position couples with the oxygen-centered
• O-O coupling (F): two phenoxyl moieties combine through their oxygen-
centered radicals.
Because of the numerous pathways through which phenoxyl radicals may react with
one another to form dimers, the oxidative coupling of unsubstituted phenol itself
results in numerous products, and there is no known process in which the yield and
selectivity to one specific product is high enough to term the process a successful
one. In addition to the above six modes of coupling, one must also bear in mind that
dimeric products that form are also capable of reacting further with either the
substrate and/or dimeric products in the reaction mixture, forming polymeric species.Oxidation mixtures of unsubstituted phenols thus result in a complex mixture of
dimeric, polymeric and unreacted compounds, often with poor carbon accountability.
All of these factors make the oxidative coupling of unsubstituted phenol itself a very
unattractive prospect when the desired product is a specific dimeric form, for
example, the industrially useful compound 4,4 ’-dihydroxybiphenyl (6).
3.1.1 Molecular Orbital Calculations for the Coupling of Phenol
In order to ascertain the likelihood that the phenoxyl radicals will couple as shown in
Schemes 22 and 23, it was deemed appropriate to calculate the relative stabilities of
the dienone dimers as well as the phenolic dimers. Since these species are all
isomeric, their relative stabilities can be obtained by comparing their heats of
formation (? f H) directly. Given in Table 3.1 are the theoretical heats of formation for
the coupled dienones (? f Hd) and phenols (? f Hp), calculated at the PM3 semi-empirical
molecular orbital (MO) level. In the final column of the table, the difference betweenthe heats of formation of the phenolic dimers and their corresponding dienones has
The peroxide PhOOPh, formed by means of O-O coupling (mode F), can be shown to
be much less stable, having a calculated heat of formation of +34.61 kcal/mol.
From this investigation, it may be concluded that both C-C and C-O dienone products
are likely to form in the oxidative coupling of unsubstituted phenol. Hence their
corresponding phenolic forms are also likely, though it must be reiterated that these
calculations do not convey any information on relative rates of formation, and so the
actual product distribution cannot really be predicted. The peroxide, PhOOPh, on the
other hand, appears unlikely to form due to its low stability relative to the other
products. These calculations thus confirm reports that the oxidative coupling of phenol results in a wide product distribution, both C-C and C-O coupled, and cannot
be used with much success when a single dimer is the desired product.
Experimentally, this work did not involve unsubstituted phenol as a substrate for the
above reasons. However, the above MO study was extrapolated to two other
substrates, 2,4-di-t -butylphenol and 2,6-di-t -butylphenol, the results of which are
discussed in the relevant sections.
3.2 THE OXIDATIVE COUPLING OF 2-t -BUTYLPHENOL
The literature contains many reports that deal with the successful C-C coupling of
disubstituted phenols, such as 2,4- and 2,6- dialkylphenols, but is, however, virtually
devoid of studies carried out on monosubstituted phenols, such as 2- t -butylphenol.
This is certainly because both 2,4- and 2,6- disubstituted phenols can each only
effectively C-C couple at one particular carbon position, namely the 6- and 4-positions, respectively, thus affording high yields and selectivities to the desired
product with these substrate types. (This is obviously assuming that C-C coupling is
less likely to take place at aromatic carbon positions that already bear a substituent,
this assumption having been confirmed by PM3 semi-empirical MO calculations which
are discussed later). However, in the case of the monosubstituted phenols, such as
2-t -butylphenol (35), there are two positions available through which C-C coupling can
occur, the 4- and 6- positions, and so the number of possible products increases and
therefore results in low yields and selectivities to any one desired product. Thus
these substrate types generally result in reactions that are not significantly successful,and hence their virtual absence of mention in the literature.
(35)
There therefore exists a need to investigate the oxidative coupling of monosubstituted
phenols in more depth in order to, at best, develop a process that leads to higher
yields of the required materials or, at worst, contribute positively towards this field of
chemistry by obtaining further information associated with this reaction, since there
does not appear to be much mention of it in the literature. To this end, a variety of
oxidizing agents were used in the investigation of the oxidative coupling of 2-t -
butylphenol in order to attempt to form the para-para C-C coupled product, 3,3’-di-t -
butyl-4,4’-dihydroxybiphenyl (molecule (39) in Scheme 25), in high yield and
selectivity.
The aim of this section of the work was to determine whether any one particular
oxidizing agent afforded optimal results compared with the other agents used, and
whether the substrate molecule could, in fact, be C-C coupled selectively through itspara position despite the additional availability of its ortho position.
3.2.1 The Range of Possible Products During the Oxidative Coupling of
2-t -Butylphenol
The reaction mechanisms possible for the oxidative coupling of substituted (mainly
disubstituted) phenols have been discussed at length in the literature.22,23 As
mentioned previously, the number of products possible with 2-t -butylphenol as a
substrate is most likely to be greater than that with 2,4- or 2,6- disubstituted phenols
as substrates. In order to hypothetically predict the types of coupling products
possible with (35), one needs to look at the possible coupling modes of the initial 2-t -
butylphenoxyl radical that is formed. Due to resonance stabilization, the radical maybe centered at either the 6- or the 4- position, or on oxygen itself. In either of the two
cases where the radical is centered on an aromatic carbon atom, a number of
products can form, as is illustrated in Schemes 24 and 25. In these schemes, only
the phenolic forms of the coupled products are given, and not the primary dienone
products, for the sake of brevity.
Scheme 24, in which the radical is centered at position 6, shows how the
monosubstituted phenoxyl radical can couple with either another radical in which the
unpaired electron is centered at the 6- or 4- position, or also on oxygen, affording
phenolic dimers (36), (37) and (38) as products, respectively.
Scheme 25 is similar but the initial phenoxyl radical is centered at position 4, thus
affording compounds (37), (39) and (40) upon combining with the various species as
3.2.2.1 Vanadium(V) oxytrichloride and vanadium(IV) tetrachloride as
coupling agents
As discussed previously, literature reports show that vanadium(V) oxytrichloride61 and
vanadium(IV) tetrachloride react with phenol to give exclusively para -coupled
products, thus implying that these agents are highly selective and para -directing in
their action. Furthermore, vanadium(V) is reported to follow a non-radical mechanism
in which an intermediate with a considerable cationic character is developed,
ultimately ensuring the exclusive formation of para -coupled products (see previous
Schemes 8 and 9).
82
However, for the purposes of our investigation, due to theprohibitive costs of these oxidants, and due to the fact that a vigorous evolution of
HCl gas is accompanied by their reaction with the substrate, implying both
economical and environmental non-viability, it was decided not to assess their effect
on 2-t -butylphenol as substrate.
3.2.2.2 Silver carbonate supported on celite as coupling agent
Silver carbonate on a celite support is also known28 to be a highly specific and
selective oxidizing agent for C-C coupling when reacted with disubstituted phenols.
For example, when silver carbonate/celite was reacted with 2,6-di-t -butylphenol, the
diphenoquinone (10) was the primary product formed. The redox potential of the
oxidant (Ag+ + e- ? Ag ~0.80 V) is thus high enough to oxidize the initially formed
4,4’-dimer (16) to the corresponding 4,4’-diphenoquinone (10). An attractive prospect
with this agent is that the reactions were performed under very mild conditions.
Though the cost of the agent is rather high, it was deemed plausible that some formof recycle would circumvent this disadvantage, and so it was decided to investigate
silver carbonate/celite as the coupling agent for 2-t -butylphenol, and thus to compare
the results obtained with those achieved with disubstituted phenols.
Thus a water-free silver carbonate/celite oxidant was prepared and the substrate
added to it, and the mixture stirred at room temperature for either 1 h (reaction 1,
Table 3.2) or 20 h (reaction 2, Table 3.2). From the results in Table 3.2, it can be
seen that reaction 1 did not afford a high selectivity to the desired coupled product
3,3’-di-t -butyl-4,4’-dihydroxybiphenyl (39). From standard curves using HPLCanalyses, it was determined that in reaction 1, the selectivity to (39) was only 25.57
%.
A GC-MS experiment of the reaction mixture (from reaction 1) showed that there were
four products, at retention times of 12.76, 13.29, 14.18 and 15.27 min, that had the
same molecular ion mass (M+) of 298 mass units (the mass of the desired product).
The GC trace and the associated mass spectra of these four isomeric products maybe observed in Figure 3.1 and Appendices 3.1-3.4, respectively.
Figure 3.1: GC trace of product mixture obtained in reaction 1, Table 3.2
An injection of the standard material for (39) showed that it eluted at 15.27 min,
confirming the presence of the desired product in the reaction mixture. The MS
fragmentation patterns of each of these products (Appendices 3.1-3.4) were found to
be, unsurprisingly, somewhat similar in that many of the mass fragments were
common to all four products. The main difference between these mass spectra was
the relative abundance of the various mass fragments in one spectrum compared to
that in another spectrum. It should be noted that products (36), (37), (38) and (40) all
have the same M+ (298 mass units) as the desired product (39). However, to identify
the exact structures of the products from their MS fragmentation patterns alone wasnot possible, and an exhaustive separation procedure would be required followed by
individual characterization. This was not deemed necessary since the objective of
this study was to increase the selectivity to (39).
In an attempt to increase the yield of (39) that was obtained in reaction 1, the reaction
time was extended from 1 h to 20 h, with all other variables remaining constant
(reaction 2, Table 3.2). The effect of this change was a dramatic increase in theconversion of the substrate (from 10.98 to 78.58 %), but with an equally dramatic
decrease in selectivity to (39) [from 25.57 to 3.56 %]. The number of moles of (39)
formed in each of these two reactions was calculated from standard curves using
HPLC analyses, and may be observed in Table 3.3.
Table 3.3 Amount of 3,3’-di-t -butyl-4,4’-dihydroxybiphenyl (39) formed in
reactions 1 and 2
Reaction
No.
Time
(h)
Mass of (35) used
(g)
Moles of (35) used
(mmol)
Moles of (39) formed
(mmol)
1 1 0.1576 1.049 0.0147
2 20 0.1611 1.072 0.0150
It is clear that in both reactions in which the amounts of substrate used was very
similar, irrespective of the reaction time, the same amount of product was formed.
Reaction time did not seem to have an effect on the yield of the desired product (39).
This may imply that the desired reaction is rather fast, and that an increase in time
after the formation of (39) merely resulted in side product formation, and hence the
overall decrease in selectivity to (39). However, it may also be probable that the
increased reaction time allowed the formed product to react further, thus also
Further scrutiny of the GC trace from reaction 1 (Figure 3.1) revealed the two
products at retention times 20.67 and 21.79 min. MS Data showed both these
products to have an M+ of 446 mass units, and their retention times alone hinted atthe possibility that these were large molecules. When the possible products with m/z
= 446 mass units were investigated, it was found that the C-O coupled product (41) in
Scheme 26 (where n=1), had this required mass. Other workers using 2,6-
dimethylphenol as the substrate and manganese oxide as the coupling agent also
found these types of compounds in their reaction mixtures (see Schemes 14 and
15).37
Scheme 26: The C-O coupling of 2-t -butylphenol to afford (41)
When C-O coupling occurs, the reaction mixture becomes much more complex with
side product formation becoming even more significant. It must also be noted that
when an oxidizing agent is reacted with 2- t -butylphenol, the polyether (41) is only one
of the possible products that could have m/z = 446 mass units. Another possibility is
the following C-C coupled product, which is also deemed highly feasible.
Furthermore, a product having both C-O and C-C coupling may also account for the
mass of 446, but mass fragmentation patterns alone do not suffice for the exact
structure determination of such molecules.
Upon completion of reaction 2, i.e., after 20 h reaction time, it was noted that these
two products with m/z = 446 mass units disappeared, an indication that they reacted
further to form longer chain polymers, and which could then not be detected by this
technique. To add credence to the latter statement, reaction mixture 2, upon workup,was very tarry, dark in colour and had a high viscosity, an indication of the presence
of polymers.
It was thus concluded that silver carbonate/celite was not suitable as a coupling agent
for 2-t -butylphenol under the conditions investigated in this study, despite its reported
success with disubstituted phenolics: the reaction was very inefficient and the
selectivity to (39) was unacceptably low. No further work was thus conducted using
this oxidant and 2-t -butylphenol.
3.2.2.3 Copper acetate, in the presence of a dicarboxylic acid, as coupling
agent
Copper acetate, in the presence of the dicarboxylic acid oxalic acid,46,76 was also
investigated as a coupling agent for 2- t -butylphenol. Cupric salts of dicarboxylic acids
have been reported to couple phenolic substrates, the products of which were
combined at unsubstituted ortho- and para- positions in a manner characteristic of single-electron oxidizing agents. The higher oxidized products such as the
diphenoquinones were generally not produced with this coupling agent. Furthermore,
only reactions of disubstituted phenols have been reported,46 due to the increased
possibility of polymer formation with substrates that are less substituted.
A slurry of the substrate and sodium lauryl sulphate in deionised water was treated
with a mixture of cupric acetate and oxalic acid, also in deionised water, and the
resultant mixture stirred rapidly for 5 min while heating to 60°C. After the addition of
sodium hydroxide (in order to achieve a pH of 9), the mixture was heated and stirred
for 10 h under an oxygen atmosphere, cooled and worked up. Reaction 3 in Table
3.2 summarizes the result of this experiment. Disappointingly, this was not promising:
although the conversion of the substrate was high (86.32 %), the amount of (39)
formed was extremely low, with a selectivity to (39) of 1.30 % being calculated. This
implied that side product formation in this reaction was highly significant. Consider
the HPLC trace obtained upon analysis of the reaction mixture (Figure 3.2):
also reported to afford mostly polymeric products.26 Despite these reports, an
investigation of this oxidant with 2-t -butylphenol was undertaken.
2-t -Butylphenol was thus treated with a manganese(III) acetate/glacial acetic acid
mixture for 1 h at 100°C. Reaction 4 in Table 3.2 is a summary of the obtained
results. A high conversion (85.38 %) indicated the high reactivity of 2-t -butylphenol
with managanese(III) acetate, but the selectivity to (39) was, once again, extremely
low at 2.14 %. From the GC trace, the products at retention times 12.76, 13.29,
14.18 and 15.27 min (with m/z = 298 mass units) were again prominent (as they were
in the Ag2CO3/celite work), with no trimeric species being observed at retention timesof 20.65 and 21.79 min (with m/z = 446 mass units). The reaction mixture was very
tarry upon work-up, and the presence of long chain polymers was thus a distinct
possibility (these not usually being observable under the GC-MS conditions used).
Thus manganese(III) acetate proved also to be unsuitable for the oxidative coupling
reactions of 2-t -butylphenol under the reaction conditions employed.
3.2.2.5 Barium manganate as coupling agent
Barium manganate is known to effectively couple substituted phenolics,73 and it was
thus decided to investigate its effect on 2-t -butylphenol. This substrate, in CH2Cl2 as
solvent, was treated with excess BaMnO 4 at room temperature for 1 h. The results
obtained are summarized in Table 3.2 (reaction 5). These show, once again, a very
low selectivity to (39) [4.55 %], despite a reasonable conversion of the substrate
(66.96 %). Once again, the GC and HPLC traces indicated that a large number of
products had formed in the reaction. From the GC trace (Appendix 3.5), the isomericproducts at retention times 12.79, 14.21 and 15.30 min were again prominent (as they
were in the silver carbonate/celite work). In addition, products at retention times
20.84, 22.00, 26.41 and 38.04 min all had the same M+ value (446 mass units), and it
has already been speculated that these compounds are isomeric trimeric species
It was concluded that BaMnO4 was therefore not an ideal coupling agent for (35) in
these conditions due to the low selectivity to (39).
3.2.2.6 Ferric chloride as coupling agent
2-Naphthol has been successfully coupled using FeCl3, known as a one-electron
oxidant, as coupling agent.77 2-t -Butylphenol was thus treated with this ferric species
for 2 h at room temperature with chloroform as the solvent. Table 3.2, reaction 6,
illustrates that no p -p coupled product was formed under these conditions, as was
observed from standard HPLC data, and confirmed by GC-MS experiments, despitetotal substrate conversion. The nature of the products obtained in this reaction also
varied substantially from those experiments already discussed, as was apparent from
the lack of common retention times and m/z values when comparing the various
traces. However, due to the poor results achieved in this reaction, these products
were not characterized.
Reports exist that claim that many phenols having steric bulk in the vicinity of the
hydroxyl moiety do not couple successfully in the presence of FeCl3. It has been
stated that this is due to prevention of formation of the phenoxyl–iron complex that is
required to form for an efficient reaction.22 Thus carbon-carbon coupling is only
favoured when the phenoxyl radicals that are produced remain complexed through
oxygen to the respective iron atoms during the coupling step. This may offer a
reasonable explanation for our findings using (35) as the substrate, since this
compound does have steric bulk, in the form of the tert -butyl group, in the ortho
position to the hydroxyl group. FeCl3 was thus not investigated further in thisinstance.
Silver oxide provided high yields and selectivities to the desired C-C coupled productswhen disubstituted substrates such as 2,6-di-t -butylphenol were employed in its
presence.78 The mechanism of this reaction, however, is possibly the least well
understood of all the oxidative coupling processes, though it is postulated to occur on
the metal surface.22
In our investigation, commercially available silver oxide was used in addition to silver
oxide that had been prepared in our laboratories by treatment of silver nitrate withaqueous sodium hydroxide. The use of this oxidant implies exorbitant costs (unless
some form of recycle may be used for the silver metal so-produced), irrespective of
the results achieved.
The alkylphenol was added to silver oxide in methanol, and the resultant reaction
mixture was stirred at ambient temperature for 1 h. Reaction 7 in Table 3.2 thus
showed a high conversion of the substrate (96.00 %), but with a selectivity of only
7.29 % to (39). The kinds of products obtained were similar to those from the silver
carbonate/celite work, as concluded from GC data. The coupled product 3,3’-di-t -
butyl-4,4’-dihydroxybiphenyl (39) was the most prominent on the GC trace, and no
other products were observed above the retention time of 16 min. It thus initially
appeared as though the selectivity to (39) would be high, but standard calculations
proved otherwise: once again, the conditions under which the GC-MS experiments
were conducted were not conducive to the detection of polymeric materials, as must
have been present in this case to account for the low calculated selectivity.
Silver oxide as coupling agent was thus set aside due to its inefficient action in this
conversion of the substrate was reasonable, and it appeared that the higher
conversions were associated with the lowest selectivities. This was not surprising
since high conversions, in these reactions, generally implied an increased possibilityfor side product and polymer formation. No further work was thus conducted on (35)
as substrate since none of the results obtained showed any promise.
3.3 THE OXIDATIVE COUPLING OF 2,6-DI-t -BUTYLPHENOL
The literature contains many reports that deal with the successful para C-para C
coupling of 2,6-di-t -butylphenol (9).
84-87
In order to understand why this substratereacts so successfully, we may predict that the reasons are due to the fact that (9)
can only effectively couple at one particular carbon position, namely the 4-position,
assuming that coupling at a carbon position already bearing a substituent is less
favoured (as is likely, due to steric implications). This latter assumption, however, will
be investigated further in order to assess its validity.
Furthermore, we may also assume that para C-O coupling will be less likely to occur
than para C-para C coupling because of the increased steric effect that would comeinto play in such a situation, due to the proximity of the t -butyl groups to the hydroxyl
moiety (and thus their proximity to the subsequent O-centered phenoxyl radical
species, effectively hindering its reaction with other radical species). We may thus
predict that side reactions, such as those resulting from para C-O coupling, ortho C-
para C coupling etc., will be less favoured, and that the resultant product mixture in
the oxidative coupling of (9) will show a high yield and selectivity to the desired
product, 3,3’,5,5’-tetra-t -butyl-4,4’-dihydroxybiphenyl (16), via the corresponding
dienone form. (Note that, depending on the oxidant, 3,3’,5,5’-tetra-t -
butyldiphenoquinone (10), may also be the isolated product, which is formed by over oxidation of (16), and that this product may readily be reduced back to the phenolic
form (16).)
Previous studies have been conducted to investigate the effect of various substituents
on the aromatic ring in coupling reactions. It has been reported87 that when larger
groups, such as t -butyl, are present, then carbon-carbon coupling predominates.
However, when the phenolic bears smaller substituents, such as a methyl group,carbon-oxygen coupling becomes more predominant. It was further reported that the
relative rate of this competitive reaction depended largely on the reaction conditions.
As confirmation of the above predictions and thus the success with which (9) can be
oxidatively coupled, this substrate has been reported to predominantly form the para
C-para C dimer in very high yields when subjected to electrochemical oxidation, as
shown in Scheme 19 previously.53
3.3.1 Molecular Orbital Calculations for the Oxidative Coupling of 2,6 -Di-
t -Butylphenol
Predictions made in the preceding section required some theoretical backup and
verification, and MO calculations were thus carried out in order to determine the
preferential mode of coupling of (9). All possible modes were taken into account, with
one exception, the oxygen-oxygen coupling mode to afford the peroxide, since thishad been shown in previous calculations to be highly unlikely. Schemes 27a and 27b
are an illustration of the coupling reactions investigated by means of PM3 semi-
Scheme 27a: Dienone and phenolic dimers from the coupling of 2,6-di-t -
butylphenoxyl radicals by modes G and H
Of the five modes possible, only two are able to result in a final phenolic form of the
product (where both rings are aromatic), modes G and H (Scheme 27a), with the
other modes (I, J and K, Scheme 27b) forming only the dienone form of the dimers
(where at least one of the rings is not aromatic), due to the nature of their structures.(Dienones from modes I, J and K would require the leaving of one or more t -butyl
groups in order to form phenolic products, which we assume will not be a highly
Scheme 27b: Dienone dimers from the coupling of 2,6-di-t -butylphenoxyl
radicals by modes I, J and K
As done previously for phenol, the relative stabilities of the various dimeric species for
coupling modes G-K were obtained by comparing their heats of formation (? f H)directly. Thus the heats of formation obtained for the coupled dienones (? f Hd) and
carbon atom already bearing a substituent are even less favourable, as was predicted
earlier, and for similar reasons.
In conclusion, we may therefore say that the formation of the para C-para C coupled
product will be the most likely when considering calculated heats of formation of the
various species. These MO calculations add credence to previous predictions made
regarding which mode of coupling will be the predominant one when (9) is used as
the substrate of choice.
2,6-Di-t -butylphenol was subjected to the action of a couple of oxidizing agents,namely Ag2O and Cu(OAc)2/oxalic acid, in order to confirm the previous predictions
experimentally in our laboratories, and to compare with the results obtained for other
substrate oxidations. The results of this investigation are reported in the next section.
3.3.2 Oxidative Coupling Reactions of 2,6-Di-t -Butylphenol Using
Various Oxidants
As mentioned previously, the successful oxidative coupling of 2,6-di- t -butylphenol has
been extensively reported in the literature.84-87 In our case, the aim of this
investigation was to determine specifically both conversion and selectivity to the
desired para C-para C coupled products, in order to be able to compare the results
with those obtained when using other substrates, and to confirm data obtained from
MO calculations. Note that, in this case, there are two possible desired products,
3,3’-5,5’-tetra-t -butyl-4,4’-biphenol (16) and 3,3’-5,5’-tetra-t -butyl-4,4’-diphenoquinone
(10). Both of these are obtained by means of coupling mode G, with the quinoneform (10) merely being an oxidized form of the biphenol. Both forms are readily
interchangeable by simple reduction or oxidation (Scheme 28), and hence the
formation of either of these two compounds or a mixture of both is equally desirable,
and a mixture of the two is not necessarily deemed a disadvantage in this reaction.
Figure 3.3: GC trace of product mixture obtained in reaction 10, Table 3.5
The major product (99 %) at a retention time of 13.77 min was identified as 3,3’,5,5’-
tetra-t -butyldiphenoquinone (10) by using both retention time and mass fragmentation
pattern comparisons with that of the prepared standard material. The molecular ion
had a mass of 408 mass units as is required for this product (see Appendix 3.6). Theother product present in the reaction mixture to a much lesser extent, with a retention
time of 14.11 min, was identified as the biphenol (16), whose retention time and mass
fragmentation pattern corresponded with that of its standard material (with M+ = 410
mass units, Appendix 3.7). The selectivity to (16) was calculated to be only 1.00 %
(Table 3.5). Since both (16) and (10) are desired products, both resulting from the
same coupling mode G (para C-para C), the overall selectivity of this oxidative
coupling reaction may be said to be 100 % (1 % + 99 %).
The predominance of (10) in this reaction is not surprising since the reaction was
carried out with sufficient oxidant for the overall transformation to (10). The
substrate:Ag2O ratio used was 1:2, which is effectively a 1:4 substrate:Ag+ ratio. It is
presumed that one mole equivalent of Ag+ per substrate molecule is required for the
coupling process itself (i.e., two moles of Ag+ are used to form every biphenol product
molecule), while two mole equivalents of Ag+ are then required for the oxidation of the
As the ratio of substrate to oxidant is increased, it is clear that the conversion of
substrate decreases, not unexpectedly. The diphenoquinone (10) remains the
predominant final product, implying that the oxidation of (16) to (10) is possibly amore facile reaction than the coupling of (9) to form (16). In addition, the overall
selectivity topara C-para C coupling remained 100 % in all cases.
The intermediate dienone, i.e., 3,3’,5,5’-tetra-t -butyl-1,1’-dihydro-2,2’,5,5’-
biscyclohexadiene-4,4’-dione (43) [see also Scheme 27a] was never observed in any
of our reaction mixtures. It has been reported88 that the oxidative coupling of 2,6-di-t -
butylphenol with silver oxide, in the absence of air, afforded isolation of this dienoneform of (16), and that this keto tautomer is stable in non-polar solvents,53 but
tautomerizes immediately to (16) in hydroxylic solvents such as methanol. In
addition, Blanchard78 reported that the coupled product (43) was also formed in the
presence of oxygen when 2,6-di-t -butylphenol was oxidized by silver oxide in a non-
polar solvent.
Due to the fact that methanol (a hydroxylic solvent) was our solvent of choice, the
absence of observation of (43) in any of our reaction mixtures (irrespective of
substrate:oxidant ratio) was thus not surprising.
Overall, our findings are thus in agreement with other reports that used silver oxide as
coupling agent, giving para C-para C coupling (mode G) exclusively, with no other
side reactions occurring. No para C-O coupling was ever observed in our
investigations. The previously shown MO calculation data predicted this behaviour.
The bulky t -butyl groups, and thus steric factors, play a significant role in the mode of coupling that (9) preferably undergoes.
3.3.2.2 Copper(II) acetate/oxalic acid as coupling agent
The reaction of copper acetate46,76 in the presence of a dicarboxylic acid, oxalic acid,
with 2,6-di-t -butylphenol was investigated. As reported previously, cupric salts of
dicarboxylic acids have been used to couple phenolic substrates in a manner characteristic of a single-electron oxidizing agent.46
A substrate/sodium lauryl sulphate/deionised water slurry was treated with a cupric
acetate/oxalic acid/deionised water mixture, and stirred rapidly for 5 min while heating
at 60°C. After sodium hydroxide addition (pH 9), the mixture was heated and stirred
for 10 h under an oxygen atmosphere, and then cooled and worked up. Reaction 11
in Table 3.5 summarizes the result of this experiment. (For this reaction, the molar
ratio of the substrate:oxidant was 50:1, the copper salt of oxalic acid thus acting in a
catalytic fashion.) After the required reaction time, no starting material remained, with
the conversion of 2,6-di-t -butylphenol to products being complete. A GC-MS analysis
of the reaction mixture showed that, once again, as with the silver oxide work, only
two products were formed, at retention times of 13.76 and 14.14 min, respectively.
These corresponded with the standards (10) and (16), and the analysis showed again
that (10) was the predominant product (with a selectivity of 96.25 %) as compared to
(16) [with a selectivity of 3.75 %]. Thus the redox potential of the copper salt wasalso high enough to oxidize the initially formed product (16) further to form (10).
These results are thus very similar to those achieved when using Ag2O, reiterating the
ease with which (9) can be oxidatively coupled with high selectivity and yield to the
desired para C-para C coupled product. Once again, the overall selectivity to the
para C-para C coupled products (16) and (10) is 100 % (3.75 % + 96.25 %), with no
para C-O coupled products being observed.
3.3.3 Concluding Remarks on the Oxidative Coupling of 2,6-Di-t -
Butylphenol
From the above information, it is clear that the number of modes in which 2,6-di- t -
butylphenol can theoretically couple is numerous, though, in practice, this substrate is
highly selective when placed under oxidative coupling conditions. High selectivities to
the desired para C-para C coupled products (16) and (10) were achieved with both Ag2O and Cu(OAc)2/oxalic acid (100 % selective in both instances). This is in
agreement with results obtained in the literature.46,78 No para C-O coupled products
were ever observed. Molecular orbital calculations confirmed these observations. In
addition, it was stated earlier that dealkylation of the coupled product and/or substrate
would not be a significant reaction pathway, and this was found to be so since no
dealkylated products were ever evident. In conclusion, it may thus be said that the
presence of the additional t -butyl group in (9), as compared with that of 2-t -
butylphenol (35), obviously plays a critical role in its choice of mode of coupling, and
steric congestion is also a major consideration in these reactions. No further work
was conducted on (9) as substrate since the results were optimal and well known to
the field, and there was thus no need for further investigation.
3.4 THE OXIDATIVE COUPLING OF 2,4-DI-t -BUTYLPHENOL
The oxidative coupling of 2,4-di- t -butylphenol has not been as well documented asthat of 2,6-di-t -butylphenol. However, as with the 2,6-analogue, it is envisaged that
2,4-di-t -butylphenol (44) will carbon-carbon couple primarily at one particular carbon
position, namely the 6-position, for similar reasons to those discussed earlier for the
2,6-analogue. Thus the oxidative coupling of (44) should lead to the ortho C-ortho C
coupled product (45) with high yield and selectivity.
Having said this, it must be bourne in mind that although (44) has only one available
unsubstituted carbon position available for coupling, the possibility of unsubstitutedortho C-O coupling occurring may be greater than that for 2,6-di-t -butylphenol since
the hydroxyl moiety (and hence the phenoxyl radical) of the 2,4-analogue has
decreased steric hindrance compared with that of the 2,6-analogue. This is because
there is only one bulky t -butyl group in close proximity to the OH group in the 2,4-
analogue, but two such bulky groups in the 2,6-analogue. One may propose that 2-t -
butylphenol (35) has similar hindrance in the vicinity of the hydroxyl moiety as that of
(44), and thus when one considers that (35), under oxidative coupling conditions,
afforded no less than four products having the same mass as that of the desired
coupled product (39) [see relevant previous section], one may come to the conclusion
that it is highly likely that some unsubstituted C-O coupling did indeed occur with (35)
[though these four isomeric products were not isolated and characterized]. It
therefore appears likely that some unsubstituted C-O coupling should also be likely
with (44). However, in previous reports,89 it was stated that when (44) was oxidized
with di-t -butyl peroxide at 140°C for 24 h, the coupled product 3,3’,5,5’-tetra-t -butyl-
2,2’-dihydroxybiphenyl (45) was solely formed, and no C-O coupled products wereobserved.
Scheme 30a: Dienone and phenolic dimers from the coupling of 2,4-di-t -
butylphenoxyl radicals by modes L and M
There are only two possible reaction modes (L and M) that are able to afford phenolic
products with both rings aromatic in nature, via the tautomeric rearrangement of the
corresponding dienone forms. The other modes (N, O, P and Q, Scheme 30b) result
only in the dienone forms of the dimers: as with 2,6-di-t -butylphenol, the formation of
the phenolic products for these latter modes would require the leaving of one or moret -butyl groups and it was assumed that this would not constitute a significant reaction
significantly less stable than that from mode L (by 17.70 kcal/mol with ? f Hp = -105.61
kcal/mol).
Coupling via the unsubstituted ortho C and a substituted para C (mode N) is
approximately 10 kcal/mol less favourable than ortho C-ortho C (with ? f Hd = -83.61
kcal/mol), obviously as a result of steric interactions. Mode O, coupling between a
substituted para C and oxygen is less desired still, having ?f Hd = -78.63 kcal/mol.
Finally, when both coupling carbons are substituted, i.e., modes P and Q, the
coupling reaction is highly disfavoured, as would be expected. In conclusion, we may
therefore say that the formation of the ortho C-ortho C coupled product is most likelyto be the favoured one according to these heats of formation data of the various
species.
2,4-Di-t -butylphenol was then subjected to the action of a variety of oxidizing agents,
and the results of these reported in the next section, and compared with the data
obtained from other substrates in a later section.
3.4.2 Oxidative Coupling Reactions of 2,4-Di-t -Butylphenol Using
Various Oxidants
In the course of this investigation, a number of oxidizing agents were selected for
study. Naturally, agents that were considered both economically and environmentally
advantageous were given priority. Furthermore, the aim here was to obtain optimal
results in terms of the coupling of (44) to (45). The experimental results obtained
were compared to those obtained from MO calculations (with mode L being predictedto be the most favourable), and to results obtained from reactions with other
substrates, where appropriate.
The standard material of the desired product, 3,3’,5,5’-tetra- t -butyl-2,2’-
dihydroxybiphenyl (45), was prepared by reacting 2,4-di-t -butylphenol with K3Fe(CN)3.
The product thus formed required purification by recrystallization. The structure of
Data obtained from MS experiments suggested that, in the case of this substrate, the
products obtained stemmed not only from the oxidative coupling process, but also
from their subsequent dealkylation, as well as that of the starting material. For example, when toluene was used as the reaction solvent, the presence of t -
butyltoluene at the retention time of 7.11 min was very significant, contrary to our
initial assumption that dealkylation would not be a main consideration in these
reactions. In addition, the product at 9.60 min was identified as t -butylphenol, most
likely from the dealkylation of the substrate. Furthermore, products at retention times
15.27 and 13.51 min were identified as the tri- and tetra- debutylated coupled
products, respectively, though these were not isolated and characterized, and so themode of coupling that occurred could not be ascertained.
When CHCl3 was used as the reaction solvent (reaction 16), a large variety of
products were noted, including chlorinated 2,4-di-t -butylphenol, t -butylphenol (due to
dealkylation) and various dealkylated coupled products. It seems that coupling does
indeed occur in these conditions but that the resultant products are further
debutylated.
Once again, the steric bulk afforded by the t -butyl group in the vicinity of the OH
group was not conducive to a clean oxidative coupling process, as verified by results
obtained in the 2-t -butylphenol work and as claimed by other workers in the field.22
Due to the poor results achieved with FeCl3, it was therefore deemed appropriate to
sideline this reaction, and not investigate its use any further with 2,4-di-t -butylphenol
as substrate.
3.4.2.2 Silver oxide as coupling agent
Silver oxide is a known one-electron transfer oxidant that converts phenols to
phenoxyl radicals.12,90 These phenoxyl radicals can then undergo the characteristic
C-C and/or C-O coupling processes. Since silver oxide showed high selectivity to the
Two products at retention times 23.84 and 25.70 min had an M+ of 408 mass units,
the required mass for (46). However, because it was not a priority in this
investigation, a separation of these was not carried out, and it is thus not certain
which of the two products was (46), if any. Whatever the case, it is obvious from this
study that the 2,4- and 2,6- di-t -butylphenols behave very differently under identical
conditions when treated with silver oxide. The 2,6-analogue was coupled highly
successfully whilst the 2,4-analogue provided rather disappointing results, with a
number of unwanted side products being formed. Reasons for this are not clear – the
mechanism at work here, as has been stated before, is not well known, but from this
investigation, it may be concluded that the positioning of the alkyl substituents on the
aromatic moiety plays a role in the subsequent reaction of the substrate. It is
plausible that steric hindrance is a factor since one would expect the 4-position of the
2,6-analogue to be less crowded than that of the 6 -position in the 2,4-analogue.
As an aside, and of peripheral interest, was the product at the retention time of 30.74min which MS data showed to have an M+ of 615 mass units (Appendix 3.8). This is
possibly due to a product of multiple coupling such as, for example, the triaryl species
(47) shown in Scheme 31 (where n=1), though the exact nature of the coupling
(whether C-C or C-O) was not verified by further characterization.
Scheme 31: A plausible multi-coupled product (47) with M+ = 615 mass units
Therefore, due to the disappointingly low selectivity to the coupled product (45) and
the subsequent significance of side reactions in these conditions, the reaction of 2,4-
di-t -butylphenol with Ag2O was not investigated further.
3.4.2.3 Potassium ferric cyanide as coupling agent
Ferric cyanide is one of the most widely used oxidizing agents for the generation of
phenoxyl radicals in alkaline solutions.90,92,93 Previous studies94 indicate that the
oxidant must be in excess relative to the substrate, and that K3Fe(CN)6 acts as a one-
electron transfer agent [reaction (A)] involving phenoxide anions (present due to the
basic medium) as the oxidizable substrate. Furthermore, it was found that the rate of oxidation was largely dependent on the basicity of the solution and on the
2,4-Di-t -butylphenol was thus treated with this ferric species in a basic medium
(NaOH) with the reaction conditions and results summarized in Table 3.10.
Table 3.10 Reactions of 2,4-di-t -butylphenol with K3Fe(CN)6
Reaction
No.
Oxidant Time
(h)
Solvent Temp. Conversion
(%)
Selectivity
to (45)
(%)
14 K3Fe(CN)6 2 MeOH/H2O R.T 96.04 83.95
17 K3Fe(CN)6 1 MeOH/H2O R.T 84.53 86.10
It was noted in this reaction that 2,4-di-t -butylphenol (44) was highly reactive under
the reaction conditions employed, with conversions of 96.04 and 84.53 % for reactions 14 and 17, respectively. These reactions were carried out under identical
reaction conditions except that reaction 14 was continued for 2 h, whilst reaction 17
was quenched after only 1 h. The longer reaction (14) afforded the higher conversion
of substrate (not surprisingly), but the selectivity to (45) decreased from 86.10 (after 1
h, reaction 17) to 83.95 % (after 2 h, reaction 14). These results were, however,
significantly superior to those achieved with ferric chloride and silver oxide.
Since potassium ferric cyanide is a known one-electron oxidizing agent, and since the
reaction takes place in a basic medium, the phenoxyl radical is thought to form from
the phenoxide anion (see (A) before). After this, the direct coupling (the FR1
mechanism highlighted earlier, Scheme 6) of two such phenoxyl radicals, in which the
radical is centered at the 6-position (through resonance), takes place to ultimately
afford (45) after tautomerization. Scheme 32 depicts the proposed mechanism at
Also notable was the presence of products at retention times of 24.13 and 31.39 min,
their mass fragmentation patterns being indicative, as before, of products (46) and
(47), respectively, which can only form by coupling modes L or M.
3.4.2.4 Cerium(IV) as coupling agent
3.4.2.4.1 Identification of Ce(IV) as the preferred oxidant
In this investigation, two considerations that had to be bourne in mind were the
economic viability and environmental impact of the oxidant of choice. Although the
oxidative coupling reaction of 2,4-di-t -butylphenol (44) with K3Fe(CN)6 gavesatisfactory results, it is not an environmentally acceptable oxidant nor would it be
industrially attractive. The indirect electrochemical oxidation of organic substrates is
becoming more and more economically viable, and there are many compounds that
are known to be capable of acting as indirect oxidants, including transition metal salts,
cobalt, manganese, iron, lead, silver and cerium. Examples of redox couples that
have been studied include Ce3+/Ce4+, Mn2+/Mn3+, and Mn2+/MnO2.95 An alternative
oxidant was thus sought, one that can be regenerated, implying that recycling thereof
would be feasible, and thus being advantageous from an economic point of view.
Furthermore, if the oxidant could be re-used, this would directly have a positive
bearing on the environment since the potential amount of waste requiring
treatment/storage would be minimized. However, the regeneration of many spent
metals to their higher oxidation states is not always effective, since many metal ion
oxidants have certain properties that make this process difficult. However, the
indirect electrochemical oxidation of phenols using a redox couple remains attractive
and, to this end, we investigated it further.
For the purposes of this study, the Ce4+/Ce3+ couple was extensively investigated,
with its recycle (Figure 3.7) being the driving force for our interest. Among the
electron carriers most commonly used for indirect oxidations, cerium salts appear to
be the most suitable when oxidations must be carried out under mild reaction
conditions. The most common electron valencies of cerium salts are three and four,96
with cerium(IV) behaving as a one-electron oxidant.83,97 The oxidation potential of the
Ce4+/Ce3+ couple is reported to be dependent on the reaction conditions, such as the
ligand. For example, the oxidation potential of this couple in 1 N perchloric, nitric,sulphuric and hydrochloric acids was observed to be -1.70, -1.61, -1.44 and -1.28
volts, respectively.98-101
Figure 3.7: Electron flow in the indirect electrochemical oxidation process
using the Ce3+/Ce4+ couple
The oxidative coupling of 2,6-disubstituted phenols using Ce(IV) in the presence of
perchloric acid is well documented.69,70 However, an alternative acid would be
preferred since perchloric acid has the potential to lead to the formation of
perchlorates which can be hazardous when in contact with organic chemicals.102 It
was thus decided to investigate the use of W.R. Grace’s technology,103 where Ce(IV)
would be reacted with our substrate in the presence of methanesulphonic acid. There
are a number of advantages of using methanesulphonic acid rather than sulphuric
and perchloric acid. These are:
• It is unreactive with both reactants and products.
• It is stable to anodic and cerium oxidations.
• Ce(III) and Ce(IV) are highly soluble in aqueous methanesulphonic acid.
• It has a high current efficiency (>90 %) at high current density (>400 mA/cm2).
source of Ce(III). The experimental setup was given previously (Figure 2.1). The
Ce(III) concentration was kept constant (0.1 M) whilst varying the methanesulphonic
acid concentration initially between 0.5 and 2.0 M. An extensive literature surveyfailed to indicate that oxidative coupling of disubstituted phenols, including 2,4-di-t -
butylphenol, had been previously investigated with Ce(IV) in the presence of
methanesulphonic acid. This work is thus entirely novel, and information gathered
from it is deemed to add to the knowledge base of this field of chemistry.
The aim of this investigation was thus to extensively investigate the oxidative coupling
of 2,4-di-t -butylphenol (44) using Ce(IV) as the oxidant in methanesulphonic acid and,in so doing, to determine the effect of the following on this oxidative coupling process:
• Varying the MeSO3H concentration.
• One or two phase systems with or without added co-solvent.
• Varying the reaction temperature.
• Varying the reaction time.
• Substrate loading.
• Varying the substrate:oxidant ratio.
• Varying the rate of oxidant addition to the reaction mixture.
On completion of these studies, it will then be possible to optimize reaction conditions
so as to improve the yield and selectivity to the desired coupled product (45).
3.4.2.4.2 Oxidation in MeSO3H mediated by Ce(IV) ions
For this investigation, the cerium carbonate concentration was kept constant at 0.1 M,
irrespective of the methanesulphonic acid concentration. The minimum concentration
of methanesulphonic acid was set at 0.5 M (MeSO3H is a mono-protic acid). Hence
the methanesulphonic acid concentration was varied between 0.5 and 2.0 M for the
electrochemical oxidation of Ce(III) to Ce(IV). The results obtained are summarized
in Table 2.3 (Experimental section). It was noted that the highest conversion of
Ce(III) to Ce(IV) occurred when the MeSO3H concentration was 1.0 M (when
substrate is very soluble, such as methanol, acetonitrile or dichloromethane. The first
two of these are water-soluble organic solvents and would ultimately afford a one
phase reaction system, whilst dichloromethane is not and would result in a two phasereaction system. To investigate the effects of these co-solvents on the coupling
process, the reactions were initially performed at R.T. for 1 hour. Since previous
results showed that optimal conversions and selectivities were obtained using 1.0 M
methanesulphonic acid concentrations, this was the concentration of choice in the
following reactions. Results so-obtained are summarized in Table 3.13.
Table 3.13 Effect of co-solvents on oxidative coupling of (44) by Ce(IV) at R.T.
The conversion of 2,4-di-t -butylphenol in aqueous solvent with added MeOH or
CH3CN was similar and high at 94.80 and 93.29 %, respectively. However, the
acetonitrile-containing system afforded a lower selectivity to (45) of 62.37 %, whereas
the methanol-containing system was higher at 69.28 %. When CH2Cl2 was used as
the organic co-solvent, there was a significant decrease in the conversion of (44)[reaction 26, 41.87 %] compared with MeOH and CH3CN. The selectivity to the
coupled product (45) was, however, much higher (85.22 %). This low reactivity of
2,4-di-t -butylphenol with Ce(IV) in H2O/CH2Cl2 is most likely due to the presence of
the two phases, where the substrate prefers to reside within the organic solvent,
implying that reaction can only occur effectively at the boundary surface of the two
phases. In such a system, stirring efficiency would be an important consideration.
Overall, MeOH as co-solvent thus gave the most promising results at ambient
temperatures.
In the next study, using 1.0 M MeSO3H and MeOH as co-solvent, the reaction was
carried out at various reaction temperatures in order to determine whether an
optimum point could be defined. All other variables were kept constant. The results
obtained are contained in Table 3.14.
Table 3.14 Temperature effect on coupling of (44) by Ce(IV) with MeOH as co-solvent
Reaction No. MeOHvolume
Ratio(substrate:oxidant)
Temperature Conversionof (44)
(%)
Selectivityto (45)
(%)
27 20 mL 1:2 0°C 92.29 73.72
28 20 mL 1:2 R.T. 93.42 78.76
29 20 mL 1:2 45°C 95.09 82.76
30 20 mL 1:2 Reflux 96.96 80.50
The total volume of the MeOH and aqueous MeSO3H was also kept constant (40 mL)
for all of these reactions.
After assessing reactions at 0°C, ambient temperature, 45°C and reflux (65°C), it wasfound that high conversions were obtained throughout, even at 0°C (92.29 %). The
highest conversion occurred at the highest temperature (65°C, 96.96 %). A plot of
conversion versus temperature more clearly summarizes the results obtained (Figure
Similarly, Figure 3.10 is a graphical summary of the results obtained in terms of
selectivity, where selectivity was plotted against temperature. From this plot, an
optimal selectivity to (45) was achieved at the reaction temperature of 45°C (reaction
29, 82.76 %). However, as seen above, the conversion of (44) at this temperature
(95.09 %) was slightly lower than that obtained at reflux (reaction 30, 96.96 %). Of
interest is the observed decreased selectivity to (45) at the lower temperatures (73.72
% at 0°C and 78.76 % at R.T.). This is a possible indication that the rate of thecoupling reaction to afford (45) increases with increasing temperature, implying that
the lower coupling rates at the lower temperatures provides the reactive species time
to undergo other side reactions, and hence resulting in the observed lower selectivity
at the lower temperatures. In other words, it appears as though the coupling reaction
to form (45) occurs rapidly at the higher temperatures and less so at the lower
temperatures relative to other side reactions: hence, when the reaction temperature
The best results were obtained in reactions 30 and 33 (Table 3.16) where the volume
of MeOH used was 20 and 10 mL, respectively. It is clear from Figure 3.11 that when
the substrate loading was high (reaction 32, where the MeOH volume used was 5mL), the conversion was a low 55.98 %, while the selectivity was high (84.86 %).
This was surprising since one would have expected that a high substrate loading
would afford high conversions of the starting material. A possible explanation for this
phenomenon is that the substrate solubility in the aqueous medium containing only
small volumes of added co-solvent is low and hence its low conversion. A low
substrate loading, when the volume of MeOH used was 40 mL (reaction 34), showed
a significant decrease in selectivity (60.70 %). This data shows that the substrateloading in these reactions is critical, and should not be too high nor too low for optimal
conversions and selectivities.
The reaction variable investigated next was the oxidant to substrate ratio. In this part
of the investigation, the following reaction variables were kept constant:
• Temperature (reflux).
• Substrate loading (20 mL of MeOH).
• Reaction time (1 hour).
• MeSO3H concentration (1.0 M).
Hence, whilst maintaining a constant substrate amount (moles) in each reaction, the
number of moles of Ce(IV) was varied so that substrate:oxidant molar ratios of 1:0.5,
1:1, 1:1.5, 1:2 and 1:5 were achieved. Refer to Table 3.17 for the results obtained.
Table 3.17 Effect of substrate:oxidant ratio on the coupling of (44) usingCe(IV)
Reaction No. MeOHvolume
Ratio(substrate:oxidant)
Temperature Conversionof (44)
(%)
Selectivityto (45)
(%)
35 20 mL 1:0.5 Reflux 35.50 94.25
36 20 mL 1:1 Reflux 72.90 91.74
37 20 mL 1:1.5 Reflux 83.35 79.29
30 20 mL 1:2 Reflux 96.96 80.50
38 20 mL 1:5 Reflux 97.12 60.31
The poorest results in terms of conversion were achieved in reaction 35 (35.50 %),
where the substrate to oxidant ratio was 1:0.5. However, this reaction also gave the
greatest selectivity to the coupled product (45) of 94.25 %. The highest conversion of the starting material was achieved in reaction 38 where the substrate to oxidant molar
ratio was the greatest (1:5). This result was to be expected since a larger amount of
oxidant in the reaction mixture increases the availability of the Ce(IV) ions to the
substrate (44), thus increasing the conversion. Furthermore, and as expected, the
selectivity to the coupled product in this reaction was also the lowest (60.31 %):
excess oxidant was obviously available for side reactions or further oxidation of the
formed product.
A graph of both percentage selectivity and conversion versus oxidant:substrate ratio
was then plotted (Fig. 3.12) by using the information in Table 3.17.
It is clear from this result that these reaction conditions afford superior results in termsof conversion, which is quantitative, and the associated selectivity to the desired
coupled product (45) is still very high at 90.35 %. Thus slow oxidant addition is
clearly favoured over that of its rapid addition.
3.4.2.4.3 Reaction mechanism for the oxidative coupling of 2,4-di-t -butylphenol
using Ce(IV)
To propose a feasible reaction mechanism for the oxidative coupling of 2,4-di-t -
butylphenol (44) using Ce(IV) in the presence of MeSO3H, all the reaction products
need to be identified. The mechanism must then account for these products. Upon a
thorough examination of the reaction mixtures in the various oxidative coupling
reactions using 2,4-di-t -butylphenol (44) as the substrate and Ce(IV) as oxidant, the
following products [(45), (47), (48), (49) and (50)] were identified from their molecular
ion masses and their mass fragmentation patterns. With the exception of (49) and
(50), all of these products were common to those found in reactions using K3Fe(CN)6
Since Ce(IV) is reacted with (44) in acidic media, the reaction mechanism probably
differs from that established for potassium ferric cyanide (basic media). (When
potassium ferric cyanide is reacted with (44) in basic medium, the hydrogen of the
phenolic OH is abstracted by base to form the anion (Scheme 32), which is then
oxidized to form the phenoxyl radical.) Since Ce(IV) acts as a one -electron oxidant
and the reaction takes place in acidic medium, the assumption can be made that one
electron is removed from the aromatic ring to form the radical cation (51), as shown inScheme 33. This is most likely a facile process due to the electron rich nature of the
aromatic ring due to the presence of electron-donor substituents such as the hydroxyl
and alkyl groups. This radical cation may then lose a proton to afford the phenoxyl
radical (52). One resonance form of (52) is where the unpaired electron is centered
at carbon position 6. Two of these then couple directly together to afford dimer (53)
Scheme 36: Formation of products (49) and (50) from a radical cation
From this scheme, the radical cation (51) produced reacts with water and, after proton
loss, affords the radical (58), which then also undergoes oxidation by losing an
electron. The cation which so forms is then transformed into (50) by the loss of a t -
butyl cation. (Since this is a tertiary carbocation, its stability is reasonably high, and
this step is thus more than plausible. However, since high selectivities to (45) were
observed in these reactions, these debutylated products, though not quantified, werenot formed in amounts significant enough to state that this was a facile process
compared with that of the formation of (45).) Hydroquinone (50) is then readily
oxidized to the quinone (49).
A general trend observed is that an increase in reaction temperature allowed for an
increase in the amount of desired product (45) formed after the same amount of
reaction time (Table 3.14). The desired reaction appears to be a rapid one, and an
increase in reaction temperature plausibly increases it even further, providing less
opportunity for side reactions (such as C-O coupling), and thus possibly accountingfor this general observation. The rate of the desired coupling reaction thus generally
increases with increasing temperature, implying that at lower temperatures, the lower
rate of formation of the desired dimer results in an increased possibility for side
product formation. The statement that the desired coupling reaction is proposed to
occur rather rapidly was also implied by a comparison of the reaction when carried
out for 60 min relative to the reaction when carried out for only 5 min (reaction 30
versus reaction 31, Table 3.15): after only 5 min, a high conversion and selectivity to(45) was obtained, and extending the reaction time to 60 min made no significant
difference to the result.
The effect of substrate loading was also significant in this reaction: the lowest
selectivity to (45) was obtained when the reaction mixture had the lowest substrate
loading (reaction 34, Table 3.16). When the substrate loading was increased, the
selectivity to the desired C-C coupled product also increased. Once again, it may be
concluded that reaction conditions that favour the rapid formation of (45) via the
desired coupling reaction, i.e., by increasing reaction temperature (see above) or
increasing the substrate concentration, disfavour side product formation, and thus an
increase in the selectivity to (45) was observed in these instances. It must be kept in
mind that too high substrate loadings are not ideal since substrate conversions are
rather low in such cases.
Schemes 33-36 therefore account for the products observed in our reaction mixtures,and they hence represent viable pathways by which these products were all formed.
3.4.3 Concluding Remarks on the Oxidative Coupling of 2,4-Di-t -
Butylphenol
From the results obtained, it is clear that 2,4-di-t -butylphenol coupled primarily by
modes L and M. This substrate was not as selective as 2,6-di-t -butylphenol, but high
selectivities to the ortho C-ortho C coupled product (45) were achieved with both the
K3Fe(CN)6 and Ce(IV) oxidants [86.10 and 96.96 %, respectively]. The use of other
oxidants such as FeCl3 and Ag2O afforded results that were less than satisfactory,
with very little, if any, desired product being formed, despite high conversions. This
was rather surprising, especially in the case of Ag2O, since this oxidant was 100 %selective towards the para C-para C coupled product when the 2,6-analogue was the
substrate. These oxidants are obviously not suitable for the purposes of forming (45),
quite possibly due to the mechanisms by which they react in combination with the
positioning of the substituents on the aromatic rings. Molecular orbital calculations
confirmed the preference for coupling mode L and, to a lesser extent, mode M. The
difference in results obtained for the 2,4- and 2,6-analogues may only be explained in
terms of steric crowding, in which the hydroxyl moiety of the 2,6-analogue is well
“surrounded” by the two bulky t -butyl groups, thus disallowing the formation of C-O
coupled products. The 2,4-analogue, on the other hand, is less crowded in the
vicinity of the OH group, and can thus also form some of the C-O coupled product,
resulting in the observed lower selectivities to the C-C coupled product as compared
with the 2,6-analogue.
The work conducted using Ce(IV) as the oxidant is entirely novel, and the results
obtained in these reactions were very promising indeed, with high selectivities andconversions to the desired coupled dimer (45) being achieved. Optimal reaction
conditions included the use of 1 M aqueous methanesulphonic acid as the medium of
choice with added co-solvent (methanol) such that the resultant solution is a single
phased reaction mixture. Furthermore, the optimal reaction temperature was
approximately 65°C (at reflux, giving high selectivities and conversions and implying
ease of application), and lengthy reaction times were not necessary, probably
because the desired coupling reaction takes place rather rapidly. The substrate
loading was an important factor: too high loadings afforded low conversions, and too
low loadings afforded low selectivities. The optimal substrate:oxidant ratio was 1:2,with a slow addition of the oxidant to the reaction medium being slightly favoured over
that of rapid addition. Finally, a mechanism was proposed for this work that
accounted for all the products observed in the reaction mixtures.
3.5 THE OXIDATIVE COUPLING OF 2,4-DIMETHYLPHENOL
From the literature, it was ascertained that the reaction of 2,4-dimethylphenol (59)with various oxidative coupling agents gave complex reaction mixtures.52,105
Furthermore, it was shown that, in the case of the 2,6-dialkylphenols, the bulkiness of
the substituents played an important role in the types of products formed, with the
larger groups, such as t -butyl, giving C-C coupled products almost exclusively.106,107
However, with smaller substituents, such as methyl, C-O coupling has been reported
to occur more readily, and long chain ethers of high molecular weight have been
obtained as products in these cases. When various 2,6-dialkylphenols were oxidizedwith cuprous chloride in nitrobenzene/pyridine, the yield of long chain ethers was
decreased to zero with an increase in bulk of the substituents [-CH3, -CH(CH3)2, -
C(CH3)3].108
Thus the decreased steric effect of the methyl groups of 2,4-dimethylphenol (59),
compared with the t -butyl groups of 2,4-di-t -butylphenol, implies that the course that
the oxidative coupling process takes by the former substrate may be different to that
taken by the latter substrate, and so resulting in different relative product ranges for
the two substrates. Although 2,4-dimethylphenol (59) has only one availableunsubstituted carbon position available for coupling, the possibility for ortho C-O
coupling occurring may be greater than for 2,4-di-t -butylphenol since the hydroxyl
moiety (and hence the phenoxyl radical) of (59) has decreased steric hindrance.
From the literature, it was also ascertained that the non-bonding interactions between
methyl groups in the transition state for coupling of methyl-substituted phenoxyl
radicals are important.109 It was observed that the oxidation of 2,4-dimethylphenol
gave a much higher yield of C-O coupled products than obtained with p -cresol.
110
Anexamination of the staggered approach (60) for the ortho-ortho coupling of 2,4-
dimethylphenoxyl radicals revealed that there were two sets of non-bonding
interactions between the methyl groups.109
Due to these methyl group interactions, a higher energy pathway for C-C coupling
results, and consequently more C-O coupling occurs because the formation of thelatter bond is much less dependent on efficient SOMO-SOMO interactions between
the two radicals.
However, the possible coupling modes of 2,4-dimethylphenol (59) with respect to
available carbon positions for coupling are similar to those of 2,4-di- t -butylphenol.
Although the oxidative coupling reactions of (59) were not investigated by means of
The assumption is made (as for the 2,4-di-t -butyl analogue) that there are two main
reaction modes that can afford phenolic products, namely modes R and S. The
assumption is also made that, for 2,4-dimethylphenol, the other coupling modes (T, U,V and W, Scheme 37b) result only in the dienone forms of the dimers, with the loss of
methyl groups not being considered to be a pathway that will be favoured by these
dienones, and thus the phenolic forms thereof not being considered significant in
these cases. (The phenolic forms of the dienones can only be formed by methyl
substituents being lost from these dienone substrates.)
The oxidative coupling of (59) was thus assessed by reacting this substrate with avariety of coupling agents, and analyzing the reaction mixtures in order to determine
percentage conversions and yields to the desired product, 3,3’,5,5’-tetramethyl-2,2’-
dihydroxybiphenyl (61).
3.5.1 Oxidative Coupling Reactions of 2,4-Dimethylphenol Using Various
Oxidants
The oxidizing agents considered during the course of this investigation were the same
as those used for 2,4-di-t -butylphenol (with the omission of Ag2O), the aim being to be
able to compare results obtained here with those obtained for the 2,4-di-t -butyl
analogue. This would then provide information on the comparative effect of the
various substituents on the aromatic ring on the coupling process.
The standard material for the desired product, 3,3’,5,5’-tetramethyl-2,2’-
dihydroxybiphenyl (61), was prepared by reacting 2,4-dimethylphenol with Ce(IV) in
the presence of methanesulphonic acid. The product thus formed was purified bymeans of column chromatography. The structure was confirmed to be that of (61) by
means of a melting point determination, and the successful comparison of this with
reported values, as well as NMR, IR and GC-MS experiments.
The optimum results obtained when 2,4-dimethylphenol (59) was treated with each of
the various oxidants are summarized in Table 3.19.
Table 3.19 Reactions of 2,4-dimethylphenol with various oxidizing agents
Reaction
No.
Oxidant Time
(h)
Solvent Temp. Conversion
(%)
Selectivity
to (61)
(%)
39 FeCl3 2 CHCl3 50°C 90.98 49.11
40 K3Fe(CN)6 2 MeOH R.T. 70.86 26.72
41 Ce4+
1 H2O R.T. 76.04 57.58
3.5.1.1 Ferric chloride as coupling agent
The one-electron oxidant, ferric chloride, was reacted with 2,4-dimethylphenol even
though it was not successful at all in coupling the 2,4-di-t -butyl analogue to the
desired coupled product (45). 2,4-Dimethylphenol was treated with this ferric species
Product (63) appears to be one in which one of the methyl substituents has been
oxidized to an aldehydic group. The product at retention time 18.15 min has an
identical m/z value compared with that of (61) (i.e., 242 mass units), and is possibly
either the isomeric Pummerer’s ketone (64) or ortho C-O coupled product (65):
When MeOH and ethyl acetate were used as solvents (reactions 42 and 44) in place
of CHCl3, the results obtained in terms of conversion (34.71 and 36.83 %,respectively) and selectivity (11.05 and 9.59 %, respectively) were very poor, and
these reactions were not further investigated. With toluene as the solvent (reaction
43), a large variety of dealkylated products were obtained, including mono-, di- and
tri- demethylated coupled products at retention times 15.23, 15.17 and 12.86 min
respectively (Figure 3.14). (These dealkylated products were not present in reaction
time 15.54 min was identified as the desired coupled product (61).) Higher elution
times 21.78 and 25.34 min both corresponded to products with an M+ of 360 mass
units, i.e., compounds that were higher than dimeric in nature. The success of thereaction of this ferric species with the 2,4-di-t -butyl analogue could thus not be
mimicked in this case, and all evidence obtained in this reaction hinted at the
significant presence of higher oligomeric species, most likely emanating from C-O
coupling (due to its greater propensity for occurring when the substituents are methyl
groups).
3.5.1.3 Cerium(IV) as coupling agent
For 2,4-di-t -butylphenol (44), coupling mode L (ortho C-ortho C) was found to be the
most dominant, but the dominance of this mode in the oxidation of (44) with Ce(IV)
depended on a number of factors such as temperature, concentration of the oxidant,
the substrate:oxidant ratio, the reaction time and substrate loading. Under the correct
reaction conditions, Ce(IV) in methanesulphonic acid afforded high conversions and
selectivities to the desired product (45) when reacted with 2,4-di-t -butylphenol (44).
However, and as discussed earlier, in the case of the 2,4-dimethyl analogue, it is
highly probable that coupling mode S (ortho C-O) may become more prominent due
to the associated decreased steric effects around the hydroxyl group. In order to
investigate this, 2,4-dimethylphenol was treated with Ce(IV) in methanesulphonic
acid. (This reaction is novel in terms of the reaction conditions used for the coupling
of this substrate by Ce(IV).) The effect of the following reaction parameters on this
oxidative coupling process was investigated:
• Varying the MeSO3H concentration.
• One or two phase systems with or without added co-solvent.
• Varying the reaction temperature.
• Varying the rate of oxidant addition to the reaction mixture.
Once again, the cerium carbonate concentration was kept constant at 1.0 M,
irrespective of the methanesulphonic acid concentration, which was varied between
0.5 and 2.0 M for the electrochemical oxidation of Ce(III) to Ce(IV). As describedearlier, an increase or decrease in the acid concentration affects the oxidation
strength of the Ce(IV) ions.70,104 The substrate to oxidant ratio (1:2) was also kept
constant throughout.
Tables 3.22 and 3.23 show what effect a change in the MeSO3H concentration has
on the coupling of 2,4-dimethylphenol (59) by Ce(IV) ions to form (61) at both room
temperature (R.T.) and 80°C, with all other variables remaining constant.
Table 3.22 Oxidative coupling of (59) by Ce(IV) at various [MeSO3H] at R.T.
It can be seen that when the concentration of MeSO3H was 1.0 M (reaction 41), the
lowest conversion of (59) was obtained (76.04%), but the selectivity to (61) wassignificantly higher (57.58%) than in the other reactions. (One will recall that for the
2,4-t -butyl analogue, this reaction under identical conditions gave much lower
conversions and selectivities as compared to this current study.) As found in prior
investigations, the solubility of 2,4-dimethylphenol was found to be low in the aqueous
reaction mixture, and so the reaction temperature was increased to 80°C to increase
Table 3.23 Oxidative coupling of (59) by Ce(IV) at various [MeSO3H] at 80°C
In general, a decrease in terms of both the conversion of (59) and the selectivity to
(61) was noted in each of these reactions when compared to those results at R.T.
This was surprising and in stark contrast to results obtained with the 2,4-di-t -butyl
analogue that afforded much higher conversions and selectivities upon raising the
reaction temperature from ambient to 80°C. Thus, in the case where (59) is the
substrate, an increased reaction temperature is not beneficial to the desired outcome
of the reaction. The reasons for this are not clear, but perhaps the fact that 2,4-
dimethylphenol is a liquid at room temperature whilst 2,4 -di-t -butylphenol is a solid
may affect the reaction in some way. However, it is still very surprising that
conversions dropped upon reaction temperature increase.
As with the 2,4-di-t -butyl analogue, various organic co-solvents were also added to
each of the aqueous mixtures. To this end, MeOH and CH3CN were used in order toafford single-phased reaction systems, whilst CH2Cl2 was used for the biphasic
system. These reactions were performed at R.T. for 1 h while keeping the MeSO3H
Table 3.24 Effect of co-solvents on oxidative coupling of (59) by Ce(IV) at R.T.
The reactivity of 2,4-dimethylphenol towards Ce(IV) ions in the presence of the
organic solvents was high, with reactions 50, 51 and 52 achieving high conversions of
(59) [90.65, 86.17 and 92.46 %, respectively], as expected. These conversions were
much higher than those obtained earlier. However, selectivities dropped significantly.
In the next study, using 1.0 M MeSO3H and CH3CN as co-solvent, the reaction was
carried out at various reaction temperatures with all other variables kept constant.The total volume of the CH3CN and aqueous MeSO3H was also kept constant (40
mL) for all these reactions. It was predicted that increased temperatures in the
presence of this co-solvent would merely decrease selectivities even further. Table
The last variable that was investigated was the rate of addition of the oxidant to the
reaction mixture. In all the previous reactions, the Ce(IV) was added to the mixture
within 30 seconds, but in reaction 55 (Table 3.26), the Ce(IV) [in 20 mL MeSO3H] wasadded to the substrate (in 20 mL CH3CN) over a time period of 30 min. Samples
were taken at 15 min and 60 min of reaction time.
Table 3.26 Effect of rate of oxidant addition
Reaction
No.
Oxidant Time
(min.)
Solvent Temp. Conversion
(%)
Selectivity
to (61)
(%)
55 Ce4+
15 CH3CN R.T. 61.10 49.85
55 Ce4+ 60 CH3CN R.T. 82.81 19.24
When one compares the result obtained after 60 min (where the oxidant was added
over 30 min) with that of reaction 51 (where the oxidant was added within 30
seconds), there is no advantage gained with the slower oxidant addition, contrary to
that found for the coupling of the 2,4-di-t -butyl analogue. In fact, slight decreases
were observed in terms of both the conversion and the selectivity to (61). Reaction
51 thus remains the reaction having the optimal results.
3.5.1.3.1 Reaction mechanism for the oxidative coupling of 2,4-dimethylphenolusing Ce(IV)
In order to propose a feasible mechanism by which this reaction occurs, all the
products of reaction 41 (Table 3.22) were identified. The GC trace has the following
Figure 3.16: GC trace of product mixture in reaction 41, Table 3.22
By making use of data obtained in mass spectral experiments and thus by
determining the molecular ion peaks and interpreting their mass fragmentation
patterns, the products (61), (62), (64), (65), (66) and (67) are proposed to have
formed in this reaction. Products eluting at 8.13 and 8.81 min were identified as (66)
and (67), respectively, while those at 14.64, 15.05 and 18.17 min all had the same M+
of 242 mass units. It is obviously not possible to fully characterize and identify theseproducts from their mass fragmentation patterns alone, but it was deemed feasible
that these peaks could be assigned to compounds (64), (65) and some other isomeric
form thereof, such as any of the dienone dimers shown in Scheme 37b. Compounds
eluting at 8.30 and 15.39 min corresponded with the starting material (59) and desired
From this scheme, the radical (68) undergoes oxidation by losing an electron to afford
the benzylic cation (69), which then reacts with the water present to form the primary
alcohol (70). This is then further oxidized to form the benzaldehyde (66). (Compound(63), detected in reactions using FeCl3 as oxidant, may thus have been a result of the
coupling of (59) with 4-hydroxy-3-methylbenzaldehyde, formed in a similar fashion to
(66) shown in Scheme 39).
3.5.2 Concluding Remarks on the Oxidative Coupling of 2,4-
Dimethylphenol
From the results obtained, it is clear that 2,4-dimethylphenol coupled primarily by
modes R and S. This substrate was not as selective as 2,6-di-t -butylphenol and 2,4-
di-t -butylphenol. Only moderate selectivities to the ortho C-ortho C coupled product
(61) were achieved with oxidants FeCl3, K3Fe(CN)6 and Ce(IV) [49.11, 26.72 and
57.58 %, respectively]. The difference in the results obtained for 2,4-di-t -butylphenol
and 2,4-dimethylphenol can be explained in terms of steric crowding, in which the
hydroxyl moiety of 2,4-di-t -butylphenol is more sterically hindered by the two bulky t -
butyl groups, thus disallowing the formation of the C-O coupled products which are
more prevalent in the oxidation reactions of 2,4-dimethylphenol. In the case of
K3Fe(CN)6, the major product seems to be that formed by C-O coupling.
Furthermore, the methyl groups themselves are quite plausibly more reactive than the
t -butyl groups of the other substrates, resulting also in the lower observed
selectivities.
Results from the work conducted using Ce(IV) as the oxidant are novel, though theefficiency of the coupling in this reaction was found to be only moderately successful
in terms of selectivity to the coupled product (61). Finally, a mechanism proposed for
this work accounted for all the products observed in the reaction mixtures.
In this section, a brief comparison of some of the more important results that have notthus far been discussed and compared, is provided for cases where common
oxidants were used for the various butylated phenol substrates, in order to draw
conclusions from similarities and/or differences observed in the results so-obtained.
(Note that many comparisons have been made in sections prior to this one, but it was
deemed inappropriate to include the discussion that now follows in those self-same
sections.)
3.6.1 Reactions of 2-t -Butylphenol and 2,6-Di-t -Butylphenol with Ag2O
and Cu(OAc)2/Oxalic Acid
Silver oxide was reacted with both 2-t -butylphenol and 2,6-di-t -butylphenol under
identical reaction conditions. The results obtained for these substrates were
significantly different with regards to their selectivity towards the respective desired C-
C coupled products. The results of these reactions are summarized in Table 3.27.
Table 3.27 Comparative data obtained for silver oxide
Reaction
No.
Substrate Oxidant Time
(h)
Solvent Temp. Conversion
(%)
Selectivity to p/p
coupled products
(%)
7
2-t -
butylphenol Ag2O 1 MeOH R.T. 96.00 7.29
102,6-di-t -
butylphenol Ag2O 1 MeOH R.T. 100.00 100.00a
aThe diphenoquinone (10) and diol (16) percentages were 96.25 and 3.75 %, respectively.
From these results, it is obvious that when silver oxide is reacted with 2-t -butylphenol
and 2,6-di-t -butylphenol under identical reaction conditions, the selectivity to the
desired para C-para C coupled products is vastly different (7.29 % and 100.00 %,
position) are highly likely for the mono-substituted phenol, thus resulting in the myriad
of products that was observed.
It was further noted that 2-t -butylphenol was less reactive in terms of conversion than
2,6-di-t -butylphenol in the presence of both silver oxide and copper acetate/oxalic
acid. The decreased reactivity of 2-t -butylphenol, as compared to 2,6-di-t -
butylphenol, may be ascribed to the fact that 2-t -butylphenol, with only one electron-
donating alkyl group (by the inductive effect), is most likely less easily oxidized than
2,6-di-t -butylphenol, which has two such alkyl groups, the latter aromatic ring being,
therefore, more electron dense (and thus more readily oxidized) than that of theformer. It was further noted that no diphenoquinone formation was observed in the
oxidation of 2-t -butylphenol, whereas with 2,6-di-t -butylphenol, the diphenoquinone
derivative was the major product (reactions 10 and 11). This is possibly a further
indication of the difference in oxidation potential of the two substrates, as well as of
their resultant coupled products.
3.6.2 Reactions of 2,4-Di-t -Butylphenol and 2,6-Di-t -Butylphenol with
Ce(IV) in MeSO3H
For the sake of this comparative study and in retrospect, it was thought appropriate to
carry out a reaction (reaction 56, Table 3.29) in which 2,6-di-t -butylphenol (9) was
reacted with Ce(IV) under identical reaction conditions to that of reaction 15, in which
the ortho C-ortho C coupled product of 2,4-di-t -butylphenol (denoted as 2,4 o,o in the
figure), which was arbitrarily assigned a value of 0 kcal/mol for the purposes of ease
of comparison.)
Figure 3.17: Relative energies of dienones (1) and coupled phenols (2)
From these relative energy profiles, it can be seen that the intermediate dienones,
when coupling the 2,4-analogue, have lower relative energies as compared to the
corresponding 2,6-analogue, implying that the 2,4-analogue more readily forms these
intermediates (because of their greater stability) than the 2,6-analogue. This is in
agreement with experimental findings (Table 3.29) where the conversion of the 2,4-
analogue was much higher than that of the 2,6-analogue in the presence of Ce(IV) as
the oxidant. The driving force for the subsequent tautomerization of these dienones
is their gain in aromaticity, and thus an increase in their stability. These calculationsalso show that the ortho C-O coupling (2,4-di-t -butylphenol) of a pair of the
appropriate phenoxyl radicals is about 10 kcal/mol more favourable than the
corresponding para C-O (2,6-di-t -butylphenol) derivative, as suggested earlier and as
a consequence of the greater steric crowding around oxygen in the 2,6-analogue.
The greater energy difference (4.05 kcal/mol, Table 3.4) between dienones formed by
para C-para C and para C-O coupling reactions of 2,6-di-t -butylphenol as compared
with the smaller energy difference (3.37 kcal/mol, Table 3.7) of dienones formed by
ortho C-ortho C and ortho C-O coupling of 2,4-di-t -butylphenol suggests that the 2,6-
analogue favours C-C coupling over C-O coupling to a greater extent than the 2,4-analogue favours C-C coupling over C-O coupling. This may be an explanation for the
observed selectivity difference between these two substrates (when using Ce(IV) and
Ag2O as oxidants), and also for the observation that the 2,4-analogue afforded C-O
coupled product whereas the 2,6-analogue did not. (Note that these MO calculations
do not provide any information on rates of reaction, but only on thermodynamic
aspects thereof.)
Overall, these theoretical considerations thus add credence to experimental findings
During the investigation into the oxidative coupling of various mono- and di-
substituted phenols under a variety of reaction conditions using a range of different
coupling agents, a number of conclusions were drawn from the observed results.
The oxidative coupling reactions of 2-t -butylphenol (35) using various oxidants
produced a large number of products, and so the number of coupling modes that 2-t -
butylphenol prefers is numerous under the conditions that were investigated. There
was no observed selectivity to any single product, and both C-C and C-O coupling
appeared to take place in these reactions, amongst others. Although a large variety
of oxidants were assessed, the selectivity to the coupled product 3,3’-di-t -butyl-4,4’-
dihydroxybiphenyl (39) was found to be low irrespective of the oxidant used. The
highest selectivities were achieved with cerium(IV) sulphate (25.99 %) and silver
carbonate on celite (25.57 %), but these selectivities were obtained at low
conversions of 2-t -butylphenol (26.61 and 10.98 %, respectively). (The reactionconditions under which the cerium work was conducted in this case is novel and has
not been reported elsewhere, as indicated by an extensive literature survey.) All the
other oxidants that were used were found to be totally ineffective in producing the
desired coupled product (39). In many of these latter cases, though, the conversion
of the substrate was reasonable. A general trend that was observed was that higher
conversions were usually associated with lower selectivities. It was therefore
concluded that 2-t -butylphenol (35), due to the number of feasible coupling modes
available to this substrate, showed no promise as a substrate for the selective
coupling to afford (39) as the desired product under the reactions conditions that were
investigated. It therefore appears highly unlikely that 2-t -butylphenol may be used as
a substrate in order to form the desired para C-para C coupled product in an
economically viable process due to the non-selectivity displayed by the substrate,
The oxidative coupling reactions of 2,6-di-t -butylphenol (9) can theoretically produce
numerous products through a number of different coupling modes (G-K). Of these
modes, however, G and H were predicted to be more facile, as shown by molecular orbital calculations. When experimentally investigated, it was found that this
substrate was indeed highly selective when placed under oxidative coupling
conditions. It was observed that high selectivities to the desired para C-para C
coupled products (16) and (10) using Ag2O, Cu(OAc)2/oxalic acid and Ce(IV) sulphate
(100 % selectivity in all instances) were achieved. Both Ag2O and Cu(OAc)2/oxalic
acid also gave high conversions of (9) [both 100 %], but Ce(IV) sulphate only
achieved a 45.01 % conversion of (9) after a reaction time of 1 h. These resultsobtained are in agreement with those reported in the literature,46,78 and the molecular
orbital calculations further confirmed these observations. Thus the presence of an
additional t -butyl group in 2,6-di-t -butylphenol (9), as compared with that of 2-t -
butylphenol (35), has a significant effect on the course of the reaction and on the
preferred mode of coupling of (9). The number of feasible coupling modes for (9) is
thus reduced by the additional substituent, and steric congestion also comes into play
when considering the absence of any C-O coupling for (9) [which was present when
the substrate was (35)]. (Note that results obtained from reactions of Ce(IV) with (9)
have not been reported previously.)
When 2,4-di-t -butylphenol (44) was oxidatively coupled using agents potassium ferric
cyanide and cerium(IV) sulphate, a high selectivity to the desired ortho C-ortho C
coupled product (coupling mode L) was observed. The C-O coupling mode also
appears to occur in these reactions (mode M). This substrate was, however, not as
selective as 2,6-di-t -butylphenol (showing 100 % selectivity to para C-para C coupledproducts). The use of other oxidants such as FeCl3 and Ag2O afforded results that
were less than satisfactory in terms of selectivity to the preferred coupled product
(45), despite high conversions of the substrate. Thus both FeCl3 and Ag2O were
found to be unsuitable for the purposes of forming (45), quite possibly due to the
mechanisms by which they react in combination with the positioning of the
substituents on the aromatic rings. Molecular orbital calculations confirmed the
preference for coupling mode L and, to a lesser extent, mode M. The difference in
results obtained for the 2,4- and 2,6-analogues may only be explained in terms of
steric crowding, in which the hydroxyl moiety of the 2,6-analogue is well “surrounded”by the two bulky t -butyl groups, thus disallowing the formation of the undesired C-O
coupled products. The 2,4-analogue, on the other hand, is less crowded in the
vicinity of the OH group, and can thus also form some of the C-O coupled product,
resulting in the observed lower selectivities to the C-C coupled product as compared
with the 2,6-analogue. The amount of steric congestion around the OH group in both
2-t -butylphenol and 2,4-di-t- butylphenol is probably somewhat similar, thus explaining
the propensity for both substrates to undergo C-O coupling in these conditions.
Once again, the work conducted with 2,4-di-t -butylphenol using Ce(IV) as the oxidant
is entirely novel, and the results obtained in these reactions were found to be very
promising indeed, with high selectivites and conversions to the desired coupled dimer
(45) being achieved. Investigation of the various reaction conditions then produced
the optimal reaction conditions which included the use of 1 M aqueous
methanesulphonic acid as the medium of choice with added co-solvent (methanol)
such that the resultant solution is a single phased reaction mixture. Furthermore, the
optimal reaction temperature was found to be approximately 65°C (at reflux), and
lengthy reaction times were not necessary, probably because the desired coupling
reaction takes place rather rapidly. The substrate loading was an important factor:
too high loadings afforded low conversions, and too low loadings afforded low
selectivities. The optimal substrate:oxidant ratio was 1:2, with a slow addition of the
oxidant to the reaction medium being favoured over that of rapid addition. This
process can be further investigated in terms of industrial viability since Ce(IV) can beregenerated electrochemically from Ce(III) successfully and since 2,4-di-t -butylphenol
as a substrate gave high selectivities to the desired coupled product.
The oxidative coupling reactions of 2,4-dimethylphenol (59) were not as selective as
that of 2,4-di-t -butylphenol or 2,6-di-t -butylphenol under identical reaction conditions.
From the results obtained, it is clear that 2,4-dimethylphenol coupled primarily by
modes R and S. Only moderate selectivities to the desired ortho C-ortho C coupled
product (61) were achieved with oxidants FeCl3, K3Fe(CN)6 and Ce(IV) [49.11, 26.72
and 57.58 %, respectively]. Once again, the difference in the results obtained for 2,4-di-t -butylphenol and 2,4-dimethylphenol can be explained in terms of steric crowding:
the hydroxyl moiety of 2,4-di-t -butylphenol is more sterically hindered by the bulky t -
butyl group, thus decreasing the amount of C-O coupling, which is more prevalent in
the oxidation reactions of 2,4-dimethylphenol (which only has the smaller methyl
substituent in the OH region). (In the case of K3Fe(CN)6, the major product seems to
be that formed by C-O coupling.) Furthermore, the lower selectivities observed for
the dimethyl derivative is also very likely a result of the presence of the benzylic C-Hgroups, which are normally rather activated, especially under radical conditions.
Results from the work conducted using Ce(IV) as the oxidant are again novel, though
the efficiency of the coupling in this reaction was found to be only moderately
successful in terms of selectivity to the coupled product (61).
The aims of this investigation have thus been realized, and the study of the various
coupling reactions, the reaction conditions and the various oxidants and substrates
has led to much new knowledge that may now be added to this field of chemistry.
Many promising results were observed, and feasible reasons given for those
reactions that were not successful. One of the major goals of these investigations
was to find an oxidant that was both environmentally and economically viable, that
afforded high conversions of the substrates and selectivities to the desired coupled
products. From this work, Ce(IV), in the presence of methanesulphonic acid,
appears to be just such a coupling agent, and many promising, novel results were
obtained in oxidative coupling reactions carried out in its presence. The feasibility of its electrochemical regeneration from Ce(III), the fact that substrates such as 2,4-di-t -
butylphenol, when reacted with Ce(IV), gave high conversions and selectivities to the
desired product, and the mild reaction conditions used, makes it an oxidant that may
be feasible for scale-up operations, though scale-up itself will require a separate