Selective oxidation of glycerol catalyzed by iron complexes

Article abstract of DOI:10.1016/j.molcata.2014.10.021

Iron complexes with the polydentate ligand bis(2-pyridinylmethyl)amine (BPA) were employed as catalysts for the oxidation of glycerol using hydrogen peroxide as oxidant. The only observed reaction products were dihydroxyacetone (DHA) and formic acid. Although the overall conversion was in all cases lower than 50%, by accurate choice of the experimental conditions selectivities in DHA up to 100% were obtained.

Full text of DOI:10.1016/j.molcata.2014.10.021

Accepted Manuscript
Title: Selective oxidation of glycerol catalyzed by iron
complexes
Author: Corrado Crotti Erica Farnetti
PII:
S1381-1169(14)00476-2
DOI:
Reference:
http://dx.doi.org/doi:10.1016/j.molcata.2014.10.021
MOLCAA 9323
To appear in:
Journal of Molecular Catalysis A: Chemical
Received date:
Revised date:
Accepted date:
11-8-2014
6-10-2014
11-10-2014
Please cite this article as: C. Crotti, E. Farnetti, Selective oxidation of glycerol
catalyzed by iron complexes, Journal of Molecular Catalysis A: Chemical (2014),
http://dx.doi.org/10.1016/j.molcata.2014.10.021
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.

Selective oxidation of glycerol catalyzed by iron complexes
a
b
Corrado Crotti and Erica Farnetti *
a
CNR – Istituto Struttura della Materia, Unità Operativa di Supporto di Trieste, S.R.14, Km163.5, 34149 Basovizza, Trieste (Italy).
Dipartimento di Scienze Chimiche e Farmaceutiche, Università di Trieste, Via L. Giorgieri 1, 34127 Trieste (Italy).
Corresponding author. Phone: +39-040-5583938; Fax: +39-040-5583903.
b
*
E-mail address: farnetti@units.it (E. Farnetti).
Selective oxidation of glycerol catalyzed by iron complexes
Corrado Crotti and Erica Farnetti
Highlights




 We describe a green process for glycerol oxidation catalyzed by iron complexes.
 Glycerol can be selectively oxidized to dihydroxyacetone.
 The catalytic reactions are carried out at r.t. with hydrogen peroxide as oxidant.
 It is the first example of iron-catalyzed glycerol oxidation to dihydroxyacetone.
Abstract
Iron complexes with the polydentate ligand bis(2-pyridinylmethyl)amine (BPA) were employed as catalysts for the oxidation of
glycerol using hydrogen peroxide as oxidant. The only observed reaction products were dihydroxyacetone (DHA) and formic acid.
Although the overall conversion was in all cases lower than 50%, by accurate choice of the experimental conditions selectivities in
DHA up to 100% were obtained.
Keywords
Iron catalysts; Glycerol valorization; Selective oxidation; Dihydroxyacetone.
Introduction
The attention of both academic and industrial world towards the conversion of biomass to fuels and chemicals has enormously
increased in the last decade, within a common effort to develop effective alternatives to fossil-based carbon sources. Biofuels, i.e.
biomass-derived fuels, are presently employed as substitutes and/or additives to petroleum-derived fuels. Biofuel production has been
long considered with some concern, due to competition of first generation biofuels for edible crops as well as land use (the fuel vs.
food dilemma); however the recently developed biofuel manufacturing procedures from non-edible crops mostly cultivated on poor
soil, from algal lipids, from animal or waste fats (second/third generation biofuels) appear to be ethically acceptable [1].
Even if at present the reasons for increasing substitution of petrol-derived diesel with biodiesel lie more on political rather than
economical ground, any way for lowering biomass-derived fuel prices would translate into enhancement of its use. A major route for
lowering biodiesel cost is the valorization of glycerol, the byproduct of industrial biodiesel production by transesterification of
biomass-derived triglycerides with methanol: at present a large surplus of glycerol is annualy produced, which makes it one of the
most attractive platform chemicals of the near future. In recent years many efforts have been employed to develop efficient catalytic
routes for glycerol valorization, such as reforming to hydrogen and carbon monoxide, hydrogenolysis to 1,2 or 1,3-propanediol,
dehydration to acrolein, oxidation etc. [2]: however, only very few of these processes are presently used for industrial production [3].
Glycerol catalytic oxidation is one of the non-commercialized routes so far, in spite of its apparent simplicity as well as the
commercial relevance of most of the possible reaction products (see Scheme 1): in fact, due to the high functionality of glycerol its
1
Page 1 of 9

selective oxidation has proved to be a hard task to achieve. Dihydroxyacetone (DHA) is the glycerol oxidation product with highest
commercial value due to its employment in cosmetics; its industrial production makes use of a fermentation process, whereas no
catalytic procedure is presently employed on an industrial scale [3,4]. Various heterogeneous catalysts have been reported to promote
glycerol oxidation, however most of these systems produce highly oxigenated componds such as glyceric acid and its derivatives,
whereas only a limited number of heterogeneous catalysts, mostly based on Pd, Pt and Au, produce significative amounts of DHA
[
2a,3,5].
Recently our group [6] as well as other scientists [7,8] published interesting results on catalytic glycerol oxidation to DHA using
iridium or palladium-based catalysts: however, the catalytic reactions so far reported are unsuitable for industrial application, due to
the high cost of the transition-metal catalysts employed. The development of new processes for glycerol oxidation based on less
expensive, more abundant and preferably environmentally safe metals is therefore highly desirable.
Homogeneous iron-based catalysts are attracting rapidly increasing interest as convenient substitutes for the expensive, less available
and often toxic precious metals counterparts [9]. Recently developed processes catalyzed by iron catalysts include reduction of C=C,
CC or C=O bonds [10] as well as asymmetric hydrogenation [11], hydrocarbon oxidation and olefin epoxidation [12].
With regard to reduction and oxidation reactions, mechanistic studies have contributed to shed light upon catalytically active iron
species [13]. Interestingly, several papers have appeared in the last three or four years, reporting alcohol oxidation catalyzed by iron
complexes [14]. Even though none of these papers deals with glycerol oxidation, the properties of some iron catalysts here described
appeared very interesting for our purposes. From a perusal of the results reported in these papers, the use of polydentate ligands in
association with iron appears to be higly effective in enhancing both catalytic activity and selectivity; years ago this type of
association was successfully employed in our lab, in the course of studies regarding iron derivatives with polydentate P,N ligands as
hydrogen-transfer catalysts [15].
Scheme 1. Products obtained by glycerol oxidation.
Two papers by Bauer and coworkers published in 2013 attracted our interest [16]: they describe the oxidation of alcohols catalyzed
by iron complexes with bis-picolylamine ligands. Such catalysts display a remarkable selectivity towards secondary OH groups in
comparison to primary hydroxyls: notably, the authors also report the oxidation of diols to the corrresponding ketoalcohols [16b].
These iron-based complexes appeared to be the perfect candidates as catalysts for glycerol oxidation to DHA.
Here we describe our results in the oxidation of glycerol catalyzed by iron complexes in association to the tridentate ligand bis(2-
pyridinylmethyl)amine (BPA) (see Figure 1). To the best of our knowledge, the data here reported represent the first example of
selective oxidation of glycerol to DHA promoted by an iron catalyst.
Figure 1. bis(2-pyridinylmethyl)amine (BPA).
2
Page 2 of 9

2
. Experimental section
2
.1. General
Some reactions and manipulations (where indicated) were performed under argon atmosphere, using standard Schlenk tube
techniques. The GC standard naphthalene was recrystallized from ethanol. All the other chemicals were reagent grade and were used
as received from the commercial suppliers.
2
.2. Instrumental
1H and 13C NMR spectra were recorded either on a Varian 500 spectrometer operating at 500 MHz and 125.68 MHz, respectively, or
on a Jeol EX400 spectrometer operating at 400 MHz and 100.4 MHz, respectively; chemical shifts were measured relative to the
residual solvent signal. Assignments were verified by COSY and HSQC spectra.
Electrospray mass spectra were obtained by an ion-trap instrument (ESI-MS Bruker Esquire 4000) equipped with an electrospray ion
source. The instrumet performed with 10.0 psi nebulizer pressure, end-plate offset -500V, capillary 4000V and capillary exit at
-1
1
2
13.3V. The drying gas (N ) flow was 5 L min and the spectral range was from m/z = 100 to 1200.
The chemical yields of the catalytic reactions were determined by GC on an Agilent 6850 instrument with helium as carrier gas and a
®
TCD detector. Reaction mixture diluted 1:1.5 in methanol was injected at 100 °C through the cool on-column injector in a Rtx -Wax
®
Restek capillary column (30 m length, 0.32 mm ID, 0.5 μm df) protected by a Restek Hydroguard FS precolumn (5 m length, 0.53
mm ID).
2
.3. Synthesis of the iron catayst
Compound [Fe(BPA) (OTf) ] was prepared according to the reported procedure [16a].
.4. Procedure for catalytic reactions
In a typical catalytic reaction, 3.54 mg (0.010 mmol) of Fe(OTf)
mmol) of BPA and, after 5 min, 0.35 ml of 1.0 M glycerol solution in water (0.35 glycerol mmol) were added through a micro-
syringe leading to a deep orange solution. Then H (100 l, 30 wt % in H O 0.98 mmol) was slowly added dropwise via micro-
syringe under vigorous stirring until addition was complete.
In case of reactions carried out under argon, the Fe(OTf) and BPA solution was deaerated by bubbling argon through a needle for
0 min. After addition of the glycerol solution, the reaction vessel was closed with a serum cap and hydrogen peroxide was added by
2
2
2
2
were dissolved in 0.35 ml of MeCN; to that solution, 5.4 l (0.030
2
O
2
2
2
1
a micro-syringe; then the vessel was heated under vigorous stirring to the chosen reaction temperature in a thermostatted oil bath.
After the desired reaction time, the catalytic reaction was stopped by cooling the Schlenk tube to -18°C and letting air in under
stirring.
2
.5. Analysis of the reaction mixtures
Qualitative analysis was accomplished by NMR; the data were compared with those of authentic samples obtained by conventional
routes or by commercial suppliers. Quantitative evaluation of product distributions was performed by GC with naphthalene as the
internal standard, using response factors previously determined by the analysis of standard solutions; the quantitative analysis thus
performed allowed a reproducibility within ±1%.
3
. Results
3
.1. Oxidation of glycerol promoted by iron catalysts
The iron complex [Fe(BPA) (OTf) ] was employed as alcohol oxidation catalyst by Bauer et al. [16b], using either H O or t-
2
2
2 2
BuOOH as oxidant. Interestingly, other similar polydentate ligands used by the same Authors gave rise to iron compounds which
were either less active or inactive in promoting alcohol oxidation.
We chose to use the same iron species as catalyst for our initial tests on glycerol oxidation; hydrogen peroxide was selected as
oxidant, a mixture H
2 3
O/CH CN as reaction medium.
[
Fe(BPA) (OTf) ] proved to be an active catalyst for glycerol oxidation: after 90 min at r.t. an overall conversion of 46% was
2
2
measured, with formation of two products, i.e. DHA and HCOOH (selectivity of 53 and 47%, respectively, Table 1 entry 1).
Table 1
a
2 2
Oxidation of glycerol with H O promoted by iron catalysts
3
Page 3 of 9

b
Conv.
%)
DHA
Sel. (%)
HCOOH
Sel. (%)
Entry
catalyst
BPA(tot) /Fe
(
1
2
3
4
5
6
7
8
9
[Fe(BPA)
2
(OTf)
2
]
2
2
2
0
0
3
3
3
3
3
46
47
45
45
18
26
27
33
5
53
49
65
2
47
Fe(OTf)
Fe(OTf)
Fe(OTf)
Fe(OTf)
Fe(OTf)
Fe(OTf)
FeCl
Fe (CO)12+ BPA
[Fe(BPA) (OTf)
2
+ BPA
51
3
+ BPA
35
2
98
3
21
80
78
34
62
81
79
2
+ BPA
+ BPA
20
3
22
2
+ BPA
66
3
38
10
2
2
] + BPA
19
19
a
Experimental conditions: [Fe]=1.0x10-2 M; [glycerol]/[Fe]=35; [H
2
O
2
2 2 2
]/[glycerol] = 2.8; H O 30 % w/w; solvent MeCN:H O=2:1;
T=25 °C; t=90 min; b Calculated as % of glycerol reacted.
In other words, formation of the desired product DHA was accompanied by that of formic acid, one of the common glycerol
overoxidation products. Interestingly, in the final reaction mixture glyceraldehyde, which in the majority of glycerol oxidation
reactions is formed together with DHA, was present only in traces, and other possible overoxidation products were not detected.
When the same reaction was repeated by using a catalyst prepared “in situ” from Fe(OTf)
2
and 2 equivalents of BPA similar results
were obtained (Table 1 entry 2); on the other hand, substitution of the Fe(II) with the corresponding Fe(III) precursor, i.e. Fe(OTf)
3
,
also produced a mixture of DHA and formic acid, although with the Fe(III) precursor the selectivity in DHA was moderately
improved (Table 1 entry 3).
It was appropriate to verify the catalytic activity of the iron triflates in the absence of BPA: both Fe(OTf) and Fe(OTf) showed
2 3
some activity in glycerol oxidation, the Fe(II) species being more active than the Fe(III) salt, but both proved to be selective towards
formic acid formation (Table 1 entries 4 and 5) ; therefore, the ligand BPA appeared to play a key role in promoting the formation of
DHA.
Such hypothesis was confirmed by repeating the reactions performed with the catalysts prepared “in situ” in the presence of a higher
ligand/Fe ratio: comparison in Table 1 of entries 6 and 7 (relative to BPA/Fe = 3) with 2 and 3 (BPA/Fe = 2), respectively, shows an
improvement in DHA selectivity by increasing the BPA/Fe ratio, although at the expenses of the overall conversion.
2 3
Other iron salts were tested as catalyst precursors in association with BPA: FeCl was less active and selective, whereas Fe (CO)12
proved to be poorly catalytically active (Table 1 entries 8 and 9). In the absence of an iron compound, reaction of glycerol with H
produced only traces of the oxidation products.
2 2
O
2 2
Finally, the reaction in which the preformed catalyst [Fe(BPA) (OTf) ] was employed (Table 1 entry 1) was repeated in the presence
of added free BPA: the results (Table 1 entry 10) show a positive effect on the selectivity (81% in DHA) although accompanied by a
decrease in glycerol conversion; in fact, the composition of the final reaction mixture appears to be the same as that obtained using
the catalytic system prepared “in situ” with the same BPA/Fe overall ratio (compare in Table 1 entries 6 and 10).
3
.2. Optimization of experimental conditions for oxidation of glycerol to DHA
Our studies proceeded with an investigation on the effect of various parameters on activity and selectivity of the catalytic reactions.
Our goal was to maximize formation of the desired product DHA. For these reactions we employed the catalytic system prepared “in
situ” from Fe(OTf) and 3 equivalents of BPA; in all cases the only detected products were mixtures of DHA and formic acid, in
2
some cases together with traces of glyceraldehyde.
We first examined the effect on the catalytic reaction of molar substrate/catalyst ratio (see Table 2 entries 1, 2). Rather surprisingly,
the conversion slightly improved by increasing the substrate concentration, but on the other hand the same variation dramatically
affected the selectivity, which changed from a DHA/formic acid ratio of 4:1 to 1:3 with sub/Fe of 35 and 100, respectively.
On the other hand, an increase of iron loading (i.e. substrate/iron molar ratio 20 or lower) failed to produce higher reaction yields, as
both conversion and selectivity were very similar to those obtained at substrate/iron=35.
Also the choice of reaction temperature influenced both conversion and selectivity of glycerol oxidation (Table 1 entries 1, 3, 4). In
comparison to the standard reaction at r.t. (entry 1), an increase of temperature produced higher conversion but lower yield of DHA
(entry 4); on the contrary, when the reaction was performed in an ice bath the conversion was only slightly affected, whereas the
selectivity increased from 80 to 97% (entry 3).
With regard to the reaction medium, an acetonitrile/water mixture appeared to be a good choice; no significant effect was observed
by changing the relative amount of the two solvents, the best results were obtained with a CH CN/H O volume ratio of 2:1. Further
3 2
tests with other solvents were not attempted, on one hand because of the restricted choice due to limited glycerol solubility, on the
other because use of solvents classified as “undesirable” from the environmental point of view [17] was not considered.
2 2
Also the initial choice of H O as oxidant proved to be appropriate: t-BuOOH was also tested in our reactions, but it turned out to be
a worse oxidant for glycerol, especially from the point of view of overall conversion (compare entries 1 and 5 in Table 2).
Table 2
a
2 2 2
Oxidation of glycerol with H O catalyzed by Fe(OTf) + BPA
4
Page 4 of 9

b
DHA
Sel. (%)
HCOOH
Sel. (%)
Entry
T (°C)
t (min)
H
2
O
2
/glycerol
glycerol/Fe
H
2
O
2
% (w/w)
Conv. (%)
1
2
3
4
25
25
0
90
90
90
90
90
90
90
90
15
15
90
2.8
2.8
2.8
35
100
35
35
35
35
35
35
35
35
35
30
30
30
30
26
32
23
30
6
80
27
97
68
80
86
100
93
83
98
92
20
73
3
40
25
25
25
25
25
25
25
2.8
32
20
14
0
c
d
e
5
2.8
2.0
1.5
2.8
2.8
2.8
2.8
70
6
7
8
30
30
10
30
30
30
21
17
25
20
19
24
7
9
17
2
f
1
0
1
f
1
8
a
c
-2
O (2:1); b calculated as % of glycerol reacted;
Experimental conditions: [Fe] = 1.0x10 M; [BPA]/[Fe] = 3; solvent MeCN:H
2
reaction carried out with t-BuOOH as oxidant; d t-BuOOH/glycerol; t-BuOOH % (w/w); reactions carried out under an argon
e
f
atmosphere.
2 2
Other important parameters which are known to affect catalytic oxidations are hydrogen peroxide concentration and H O /substrate
ratio. We examined the effect of both factors on our catalytic reactions. The hydrogen peroxide/glycerol molar ratio was varied from
.8 to 2.0 and 1.5. The data reported in Table 2 (entries 1, 6, 7) show an evident effect on both reaction conversion and selectivity:
the former decreased and the latter increased by decreasing the excess of hydrogen peroxide employed: notably, at H /glycerol =
.5 formation of formic acid was completely suppressed, with a resulting selectivity in DHA of 100%.
Also the concentration of the hydrogen peroxide solution proved to significantly influence the catalytic reaction: by lowering the
concentration of aqueous H from 30% to 10% a positive effect was obtained on both conversion and selectivity, the latter raising
to a value of 93% in DHA (Table 2 entries 1 and 8). Since the efficiency of H as oxidant can be affected by the pH of the medium
14b,18], we performed the same reaction of Table 2 Entry 1, but replacing water with a buffer solution at pH=1: in the acidic
2
2
O
2
1
2
O
2
2
O
2
[
solution apparently a different catalyst was formed (pale yellow instead of deep purple solution) which proved to be poorly selective
towards DHA formation (about 65% of formic acid obtained).
With regard to reaction time, it proved useless to carry out the catalytic reactions at longer reaction times in order to increase the
conversion: as a matter of fact most of glycerol consumption occurred in the first 15 minutes, whereas at longer reaction times the
conversion increase was not substantial (compare entries 9 and 1 in Table 2).
Finally, it was of interest to examine the effect of performing the catalytic reaction with the exclusion of air, i.e. under an argon
atmosphere. The effect of oxygen on iron-catalyzed oxidations can be various, ranging from possible oxidation of the iron species to
concurrent catalytic reactions which make use of oxygen as oxidating agent. Interestingly, in our case the catalytic oxidation of
glycerol carried out under argon proved to be more selective than the corresponding reactions in air, with selectivities in DHA up to
9
8% (in Table 2, compare entries 9 and 1 with 10 and 11, respectively). Moreover, for the reactions performed in inert atmosphere
the trend of conversion vs. reaction time was the same as that observed in the presence of air (see Table 2 entries 10, 11).
At this point, the effect of the parameters on reaction conversion and selectivity was rather clear: in order to obtain high selectivities
in DHA the reactions must be carried out with low substrate/iron ratio, at room temperature or lower, using a moderate excess of
H O as 10% solution, preferably under an argon atmosphere. Various other catalytic reactions were then performed, with the aim of
2 2
increasing the conversion without significative loss in selectivity: unfortunately, all the attemped combinations of parameters,
although in most cases leading to very high selectivities, failed to raise the overall conversions.
3
.3. Attempted oxidation of a glycerol acetal
An alternative possible route for enhancing the selectivity of glycerol oxidation might make use of a glycerol acetal as starting
material. We recently reported a selective procedure for glycerol acetalization to five and six-membered cyclic acetals and ketals
[
19]: these molecules are potentially useful substrates for selective transformations of the single non-protected hydroxyl group.
Replacement of glycerol with one of these derivatives in oxidation reactions might in principle give a double advantage: on one
hand, no competition between primary and secondary OH groups would be possible; on the other, the acetal is expected to be less
subject to overoxidation and subsequent decomposition.
We therefore carried out a number of reactions using the six-membered cyclic acetal 2-phenyl-[1,3]-dioxan-5-ol (Figure 2) as starting
2
material and the catalyst prepared “in situ” from Fe(OTf) and three equivalents of BPA. Unfortunately, analysis of the final reaction
mixtures only revealed the presence of an isomeric mixture of five and six-membered acetals as well as the deacetalization products
glycerol and benzaldehyde, whereas DHA was only detected in traces. In fact, attempts in our previous work to substitute glycerol
with one of its acetals had been frustrated by either low reactivity of the substrate, or by its decomposition [20].
5
Page 5 of 9

Figure 2. cis, trans-1,3-O-benzylideneglycerol.
3
.4. ESI-MS data
In order to obtain some insight on the nature of the iron species formed in solution we carried out a series of ESI-MS measurements.
The spectra of an acetonitrile solution containing Fe(OTf) and 3 equivalents of BPA showed in positive ion mode signals at m/z 603,
2
+
+
+
4
2
04, 254 which were assigned to [Fe(BPA) (OTf)] , [Fe(BPA)(OTf)] and [Fe(BPA)] , respectively. The presence of free BPA
ligand was revealed by the peak at m/z 200 [M+H]. When the same solution was treated with aqueous glycerol solution reproducing
the same concentration and molar ratio used in the catalytic reactions, the MS spectra contained, aside the peaks formerly observed,
only few minor signals of very low intensity, none of which could be related to iron-glycerol adducts.
Other spectra of reaction mixtures (Table 1 entries 6 and 10, see Supplementary Information) were also performed in search for
+
signals related to either peroxo or oxygen containing adducts of the type [Fe(BPA)(glycerol)(OTf)] , however such peaks were never
observed. In fact, as the evolution of the catalytic reactions under investigation rapidly leads to formation of inactive species, the MS
spectra of reaction mixtures probably show signals mainly – if not only – related to inactive iron complexes.
4
. Discussion
The capability of iron of promoting alcohol – including glycerol – oxidation is no recent discovery. The first report on iron-catalyzed
glycerol oxidation by hydrogen peroxide dates back to 1899 [21]; other studies on alcohol and/or glycerol oxidation in the presence
of iron salts followed in the last century [22] and appeared also recently [23]. A common feature of the iron-promoted glycerol
oxidation so far reported was the formation of mixtures of oxidation products of doubtful usefulness.
Therefore, if on one hand the capability of iron to promote glycerol oxidation has long been known, on the other appropriate control
on the catalyst selectivity must be applied, in order to make it suitable for the selective production of glycerol oxigenates. In this
respect, the recent studies published on selective alcohol oxidation (see Introduction) represent a promising starting point, as they
demonstrate that iron atoms can be tamed into active and selective catalytic sites by association with the suitable ligands and reaction
conditions.
The present attractiveness of iron-based catalysts is due on one hand to low price and availability of this metal, on the other by the
increasing awareness of the scientific community of the environmental issues: from this point of view, the properties of iron in terms
of low toxicity and environvental safety make it an excellent choice for the development of green processes. Moreover, the interest
towards iron-based catalysts is enhanced by the desire of mimicking natural iron-containing enzymes [24].
The catalytic reactions here described fit well into our current interest for the development of catalytic processes in coherence with
the Green Chemistry principles [25]. All the experimental conditions employed are environmentally benign: catalyst, reaction
2 2
medium and temperature, oxidating agent (H O , which only gives water as byproduct). The catalytic reaction itself is highly atom-
economical, which favorably compares with some industrial processes for the production of glycerol oxigenates that make use of
stoichiometric reagents [5].
In the Results section we have described the application of the catalytic system iron/BPA to the oxidation of glycerol. A first
comment on the results is related to the reaction selectivity: of the various possible products only DHA and formic acid are formed,
in variable amounts which depend on the experimental parameters. Formic acid is a rather common product of glycerol oxidative
degradation [26]; in our case, as DHA is the glycerol oxidation product of highest commercial value, our efforts have been directed
towards its selective formation.
Apparently, the catalytic system under investigation is composed of at least two active species, each one promoting formation of one
of the two products. Preferred oxidation of the secondary hydroxyl group of glycerol leads to formation of DHA; it is reasonable to
assume that this reaction is promoted by an iron catalyst with at least one coordinated BPA ligand, as we can gather from a
comparison of the results obtained with and without added BPA, reported in Table 1. DHA thus formed is not subsequently oxidized
in significative amount, as confirmed by the results obtained in catalytic reactions in which glycerol was replaced by DHA: in
standard catalytic conditions only minor amounts of formic acid (<8% of the original DHA) were detected.
At variance, formic acid is probably formed via initial oxidation of a primary hydroxyl group of glycerol to give glyceraldehyde (See
Scheme 2), which is subsequently oxidized to formic acid (in fact, glyceraldehyde is known to be less stable to oxidation than DHA):
this reaction is promoted by a different catalyst, probably having no BPA coordinated to iron. As a matter of fact, formation of
formic acid is also observed when no BPA is added to Fe(II) or Fe(III) triflate. The latter reaction is disfavoured in the presence of
2 2
excess BPA, as well as at low temperature and low concentration of H O : that is the reason why in these experimental conditions the
selectivity in DHA is high. A beneficial effect on selectivity is also observed when oxygen is excluded from the reaction mixture:
these findings might be interpreted in terms of oxygen playing some role either in glyceraldehyde oxidation or in possible reactions
with the catalyst to give less selective iron species.
6
Page 6 of 9

Scheme 2. Production of DHA and formic acid via oxidation of secondary and primary hydroxyl groups, respectively.
With regard to the mechanism of the iron-catalyzed alcohol oxidation, previous studies proposed various mechanistic pathways [16a,
2
7]. There is a common agreement about initial reaction of the iron compound with the peroxide to give a species containing the Fe-
O-O-R mojety (R = H or alkyl): this iron complex is not catalytically active, but it can generate the active catalyst (or catalysts). It
has been proposed that the fate of the iron-peroxo intermediate can undergo at least two different decomposition routes, namely
homolytic and heterolytic reactions. The former route produces radical species which promote oxidation reactions with radical
mechanism, typically poorly selective. At variance, the heterolytic decomposition probably forms high valent iron-oxo species, the
catalytic activity and selectivity of which are strongly influenced by the nature of the other ligands coordinated to iron; the role of
high valent iron-oxo species in oxidation catalysis has been the object of both experimental and theoretical studies [13a, 28].
Evidence in favour of substrate coordination is uncommon for iron catalyzed alcohol oxidation [14b], on the contrary most studies
propose catalytic mechanisms which exclude alcohol coordination to the iron centre [16a, 29]. In the present case, ESI-MS spectra of
solutions containing the iron salt, BPA and glycerol did not provide evidence for coordination of glycerol to the metal; on the other
hand, in the MS spectra of catalytic mixtures no peaks could be assigned to iron-BPA oxo or peroxo complexes.
Finally, our findings relative to the nature of the iron precursor, i.e. the similar results obtained using Fe(II) and Fe(III) triflates, are
compatible with formation of Fe(IV) or Fe(V) oxo species, indipendently from the initial iron oxidation state. The worse catalytic
results observed when using FeCl
previously proposed [30].
2
as precursor can be explained in terms of formation of dimeric inactive or less active species, as
5
. Conclusions
The results here reported describe the properties of iron complexes with the polydentate ligand BPA as effective catalysts for the
oxidation of glycerol. In all the catalytic reactions only two products were detected, namely dihydroxyacetone and formic acid; such
products appear to be obtained by two different pathways, promoted by different catalytic species. By tuning the experimental
conditions, although conversions exceeding 50% were never obtained, complete suppression of the reaction leading to formic acid
was achieved: in these conditions DHA was formed with up to 100 % selectivity. To the best of our knowledge, this study represents
the first example of iron-catalyzed selective oxidation of glycerol to DHA; moreover, the catalytic reaction here described is a
remarkable example of green process, from the point of view of metal, oxidant, reaction medium as well as of the experimental
conditions employed.
Acknowledgements
The Authors thank Dr. Fabio Hollan for the ESI-MS spectra. Finantial support from the University of Trieste (FRA 2012) is
gratefully acknowledged.
Notes and references
[1] (a) O. R. Inderwildi, D. A. King, Energy Environ. Sci 2 (2009) 343-346; (b) R. Luque, L. Herrero-Davila, J. M. Campelo, J. H. Clark, J. M.
Hidalgo, D. Luna, J. M. Marinas, A. A. Romero, Energy Environ. Sci. 1 (2008) 542-564.
[2] (a) C.-H. Zhou, J. N. Beltramini, Y.-X. Fan, G. Q. Lu, Chem. Soc. Rev. 37 (2008) 527-549; (b) M. Pagliaro, R. Ciriminna, H. Kimura, M. Rossi,
C. Della Pina, Angew. Chem. Int. Ed. 46 (2007) 4434-4440; (c) A. Behr, J. Eilting, K. Irawadi, J. Leschinski, F. Lindner, Green Chem. 10 (2008)
13-30; (d) A. E. Diaz-Alvarez, J. Francos, B. Lastra-Barreira, P. Crochet, V. Cadierno, Chem. Commun. 11 (2011) 6208-6227; (e) N. H. Tran, G.
S. Kamali Kannangara, Chem. Soc. Rev. 42 (2013) 9454-9479.
[3] B. Katryniok, H. Kimura, E. Skrzynska, J.-S. Girardon, P. Fongarland, M. Capron, R. Ducoulombier, N. Mimura, S. Paul, F. Dumeignil, Green
Chem. 13 (2011) 1960-1979.
[
[
[
[
4] F. Jerome, Y. Poilloux, J. Barrault, ChemSusChem 1 (2008) 586-613.
5] M. Besson, P. Gallezot, C. Pinel, Chem. Rev. 114 (2014) 1827-1870 and references cited therein.
6] (a) E. Farnetti, J. Kašpar, C. Crotti, Green Chem. 11 (2009) 704-709; (b) C. Crotti, J. Kašpar, E. Farnetti, Green Chem. 12 (2010) 1295-1300.
7] (a) A. Azua, J. A. Mata, E. Peris, Organometallics 30 (2011) 5532-5536; (b) A. Azua, J. A. Mata, E. Peris, F. Lamaty, J. Martinez, E. Colacino,
Organometallics 31 (2012) 3911-3919.
[8] (a) R. M. Painter, D. M. Pearson and R. M. Waymouth, Angew. Chem. Int. Ed. 49 (2010) 9456-9459; (b) A. G. De Crisci, K. Chung, A. G.
Oliver, D. Solis-Ibarra, R. M. Waymouth, Organometallics 32 (2013) 2257-2266.
[9] S. Enthaler, K. Junge, M. Beller, Angew. Chem. Int. Ed. 47 (2008) 3317-3321.
[10](a) K. Junge, K. Schroeder, M. Beller, Chem. Commun. 47 (2011) 4849-4859 and references cited herein; (b) B. A. F. Le Bailly, S. P. Thomas,
RCS Advances 1 (2011) 1435-1445; (c) G. Wienhoefer, F. A. Westerhaus, R.V. Jagadeesh, K. Junge, H. Junge, M. Beller, Chem. Commun. 48
(2012) 4827-4829; (d) L.-Q. Lu, Y. Li, K. Junge, M. Beller, Angew. Chem. Int. Ed. 52 (2013) 8382-8386; (e) S. Fleischer, S. Zhou, K. Junge, M.
Beller, Angew. Chem. Int. Ed. 52 (2013) 5120-5124; (f) S. Warratz, L. Postigo, B. Royo, Organometallics 32 (2013) 893-897; (g) R. Langer, G.
Leitus, Y. Ben-David, D. Milstein, Angew. Chem. Int. Ed. 50 (2011) 2120-2124.
[
[
[
[
11](a) R. H. Morris, Chem. Soc. Rev. 38 (2009) 2282-2291 and references cited therein; (b) Y. Li, S. Lu, X. Wu, J. Xiao, W. Shen, Z. Dong, J. Gao,
J. Am. Chem. Soc. 136 (2014) 4031-4039.
12](a) M. R. Bukowski, P. Comba, A. Lienke, C. Limberg, C. Lopez de Laorden, R. Mas-Ballestre, M. Merz, L. Que Jr., Angew. Chem. Int. Ed. 45
(2006) 3446-3449; (b) O. Cussò, I. Garcia-Bosch, X. Ribas, J. Lloret-Fillol, M. Costas, , J. Am. Chem. Soc. 135 (2013) 14871-14878.
13](a) A.-R. Han, Y. J. Jeong, Y. Kang,J. Y. Lee, M. S. Seo, W. Nam, Chem. Comm. (2008) 1076-1078; (b) S. Chakraborty, H. Guan, Dalton
Trans. 39 (2010) 7427-7436; (c) A. A. Mikhailine, M. I. Mazharul, A. J. Lough, R. H. Morris, J. Am. Chem. Soc. 134 (2012) 12266-12280.
14](a) H. Hosseini-Monfared, C. Nather, H. Winkler, C. Janiak, Inorganica Chimica Acta 391 (2012) 75-82; (b) B. Biswas, A. Al-Hunaiti, M. T.
Raisanen, S. Ansalone, M. Leskela, T. Repo, Y.-T. Chen, H.-L. Tsai, A. D. Naik, A. P. Railliet, Y. Garcia, R. Ghosh, N. Kole, Eur.J. Inorg.
Chem. (2012) 4479-4485; (c) A. Bekish, Tetrahedron Letters 53 (2012) 3082-3085; (d) R. R. Fernandes, J. Lasri, M. F. C. Guendes da Silva, J.
A. L. da Silva, J. R. Frausto da Silva, A. J. L. Pombeiro, J. Mol. Catal. A: Chem 351 (2011) 100-111; (e) S. A. Moyer, T. W. Funk, Tetrahedron
7
Page 7 of 9

letters 51 (2010) 5430-5433; (f) A. Al-Hunaiti, T. Niemi, A. Sibaouih, P. Pihko, M. Leskela, T. Repo, Chem. Commun. 46 (2010) 9250-9252; (g)
M. Lenze, S. L. Sedinkin, E. B. Bauer, J. Mol. Cata. A: Chem. 373 (2013) 161-171.
[15]C. Bianchini, E. Farnetti, M. Graziani, M. Peruzzini, A. Polo, Organometallics 12 (1993) 3753-3761.
[16](a) M. Lenze, E. T. Martin, N. P. Rath, E. B. Bauer, ChemPlusChem 78 (2013) 101-116; (b) M. Lenze, E. B. Bauer, Chem. Commun. 49 (2013)
5889-5891.
[
[
[
[
[
[
[
[
[
[
17](a) R. A. Sheldon, Chem. Soc. Rev. 41 (2012) 1437-1451; (b) P. J. Dunn, Chem. Soc. Rev. 41 (2012) 1452-1461.
18]R. Noyori, M. Aoki, K. Sato, Chem. Commun. (2003) 1977-1986.
19]C. Crotti, E. Farnetti, N. Guidolin, Green Chem. 12 (2010) 2225-2231.
20]C. Crotti, E. Farnetti, S. Licen, P. Barbieri, G. Pitacco, J. Mol. Catal. A: Chemical 382 (2014) 64-70.
21]H. J. H. Fenton, H. Jackson, J. Chem. Soc. Trans. 75 (1899) 1-11.
22]e.g. see F. Haber, J. J. Weiss, Proc. R. Soc. London 147 (1934) 332-351.
23]e.g. see S. E. Martin, D. F. Suarez, Tetrahedron Letters 43 (2002) 4475-4479.
24]S. Enthaler, K. Junge, M. Beller, Angew. Chem. Int. Ed. 47 (2008) 3317-3321.
25](a) P. T. Anastas, M. M. Kirchhoff, Acc. Chem. Res. 35 (2002) 686-694; (b) P. T. Anastas, N. Eghbali, Chem. Soc. Rev. 39 (2010) 301-312.
26](a) Y. Kwon, Y. Birdja, I. Spanos, P. Rodriguez, M. T. M. Koper, ACS Catalysis 2 (2012) 759-764; (b) J. Xu, Y. Zhao, H. Xu, H. Zhang, B. Yu,
L. Hao, Z. Liu, Applied Catal. B: Environmental 154-155 (2014)267-273; (c) P. Pullanikat, J. H. Lee, K. S. Yoo, K. W. Yung, Tetrahedron
Letters 54 (2013) 4463-4466; (d) Y. Zhang, M. Zhang, Z. Shen, J. Zhou, X. Zhou, J. Chem. Tech. Biotechn., 88 (2013) 829-833; (e) E. G.
Rodrigues, M. F. R. Pereira, J. J. M. Orfao, Catalysis Commun. 25 (2012) 110.
[27](a) E. Balogh-Hergovich, G. Speier, J. Mol. Catal. A: Chem. 230 (2005) 79-83; (b) N. Y. Oh, Y. Suh, M. J. Park, M. S. Seo, J. Kim, W. Nam,
Angew. Chem. Int. Ed. 44 (2005) 4235-4239.
[
28](a) J. T. Groves, R. C. Haushalter, M. Nakamura, T. E. Nemo, B. J. Evans, J. Am Chem. Soc. 103 (1981) 2884- 2886; (b) W. J. Song, Y. O. Ryu,
R. Song, W. Nam, J. Biol. Inorg. Chem. 10 (2005) 294-304; (c) S. Shaik, H. Hirao, D. Kumar, Acc. Chem. Res. 40 (2007) 532-542.
29]C. Fabbri, C. Aurisicchio, O. Lanzalunga, Centr. Eur. J. Chem. 6 (2008) 145-153.
[
[
30](a) C. J. Davies, G. A. Solan, J. Fawcett, Polyhedron 23 (2004) 3105-3114; (b) K. Visvaganesan, E. Suresh, M. Palaniandavar, Dalton Trans.
(2009) 3814-3823.
Figure Caption list
Scheme 1. Products obtained by glycerol oxidation.
Figure 1. bis(2-pyridinylmethyl)amine (BPA).
Figure 2. cis, trans-1,3-O-benzylideneglycerol.
Scheme 2. Production of DHA and formic acid via oxidation of secondary and primary hydroxyl groups, respectively.
8
Page 8 of 9

Selective oxidation of glycerol catalyzed by iron complexes
Corrado Crotti and Erica Farnetti
9
Page 9 of 9