Contribution to elucidation of the mechanism of preparation of 2-isobutyl-4-methyltetrahydro-2H-pyran-4-ol

Article abstract of DOI:10.1007/s11164-015-2052-z

The effects of catalyst type (sulfuric acid, Amberlyst 15, p-toluenesulfonic acid, and p-dodecylbenzenesulfonic acid) and of addition of water on the preparation of 2-isobutyl-4-methyltetrahydro-2H-pyran-4-ol by Prins cyclization were studied. It was discovered that in the absence of added water the amphiphilic character of the acid had the main effect and that organic acids were unsuitable for preparation of this compound. From the perspective of the ratio of the amount of the desired product to that of dihydropyrans, after addition of water the results obtained by use of p-toluenesulfonic acid (ratio 3.5, selectivity 67 %) were comparable with those obtained by use of sulfuric acid and were better than those obtained by use of Amberlyst 15 (ratio 2.3, selectivity 68 %). A detailed study of the mechanism confirmed that addition of water to the double bonds of dihydropyrans does not occur.

Full text of DOI:10.1007/s11164-015-2052-z

Res Chem Intermed
DOI 10.1007/s11164-015-2052-z
Contribution to elucidation of the mechanism
of preparation of 2-isobutyl-4-methyltetrahydro-
2H-pyran-4-ol
1
Eliska Vyskocilova Lucie Rezkova
1
•
•
ˇ
ˇ
´
´
Eva Vrbkova Iva Paterova Libor Cerveny
1
1
1
ˇ
•
•
´
´
´
Received: 29 October 2014 / Accepted: 7 April 2015
Ó Springer Science+Business Media Dordrecht 2015
Abstract The effects of catalyst type (sulfuric acid, Amberlyst 15, p-toluenesul-
fonic acid, and p-dodecylbenzenesulfonic acid) and of addition of water on the
preparation of 2-isobutyl-4-methyltetrahydro-2H-pyran-4-ol by Prins cyclization
were studied. It was discovered that in the absence of added water the amphiphilic
character of the acid had the main effect and that organic acids were unsuitable for
preparation of this compound. From the perspective of the ratio of the amount of the
desired product to that of dihydropyrans, after addition of water the results obtained
by use of p-toluenesulfonic acid (ratio 3.5, selectivity 67 %) were comparable with
those obtained by use of sulfuric acid and were better than those obtained by use of
Amberlyst 15 (ratio 2.3, selectivity 68 %). A detailed study of the mechanism
confirmed that addition of water to the double bonds of dihydropyrans does not
occur.
Keywords Prins cyclization Á Water addition Á Florosa Á Acid catalysis
Introduction
Florosa or Florol (2-isobutyl-4-methyltetrahydro-2H-pyran-4-ol, Fig. 1, VI) is an
important compound in the fragrance industry. It can be prepared by Prins
condensation, a cyclization reaction between an aldehyde and an unsaturated
alcohol [1, 2]. Prins cyclization is a very efficient reaction for production of
substituted tetrahydropyranols or the dihydropyran ring. These two types of product
may both be formed during the reaction and the selectivity depends on the reaction
ˇ
ˇ
´
Eliska Vyskocilova
eliska.vyskocilova@vscht.cz
&
1
Department of Organic Technology, University of Chemistry and Technology Prague,
Technicka 5, Prague 6 166 28, Czech Republic
123

ˇ ´
E. Vyskocilova et al.
O
OH
HO
O
+
+
O
IV
I
III
II
OH
+
O
O
VI
V
O
VII
Fig. 1 Mechanism of the preparation of Florosa (I, isovaleraldehyde; II, isoprenol; III, hemiacetal; IV,
carboxonium ion; V, carbocation; VI, 2-isobutyl-4-methyltetrahydro-2H-pyran-4-ol; VII, 2-isobutyl-4-
methyldihydropyran)
conditions, especially the type of catalyst [3]. Lewis [4–7] or Brønsted acids
(inorganic [1, 8], organic [9], or ion exchangers [10, 11]) may be used as catalysts;
occasionally, heteropolyacids [12], organometallic complexes [13] or heterogeneous
aluminosilicates [14] are used. The accepted mechanism [15] involves hemiacetal
formation in the first step followed by cyclization to form the substituted
tetrahydropyranol or dihydropyrans (Fig. 1).
It is obvious that the presence of water can affect the composition of the final
reaction mixture. This fact is used when dihydropyrans are the desired product. The
water formed by dehydration of the substituted tetrahydropyranol is removed, e.g.,
by azeotropic distillation [9] with a solvent (e.g. toluene). Selective preparation of
dihydropyrans with sulfuric acid (75 % concentration) as catalyst has also been
described [15]. Formation of dihydropyrans is also favored by use of high
temperatures [1] compared with the temperature used for substituted tetrahydropy-
ranol formation. Sometimes (e.g., condensation of benzaldehyde with isoprenol) it
is possible to use a strong acid catalyst to selectively prepare the substituted
tetrahydropyranol [1, 15]. Substituted tetrahydropyrans may also be prepared
selectively by use of strong nucleophiles [2, 6, 7, 13]. If the desired product is a
substituted tetrahydropyranol (in this work Florosa), the literature offers only
information about changes of the reaction conditions (lower temperature) or
addition of water. The latter is mentioned only for preparation of substituted
tetrahydropyranols with Amberlyst 15 as a catalyst [11]; the authors recommended
addition of either 10 mol% water to the reaction mixture or the amount of water that
is soluble in the reaction mixture.
In this paper we report comparison of the use of different Brønsted acid catalysts
for preparation of Florosa, and the effect of the addition of water to the different
reaction mixtures.
123

Contribution to the mechanism of preparation of…
Experimental
Materials
Isoprenol, isovaleraldehyde, p-dodecylbenzenesulfonic acid, and Amberlyst 15 (wet,
15 wt% water) were purchased from Sigma–Aldrich. Sulfuric acid and p-toluenesulfonic
acid were purchased from ChemapolCZ. All materials were used as obtainedwithout any
purification. Water purified by osmosis (\3 mS m-1) was obtained from UCT storage.
Prins condensation
In a typical reaction, equimolar amounts of isovaleraldehyde (2.15 g; 0.025 mol)
and isoprenol (2.15 g; 0.025 mol) were placed in a flask equipped with a condenser.
Water and the catalyst (for heterogeneous catalysts, 2.5 wt% calculated on the basis
of the amount of isoprenol; for homogeneous catalysts 1 mol%; for sulfuric acid
0.5 mol%) were added to the reactants and the reaction mixture was immediately
placed in a preheated (typically 70 °C) oil bath and stirred. Samples taken from the
homogeneous reaction mixture were neutralized by addition of sodium carbonate,
diluted with toluene, centrifuged, and analyzed by GC (Shimadzu 2010; non-polar
ZB-1 column (60 m, 0.25 mm, 0.25 lm)). Mesitylene was used as internal standard.
Results and discussion
The effect of addition of water in the preparation of Florosa was tested for different
catalysts with the same catalytically active group, –SO3H. The catalysts were a
mineral acid (sulfuric acid), organic acids (p-toluenesulfonic acid and dodecylben-
zenesulfonic acid), and a solid catalyst with sulfonic groups as the catalytically
active groups (Amberlyst 15). Catalysts were chosen on the basis of their different
acidity (all results are for aqueous solutions at 25 °C: sulfuric acid pKa = -3
(-1.9) [16], p-toluenesulfonic acid pKa = -2.8 [17], p-dodecylbenzenesulfonic
OH
OSO2R
+
+ H2O
-OSO2R
+
O
O
O
VIII
VI
V
-HSO2R
O
VII
Fig. 2 Possible action of the catalyst in the reaction. R = –OH, –C6H5, –C18H29, or styrene
divinylbenzene copolymer (V, carbocation; VI, 2-isobutyl-4-methyltetrahydro-2H-pyran-4-ol; VII,
2-isobutyl-4-methyldihydropyran; VIII, intermediate formed from carbocation and catalyst)
123

ˇ ´
E. Vyskocilova et al.
acid pKa = -1.8 [18], Amberlyst 15 acidity = 4.7 mequiv H?/g on the basis of
supplier’s data, pKa \ 1 [19]) and on the basis of their solubility in water and
organic media. p-Dodecylbenzenesulfonic acid may act as a phase-transfer catalyst.
Using these catalysts, different amounts of water were added to the reaction
mixture and the selectivity at 70 % conversion, the ratio of substituted tetrahy-
dropyranol to the sum of dihydropyrans (F/DHP), and TOF were calculated. F/DHP
was calculated because of the problematic separation of these two substances from
the final reaction mixture and also to specify the amount of Florosa in the reaction
mixture.
On the basis of the postulated mechanism [20, 21], addition of the catalysts on
the basis of sulfuric acid was proposed. Formation of an intermediate (Fig. 2, VIII)
from the conjugate base and the carbocation intermediate can be predicted. In the
presence of water, a byproduct of formation of the carboxonium intermediate
(Fig. 1, IV), hydrolysis of intermediate VIII and formation of substituted
tetrahydropyranol occur. If water is separated from the reaction mixture by
azeotropic distillation or by any other means the catalyst is regenerated by
acquisition of a proton from the pyran ring and dihydropyran products are formed.
The results obtained are summarized in Table 1.
Comparison of acids without addition of water to the reaction mixture
From Table 1 (rows 1, 5, 11, 17) it is clear that the composition of the reaction
products and the rate of the reaction depend on the type of catalyst. When
concentrated sulfuric acid was used (no addition of water to the reaction mixture;
results not included in Table 1), polymerization of isoprenol and formation of
products with higher boiling points occurred. The products with higher boiling
points were diacetal, formed by reaction of isovaleraldehyde with two molecules of
isoprenol, and another compound with the same molecular weight (240 g/mol; the
exact structure was not determined).
It should be mentioned that it is difficult to compare p-toluenesulfonic acid and p-
dodecylbenzenesulfonic acid catalysts with sulfuric acid and Amberlyst 15 under
water-free conditions, because the last two are always accompanied by water. This
was probably one reason why the rates of the reaction and the selectivity were
substantially lower when organic acids were used. The low miscibility and lowest
acidity of p-dodecylbenzenesulfonic acid, especially, made it a poor catalyst that
was unable to trap the carboxonium ion (Fig. 1, IV) while forming the desired
substituted tetrahydropyranol.
Moreover, the organic acids used are amphiphilic in character, which may be
associated with problematic transport of reactants to the active sulfonic center and
with trapping of water originating during formation of the hemiacetal (Fig. 1, III).
Both of these phenomena negatively affected reaction rate and selectivity.
As mentioned above, the greater selectivity for sulfuric acid and Amberlyst 15
was a result of two factors:
1
2
these compounds were not amphiphilic in character; and
water was always present in the reaction mixture.
123

Contribution to the mechanism of preparation of…
Table 1 Effect of amount of water and type of catalyst on selectivity and rate of the reaction
Entry
Catalyst
Amount of
watera (mol%)
Selectivity
(%)
TOF
(min-1
Ratio
F/DHP
)
1
Sulfuric acid (12 %)c
Sulfuric acid (10 %)c
Sulfuric acid (2.5 %)c
Sulfuric acid (1 %)c
5
10
50
100
0
55
52
68
63
33
59
61
61
67
62
26
34
46
56
60
49
62
67
68
63
61
56
13.3
11.8
10.6
6.1
3.6
7.9
7.2
6.8
5.6
1.2
2.1
5.1
5.3
4.8
4.3
0.4
2.0
1.9
3.2
3.2
0.8
1.1
2.0
2.6
3.5
2.4
0.9
1.0
1.7
2.5
2.7
1.7
1.6
2.2
2.3
2.3
1.6
1.3
2
3
4
5
p-Toluenesulfonic acid
p-Toluenesulfonic acid
p-Toluenesulfonic acid
p-Toluenesulfonic acid
p-Toluenesulfonic acid
p-Toluenesulfonic acid
p-Dodecylbenzenesulfonic acid
p-Dodecylbenzenesulfonic acid
p-Dodecylbenzenesulfonic acid
p-Dodecylbenzenesulfonic acid
p-Dodecylbenzenesulfonic acid
p-Dodecylbenzenesulfonic acid
Amberlyst 15
6
5
7
10
20
50
100
0
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
a
5
10
20
50
100
0
2.3 9 10-3b
2.3 9 10-3b
2.4 9 10-3b
1.8 9 10-3b
8.3 9 10-4b
3.3 9 10-4b
Amberlyst 15
5
Amberlyst 15
10
20
50
100
Amberlyst 15
Amberlyst 15
Amberlyst 15
Calculated on total amount of reaction mixture
b
c
Rate of reaction using solid catalyst calculated in mmol/mg min
Concentration of sulfuric acid in water
Lower selectivity when sulfuric acid, rather than Amberlyst, was used was
explained by the formation of products with higher boiling points (especially
diacetal) and almost no formation of this type of undesirable product when solid
Amberlyst 15 was used. The F/DHP ratio was higher when sulfuric acid was used,
however, probably because of the greater amount of water in the reaction mixture.
Effect of addition of water
Severe phase-transfer problems were apparent when Amberlyst 15 was used.
Decreases in the rate of reaction and selectivity were observed (Table 1, rows
17–22) when the amount of water was increased from 10 mol%. This was because
of coating of the catalyst surface with water and by problematic transfer of organic
compounds to the active centers on the surface. Agglomeration of the catalyst in the
aqueous phase of the reaction mixture was also apparent when the amount of water
123

ˇ ´
E. Vyskocilova et al.
added exceeded 10 mol%. The decrease of selectivity when this catalyst was used
corresponds to the decrease of the F/DHP ratio. In comparison with the maximum
F/DHP values obtained for the other catalysts, that for Amberlyst 15 (2.3) was
lowest (3.5 was observed for p-toluenesulfonic acid and 3.2 for sulfuric acid).
The amount of sulfuric acid, calculated on the basis of the amounts of reactants,
was the same for all the experiments, only the amount of water was different (and
therefore, also, the concentration of the acid, as is apparent from Table 1, rows 1–4).
The decrease in the rate of the reaction in this instance was caused by dilution of the
acid by the water and by transfer of the sulfuric acid in the water to the organic
phase. The selectivity for substituted tetrahydropyranol increased as the amount of
water was increased to 50 mol%. When 100 mol% water was used a slight decrease
of selectivity was observed, because of increased formation of products with higher
boiling points.
After addition of water to reaction mixtures in which organic acids were used as
catalysts, substantial increases of the rate of the reaction and selectivity were
observed (increases from 33 and 26 % to 67 and 60 % for p-toluenesulfonic acid
and p-dodecylbenzenesulfonic acid, respectively). When p-toluenesulfonic acid was
used (Table 1, rows 5–10) the highest rate of the reaction was observed for use of
5 mol% water. The same trend as when sulfuric acid was used was observed with
increasing amounts of water in the reaction mixture, i.e., a decrease in the rate of the
reaction. The difference between sulfuric and p-toluenesulfonic acid was probably
because of the different solubility of these compounds in water and the amphiphilic
character of the latter, because their acidity was comparable. Selectivity for the
desired compound increased on addition of water, and the highest value was
observed when 50 mol% water was used. After addition of 100 mol% water a
decrease in selectivity was observed, because of the decrease of the F/DHP ratio.
When p-dodecylbenzenesulfonic acid was used, one interesting observation was
that addition of water gradually increased selectivity but, again, only up to
50 mol%. It is possible that at this concentration water makes the mixture
homogeneous and that this catalyst also partially serves as a phase-transfer catalyst.
When this catalyst was used, the highest rate of the reaction was observed after
addition of 10 mol% of water. The rate of the reaction decreased with increasing
amounts of water, and when 100 mol% of water was added the rate of the reaction
was even lower than when no water was added. Addition of 100 mol% water had a
detrimental effect on both the rate of the reaction and on selectivity, possibly
because micelle formation made the acid head group unavailable for catalysis. The
different rates of reaction for p-toluenesulfonic acid and p-dodecylbenzenesulfonic
acid were because of the lower acidity and significant amphiphilic character of the
latter.
In the presence of water the catalysts may be ordered according to decreasing
selectivity as follows sulfuric acid & Amberlyst 15 & p-toluenesulfonic acid [ p-
dodecylbenzenesulfonic acid and according to F/DHP ratio as follows p-toluene-
sulfonic acid & sulfuric acid [ p-dodecylbenzenesulfonic acid [ Amberlyst 15.
This means that use of Amberlyst 15 resulted in the smallest amount of products
with higher boiling points. The largest amount of this type of byproduct was
obtained when sulfuric acid was used.
123

Contribution to the mechanism of preparation of…
To explain the positive shift of the yield of the desired substituted tetrahydropy-
ranol and, in the presence of water, suppression of dehydration of substituted
tetrahydropyranol to undesired dihydropyrans, it can be assumed the water served as
a hydrolyzing agent (Fig. 2, VIII).
Calculation of rate constants
The purpose of the work discussed in this paper was to discover how the presence of
water affected the composition of the products of the reaction. Calculation of
power-law kinetics by use of the software ERA 2.0 led to a possible reaction
mechanism (Fig. 3).
Although the intermediate formed by the catalyst (compound 7 in Fig. 3) was not
detected in the reaction mixture, the substituted tetrahydropyranol and dihydropy-
rans are formed in this step, and undesired products with a higher boiling point (6,
Fig. 3) are probably formed directly from the hemiacetal (3). Because the
intermediate was not observed in the reaction mixture, calculation of rate constants
for formation of substituted tetrahydropyranol and of dihydropyrans was not
possible. The ‘‘selectivity’’ s3 of the reaction rate was introduced into the calculation
and, hence, the rates of formation of compounds (4) and (5) were determined as
follows (Eqs. 1 and 2):
r3Ã ¼ k3Ã Â c3 ¼ k2 Â s3 Â c3
r4Ã ¼ k4Ã Â c3 ¼ k2 Â ð1 À s3Þ Â c3
ð1Þ
ð2Þ
where r3* and r4* and k3* and k4* are the apparent reaction rates and rate constants
for formation of compounds 4 and 5, respectively, k2 is the rate constant for for-
mation of intermediate 7, s3 is the selectivity for formation of product 4, and c3 is
the concentration of the intermediate hemiacetal 3. Formation of products with
higher boiling points was determined by the concentration of hemiacetal (3) and by
the concentration of isovaleraldehyde, because during the course of the reaction
isovaleraldehyde was consumed quickly and the amount of acetal was lower than
the amount of the unidentified product of molar weight of 240 g/mol.
OH
OSO2R
*
O
OH
r2
r3
HO
r1
r6
O
+
O
O
3
4
1
2
*
r5
7
r4
higher boiling products
6
O
5
Fig. 3 Suggested mechanism for calculation of rate constants
123

ˇ ´
E. Vyskocilova et al.
Table 2 Examples of calculated rate constants (power-law kinetics)
Entry
Acid/amount
of water
(mol%)
k1
k2
(min-1
s3 (–)
k5 9 10-4
k6
(mmol-1
)
(min-1
)
(mmol-1
min-1
)
min-1
)
1
2
3
4
5
6
7
8
9
10
pTS/0
0.93
5.33
3.17
1.28
0.76
1.41
3.61
3.32
2.77
0.45
0.113
0.829
0.442
0.162
0.095
0.15
0.385
0.588
0.632
0.739
0.605
0.712
0.602
0.636
0.727
0.769
0
10.3
9.2
pTS/5
1
pTS/10
1
4.12
2.12
4.03
1.48
0
pTS/50
0
pTS/100
pDBS/50
Amb/0
0
0
0.30
8.4
3.2
2.6
0
Sulfuric/10
Sulfuric/50
Sulfuric/100
0.858
1.89
10.3
2.2
0.04
0.36
From the calculations (Table 2) it is obvious for homogeneous catalysts that with
increasing amounts of water the overall rate of the reaction decreases (rows 1–5 and
8–10). The rate of dehydration described by r5 was almost 0, except for Amberlyst
15 as catalyst without addition of water and for sulfuric acid, with which
dehydration was expected (rows 7 and 8). The value of s3 monitoring the rate of
production of substituted tetrahydropyranol also increased with increasing amount
of water up to 100 % mol excess water, when dilution of the acid had an effect. The
rate of formation of products with higher boiling points also decreased with
increasing amount of water.
Conclusions
Preparation of 2-isobutyl-4-methyltetrahydro-2H-pyran-4-ol, also called Florosa,
was discussed. Acid-catalyzed Prins condensation of isoprenol and isovaleraldehyde
was used for the reaction. The study was focused especially on the effect of catalyst
type and of addition of water to the reaction mixture. It was discovered that the
character of the catalyst, with an active sulfuric or sulfonic acid group, has a
significant effect on the composition of the reaction mixture and it is difficult to
compare sulfuric acid and Amberlyst 15 with p-toluenesulfonic and p-dodecylben-
zenesulfonic acids because the first two are always accompanied by water. Addition
of water also had a significant effect, which was most apparent for the catalyst with
amphiphilic character—p-dodecylbenzenesulfonic acid. With this catalyst and
addition of 100 mol% water, possible micelle formation made the acid head group
unavailable for catalysis.
The highest selectivity (68 %) was obtained by use of Amberlyst 15 as catalyst
with addition of 5–10 mol% water, or by use of sulfuric acid (2.5 % concentration
in water) and p-toluenesulfonic acid as catalysts, both with addition of 50 mol%
water. The highest F/DHP ratio (3.5) for the desired substituted tetrahydropyranol
was obtained by use of p-toluenesulfonic acid as catalyst with addition of 50 mol%
123

Contribution to the mechanism of preparation of…
water. The low F/DHP ratio (2.3) at comparable selectivity for Amberlyst 15 can be
explained by the presence of a small amount of products with higher boiling points
compared with use of the other catalysts. It was confirmed that water has a positive
effect on the reaction, especially when homogeneous catalysis is used, but the water
does not interact with dihydropyrans during the formation of Florosa.
References
1. P.H. William, S.A. Ballard, US 2 422 648 (1947)
2. C. Olier, M. Kaafarani, S. Gastaldi, M.P. Bertrand, Tetrahedron 66, 413 (2010)
´
3. M. Breugst, R. Gree, K.N. Houk, J. Org. Chem. 78, 9892 (2013)
4. F.K. Chio, J. Warne, D. Gough, M. Penny, S. Green, S.J. Coles, M.B. Hursthouse, P. Jones, Te-
trahedron 67, 5107 (2011)
5. P. Borkar, P. Weghe, B.V.S. Reddy, J.S. Yadav, R. Gree, Chem. Commun. 48, 9316 (2012)
´
6. K. Yheng, X. Liu, S. Qin, M. Xie, L. Lin, Ch. Hu, X. Feng, J. Am. Chem. Soc. 134, 17564 (2012)
7. K.R.K.K. Reddy, I.M.L. Rosa, A.C. Doriguetto, E.L. Bastos, L.F. Silva Jr, Molecules 18, 11100
(2013)
8. J.S. Yadav, B.V.S. Reddy, M.S. Reddy, N. Niranjan, J. Mol. Catal. A 210 (1–2), 99 (2004)
9. A. Umeda, Y. Ataka, S. Tanaka, K. Naito, K. Mine, EP 149 3737 (2005)
10. G.P. More, M. Rane, S.V. Bhat, Green Chem. Lett. Rev. 5, 13 (2012)
11. G. Gralla, K. Beck, M. Klos, U. Griesbach, US 20110306779 (2011)
12. G. Li, Y. Gu, Y. Ding, Z. Yong, W. Hanpeng, G. Jianming, Y. Qiang, S.J. Liang, J. Mol. Catal. A
218(2), 147 (2004)
13. K. Tadpetch, S. Rychnovsky, Org. Lett. 10, 4839 (2008)
´
14. S. Telalovic, A. Ramanthan, J. Fei Ng, R. Maheswari, C. Kwakernaak, F. Soulimani, H.C. Brouwer,
G.K. Chuah, B.M. Weckhuzsen, U. Hanefeld, Chem. Eur. J. 17, 2077 (2011)
15. B.J.F. Hudson, G. Schmerlaib, Tetrahedron 1, 284 (1957)
16. D.J. Evans, pKa Table, http://evans.harvard.edu/pdf/evans_pKa_table.pdf. Accessed 9 Aug 2014
17. J.P. Guthrie, Can. J. Chem. 56, 2342 (1978)
18. D.S. Wishart, T. Jewison, A.C. Guo, M. Wilson, C. Knox, et al., HMDB 3.0—The Human Meta-
bolome Database in 2013. Nucl. Acids Res. 41(D1):D801–D807 (2013). http://www.hmdb.ca/
metabolites/HMDB59915. Accessed 9 Aug 2014
19. G.H. Sankey, US 5 449 772 (1995)
20. C.S. Barry, N. Bushby, J.R. Harding, R.A. Hughes, G.D. Parker, R. Roe, C.L. Willis, Chem.
Commun. 29, 3727 (2005)
21. R.W. Alder, F. Carta, C.A. Reed, I. Stoyanova, C.L. Willis, Org. Biomol. Chem. 8, 1551 (2010)
123