A new autocatalytic thioacetate-enal addition reaction: A Michael addition or not?

Article abstract of DOI:10.1002/adsc.201000391

Rather than proceeding through a Michael-type or 1,4-addition, thioacetic acid adds across unsaturated aldehydes in an autocatalytic manner and involving a double exotherm, as demonstrated by both adiabatic and reaction calorimetry. NMR studies show that an intermediate acyl-thio-hemiacetal is involved and that the product continues to react competitively with thioacetic acid.

Full text of DOI:10.1002/adsc.201000391

COMMUNICATIONS
DOI: 10.1002/adsc.201000391
A New Autocatalytic Thioacetate-Enal Addition Reaction:
A Michael Addition or Not?
Gennadiy Ilyashenko,a Andrew Whiting,a,* and Allen Wrightb
a
Department of Chemistry, Durham University, Science Laboratories, South Road, Durham DH1 3LE, U.K.
Fax : (+44)-(0)191-384-4737; phone: (+44)-(0)191-334-2081; e-mail: andy.whiting@durham.ac.uk
Department of Chemical Engineering, Newcastle University, Newcastle upon Tyne, NE1 7RU, U.K.
b
Received: January 21, 2010; Published online: August 16, 2010
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/adsc.201000391.
Abstract: Rather than proceeding through a Mi-
chael-type or 1,4-addition, thioacetic acid adds
across unsaturated aldehydes in an autocatalytic
manner and involving a double exotherm, as dem-
onstrated by both adiabatic and reaction calorime-
try. NMR studies show that an intermediate acyl-
ACHTUNGTRENNUNGthio-hemiacetal is involved and that the product
continues to react competitively with thioacetic
acid.
terpretation of this, and perhaps other hetero-Michael
additions, may be required.
Initially, adiabatic calorimetry[13] was employed for
monitoring the progress[14,15] of the reaction shown in
Eq. (1) using hexenal 2a in order to correlate with the
expected mechanism,[16,17] and to carry out process
modelling and kinetic fitting[18] as a prelude to devel-
oping an asymmetric catalytic entry to the b-thioalde-
hyde 3a[19] which is used as a fragrance/flavour com-
Keywords: aldehydes; autocatalysis; Michael addi-
tion; sulfur
The Michael addition is one of the oldest[1,2] and most pound.[10]
widely used organic reactions.[3–5] The accepted mech-
Adiabatic calorimetry, commonly used for thermo-
anism involves a direct addition of a nucleophile to chemical safety studies,[13] was applied using a simple
the remote carbon of the conjugated C=O, or related Dewar flask and a PC-data high-frequency logging
system[6,7] and this view has changed little since its dis- thermocouple to obtain adiabatic temperature profiles
covery. The utility of the Michael addition is due to (Figure 1) for the reaction of 1 with 2a in different
its broad applicability in terms of both electrophile solvents [Eq. (1)]. Unexpectedly, for several solvents,
and nucleophile, with hetero-Michael reactions[8,9] the reaction clearly exhibited a two-stage temperature
being heavily investigated and exploited for accessing rise, indicating
a
two-stage reaction process
b-oxy-, aza- and thiocarbonyl compounds from unsa- (Figure 1). An adiabatic temperature rise normally
turated carbonyl derivatives. In particular, the hetero- exhibits self-acceleration due to positive feedback of
Michael addition reaction between thioacetic acid 1 temperature with Arrhenius kinetics. The presence of
and trans-2-hexenal 2a to derive adduct 3a [Eq. (1)] is a distinct second adiabatic temperature rise is usually
of interest because it occurs under both general acid- caused by the primary reaction raising the system
and base-catalysed conditions,[10,11] however, it has yet temperature to a sufficient level that a secondary re-
to succumb to asymmetric catalysis.[12] Herein, we action is initiated. However, closer examination of
report that the formal Michael addition outlined in some of the experiments revealed that this was not
Eq. (1) does not proceed as a simple bimolecular ad- the case.
dition reaction as might be expected based on litera-
In all the Dewar experiments (Figure 1) there was a
ture precedents.[6,7] Indeed, the reaction is multi-stage, small deviation from ideal adiabatic reactor behaviour
involves at least two intermediates and is autocatalyt- due to heat loss, with gradual cooling of the system as
ic, hence, these unexpected results suggest that a rein- the reaction proceeded to completion. As a result of
this cooling, the temperature derivative, dT/dt, is re-
1818
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2010, 352, 1818 – 1825

A New Autocatalytic Thioacetate-Enal Addition Reaction: A Michael Addition or Not?
Figure 1. Temperature versus time for the reaction shown in Eq. (1) in different solvents (Note: DMSO was omitted because
of safety concerns[20]).
duced between the first and second transients. With conversion, qt, calculated as qt=(heat evolved up to
higher levels of heat loss the derivative tends to zero time t)/(total heat evolved to reaction completion)
or even becomes negative (see Supporting Informa- [Eq. (2)].
tion for the example of this effect in the case of tolu-
ene). This is an important observation: if dT/dtꢀ0,
positive feedback of temperature to produce self-ac-
celeration cannot occur, hence, the second tempera-
ture rise is not an onset Arrhenius effect. This self-ac-
celerating reaction suggests that the inherent reaction
mechanism is in some way autocatalytic.
In order to demonstrate this effect more clearly, ad-
ditional isothermal experiments were carried out
The fact that the chemical conversion curves for
using an HEL SIMULARTM reaction calorimeter hexenal 2a do not match the overall reaction thermal
using power compensation[20] to measure the reaction conversion is not surprising and supports the view
exotherm, Qr (W). Precise temperature control was that more than one reaction is taking place, each with
achieved using an external cooling system together a different heat of reaction.
with regulated electrical heating inside the vessel.
These mergers of the double peaks of the Qr signal
Changes in the heater base load (power compensa- for the first addition into a single immediate exo-
tion) were monitored to obtain a direct measurement therm following the second addition strongly supports
of Qr. The reaction mixture was also sampled during the concept of autocatalytic behaviour as suggested
these runs and analysed with respect to hexenal. by the adiabatic calorimetry. The challenge, therefore,
When thioacetic acid 1 was first added to a solution was to elucidate a suitable reaction mechanism that
of trans-2-hexenal 2a in toluene the reaction produced could account for this observed behaviour.
a double exotherm. Once this reaction was complet-
In order to provide an insight into the possible re-
ed, additional trans-2-hexenal 2a and thioacetic acid action pathways and to explain the observations
were added. This time, only a single exotherm was ob- shown in Figure 1 and Figure 2, dynamic (time re-
served (Figure 2, a). For the second addition, only solved) NMR spectroscopy experiments were under-
two-thirds the quantities of 1 and 2a were added in taken. trans-2-Hexenal 2a and crotonaldehyde 2b
order that the power compensation could cope with both exhibit the same adiabatic behaviour, therefore,
the sudden large exotherm. The total energy released the latter was selected as substrate of choice for
for each addition was calculated from the integral of NMR studies, thus simplifying NMR spectra interpre-
the Qr curves (Figure 2, a) and showed that the over- tations. The results from the CDCl3-based reactions
all heat of reaction was approximately the same for revealed the presence of two new species (A and B),
each addition. Figure 2, b compares the chemical con- in addition to starting materials and product (see Sup-
version of hexenal with the corresponding thermal porting Information). Indeed, after careful examina-
Adv. Synth. Catal. 2010, 352, 1818 – 1825
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1819

Gennadiy Ilyashenko et al.
COMMUNICATIONS
Figure 2. a (top) Exotherm (Qr, shown as a continuous line) and total energy released during isothermal calorimetry experi-
ment (shown as a broken line). b (bottom) Thermal conversion (shown as a broken line) and chemical conversion (shown as
triangles). Note, the second addition of reagents was made at 150 min.
1
tion of each of the H NMR spectra, it was possible to
Running this reaction using trans-2-hexenal 2a in
postulate that species A and B correspond to struc- place of crotonaldehyde 2b with thioacetic acid 1 pro-
tures 4b and 5b, respectively.
duced the same results (see Figure S1 and Table S2 in
Supporting Information) and, despite peak overlap, it
was possible to decipher sufficient characteristic
NMR signals to show that this reaction proceeded
with two species equivalent to A and B being in-
volved and proposed to have structures 4a and 5a, re-
spectively. The fact that the reaction is autocatalytic
was also unequivocally confirmed by adding 10 mol%
of adduct 3b to the reaction of 1 with 2b prior to the
initiation of the reaction, which resulted in a dramatic
acceleration of the product formation and reduction
A further examination of the time-resolved NMR in the formation of species 4b and 5b. We postulated
spectroscopic studies on crotonaldehyde 2b showed that it might be the non-conjugated aldehyde function
that both the appearance and disappearance of two in the product 3 that was responsible for autocatalysis.
species A and B (i.e., 4b and 5b, respectively) could In order to probe this, the thio-Michael addition was
be followed over time in both toluene-d8 and CDCl3, carried out in the presence of both 4-nitrobenzalde-
although due to space constraints the focus is on re- hyde (10 mol%) and propionaldehyde (10 mol%) as
sults obtained from reactions carried out in CDCl3 models of 3. In both cases, the reactions were slower
only (Figure 3).
than in the uncatalysed reaction (Figure 4, d and e). It
1820
asc.wiley-vch.de
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2010, 352, 1818 – 1825

A New Autocatalytic Thioacetate-Enal Addition Reaction: A Michael Addition or Not?
1
Figure 3. Representative data for H NMR spectroscopy studies of the reaction between thioacetic acid 1 and crotonalde-
hyde 2b in CDCl3: changes in the aldehyde region (appearance of the product) and changes in the alkene region (appear-
ance and disappearance of intermediates), time at which spectrum was collected (from front to back): t1 =7 min, t2 =14 min,
1
t3 =26 min, t4 =160 min; (For the complete summary of H NMR characterisation data of reactants 1 and 2b, species A and
B, and product 3b see Supporting Information) and structures of species A and B.
appears that both 4-nitrobenzaldehyde and propional- tion of the thioacetic acid to the aldehyde, which
dehyde compete with crotonaldehyde 2b for the ini- could either be followed by a 1,3-sulfur rearrange-
tial addition of thioacetic acid 1 presumably forming ment[22] or this could simply compete with the slower
thioacyl hemiacetal species, hence, removing available direct 1,4-addition (Michael) process. Although it is
acid from the reaction mixture. Since the thioacyl known that allylic phenyl sulfides[23] undergo thermal
hemiacetal species exist in equilibrium with the alde- and acid-catalysed rearrangements (of the proposed
hyde, thioacetic acid 1 is released slowly during the mechanisms,[24–26] an associative process is supported
reaction when its concentration is decreased suffi- by kinetic isotope effects[25,26]), to the best of our
ciently enough to shift the equilibrium. As expected, knowledge, this particular rearrangement is unknown.
an electron-withdrawing nitro group favours the for- Associative mechanisms have been proposed, but
mation of thioacyl hemiacetal, hence, 4-nitrobenzalde- only in the Michael addition of amines.[28] Therefore,
hyde decreases the rate of the thio-Michael reaction in order to probe how this rearrangement might pro-
more than propionaldehyde. Although these observa- ceed, crotonaldehyde diethyl acetal 6 was reacted
tions do not explain the nature of the autocatalytic with 1 equivalent of thioacetic acid 1 (Scheme 1),
process, they eliminate the possibility of simple catal- which surprisingly derived the olefin addition product
ysis by either an aldehyde or its thioacyl hemiacetal 8. At first glance, this looks like a general acid-cata-
derivative.
lysed direct addition of thioacetic acid to the alkene
The finding that these addition reactions are two- of 6, however, following the reaction by NMR spec-
stage (Figure 1 and Figure 2), involve two species troscopy clearly revealed that the reaction proceeded
(vide supra) and are autocatalytic indicates that the quickly and cleanly to give mixed thioacyl hemiacetal
reaction is unlikely to involve a simple Michael-like 7, which slowly rearranged to the product 8. Impor-
1,4-addition of thioacetic acid to the unsaturated alde- tantly, the vinyl ether 11 was not observed, which sug-
hydes. Indeed, the reaction proceeds with a 1,2-addi- gests that the expected intermediate oxonium ion 10
Adv. Synth. Catal. 2010, 352, 1818 – 1825
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1821

Gennadiy Ilyashenko et al.
COMMUNICATIONS
1
Figure 4. Representative data for H NMR spectroscopy studies of the reaction between thioacetic acid 1 and crotonalde-
hyde 2b in CDCl3: a) concentration profile for the thermal addition of thioacetic acid; b) concentration profile for the cata-
lysed addition of thioacetic acid (10 mol% of product was added prior to the start of the reaction); c) concentration profile
for the addition of thioacetic acid in the presence of 4-nitrobenzaldehyde (10 mol%); d) concentration profile for the addi-
tion of thioacetic acid in the presence of propionaldehyde (10 mol%).
formed from protonation (via 9) of 6 can only account of thioacetate to give the observed intermediate acyl
for the appearance of acetal adduct 9 via path b hemiacetal 4. This type of species can also undergo
(Scheme 1). The formation of 7 must involve a rela- fragmentation back to 13 and recombination via a
tively fast equilibrium process to account for both its slower irreversible addition of thioacetate to derive
formation, and its subsequent reaction, back through enol 14 and hence aldehyde 3 via keto-enol tautomer-
oxonium ion 10, addition of ethanol to give 9, fol- isation. The formation of dithioacetate adduct 5 can
lowed by direct irreversible addition of thioacetate to then compete for thioacetic acid in a further fast re-
the proton activated olefin. However, since this versible addition reaction. Enol 14 is not observed,
amounts to a symmetry disallowed 1,3-H-shift, the not surprisingly, since enol derivatives of aldehydes
protonated acetal 9 could exist as a dimeric H- neither exhibit stability nor high enol concentrations
bonded species (such as 12) to which the addition of due to rapid keto-enol equilibration favouring the al-
thioacetate derives 8.
dehyde form.[28] It is likely that the conversion of 13
Elucidation of the reaction of thioacetic acid 1 with to product 3 does not solely proceed through the enol
crotonaldehyde dimethyl acetal (Scheme 1) may also from addition path b (Scheme 2), rather, species 13
clarify the mechanism operating in the apparent Mi- could transform directly to adduct 3 in a manner not
chael addition [Eq. (1)], i.e., as outlined in Scheme 2. unrelated to that proposed in the acetal system (see
Hence, reversible protonation of the unsaturated al- Scheme 1), i.e., via a species such as 15. Indeed, this
dehydes 2 derives 13, which undergoes rapid addition seems to explain the origin of the autocatalytic effect
1822
asc.wiley-vch.de
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2010, 352, 1818 – 1825

A New Autocatalytic Thioacetate-Enal Addition Reaction: A Michael Addition or Not?
Scheme 1. Proposed mechanism for the reaction of thioacetic acid 1 with crotonaldehyde diethyl acetal 8.
Scheme 2. Proposed mechanism for the reaction of thioacetic acid 1 with aldehydes 2.
since it involves the dithioacetate adduct 5 acting as of thioacetate to the activated oxonium ion and this
the autocatalytic species via hydrogen bonding to ion explanation fits the observed results.
pair 13. This results in a fast, intramolecular delivery
Adv. Synth. Catal. 2010, 352, 1818 – 1825
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1823

Gennadiy Ilyashenko et al.
COMMUNICATIONS
In summary, the addition of thioacetic acid 1 to un- General Procedure for the Addition Reactions
saturated aldehydes clearly does not proceed through Carried Out in the Thermos Flask
a simple, direct addition mechanism involving a thio-
Michael addition and as revealed by both adiabatic
All adiabatic reactions were carried out using a Thermos
Model 32-34-50 flask, Filler 32-50F with a 0.5 L capacity
and reaction calorimetry, it appears to be a two-step,
autocatalytic process. Whilst in Figure 1, two-step be-
haviour is clearly visible for toluene and chloroform,
close examination of the transients for petroleum
ether, DMF and THF also show an inflection in the
curve, although to a lesser extent. Propionitrile, how-
ever, shows a single linear response during its temper-
ature rise and acetonitrile shows only a single curve
achieving a considerably lower final temperature.
Indeed, it is clear that the system under study forms a
complex network of reactions each of which may well
proceed at different rates depending on the solvent. If
the rates involved in the first step in the reaction are
reduced by these solvents, only a single curve would
result.
and equipped with magnetic stirring. All temperature read-
ings were made using a Testo 946 digital thermometer with
a Testo Type T temperature probe linked to a PC via an
RS232/USB connector. Data were logged every 2 seconds
and stored directly on the PC using the Testo Comfort soft-
ware and exported to MS Excel for data processing. The
Dewar flask, fitted with its temperature probe was closed
with a cotton wool plug to reduce evaporative heat loss.
Solvent Screening
Solvent (100 mL), trans-2-hexenal 2a (11.0 mL, 0.094 mol)
and thioacetic acid 1 (6.6 mL, 0.094 mol) were added to the
Thermos flask and the temperature change was recorded
over time. Solvents employed were: 1) petroleum ether (80/
100); 2) THF; 3) chloroform; 4) toluene; 5) DMF; 6) pro-
piononitrile; 7) acetonitrile (see Figure 1).
Examination of the corresponding addition reaction
of thioacetic acid to an allylic acetal system uncovers
an intriguing rapid acetal exchange process, which
precedes a slower addition to the olefin. Using this as
a model, a mechanistic model as outlined in Scheme 2
can be proposed to explain the observed addition
compounds, and the important finding that aldehyde
addition precedes a formal thio-Michael addition re-
action to derive adduct 3. The exact basis of how the
formal Michael addition proceeds, i.e., via a direct
slow, 1,4-addition or a fragmentation–recombination
is not clear, however, the fact that there is an autoca-
talytic process associated with dithioacetate adduct 5
strongly indicates an assisted process such as de-
scribed by 15. It is likely that these observations will
not only have repercussions upon asymmetric SH ad-
ditions[5] since the conversion of species 4 or 6 to 3
does not appear to be a simple process, but may also
have an impact on the development of new catalytic
asymmetric Michael additions in general. Indeed, it
may be that other types of Michael reactions also pro-
ceed through similar mechanisms. Further studies to
elucidate the mechanistic, kinetic and autocatalytic
details of these processes are underway and will be
reported in due course.
General Experimental Procedure for the Reaction of
Thioacetic Acid 1 with Crotonaldehyde 2b
A solution of crotonaldehyde 2b (62 mL, 0.75 mmol) and
DCM (48 mL, 0.75 mmol) in toluene-d8 (586 mL) was pre-
pared and analysed by 1H NMR spectroscopy. Thioacetic
acid 1 (54 mL, 0.75 mmol) was then added to the reaction so-
1
lution and sample was monitored by H NMR spectroscopy
for up to 9.5 h. Once the data were assigned to correspond-
ing species, the reaction concentration profiles were ob-
tained.
Reaction of Thioacetic Acid 1 with trans-2-Butenal
Diethyl Acetal 6
A
solution of trans-2-butenal diethyl acetal 6 (20 mL,
0.14 mmol) in toluene-d8 (550 mL) was prepared and ana-
lysed by 1H NMR spectroscopy. Thioacetic acid 1 (10 mL,
0.14 mmol) was then added to the reaction solution and
1
sample was monitored by H NMR spectroscopy initially for
a period of 12 h. During that time all of the starting material
1
6 was converted to intermediate 7 (as assigned by H NMR
spectroscopy). The reaction was left for a further 48 h
during which all of the intermediate 7 was converted to
product 8. The mixture was then diluted with toluene
(3 mL), washed with saturated solution of K2CO3 (2ꢁ5 mL),
dried over MgSO4 and concentrated under vacuum to give
the product as pale yellow oil; yield: 17 mg (57%). The final
product contained traces of residual toluene and thioadduct
3b which arises from slow decomposition of 8.
Experimental Section
Synthesis of 3-Acetylthiohexenal (3a)
A mixture of chloroform (15mL), and trans-2-hexenal Acknowledgements
2a (1.64mL, 0.014 mol) was treated with thioacetic
We thank the EPSRC for a grant (GR/S85368/01) and Mr S.
Walton for technical support.
acid 1 (1mL, 0.014 mol). After 12h, the mixture was
evaporated and purified by silica gel chromatography
(hexane:ethyl acetate, 9:1, as eluent) to give 3a as a
yellow oil; yield: 2.19g (90%). All spectroscopic and
analytical data were identical to those reported.[29]
1824
asc.wiley-vch.de
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2010, 352, 1818 – 1825

A New Autocatalytic Thioacetate-Enal Addition Reaction: A Michael Addition or Not?
References
[15] R. N. Landau, Thermochim. Acta 1996, 289, 101–126.
[16] G. L. Khatik, R. Kumar, A. K. Chakraborti, Org. Lett.
[1] A. Michael, J. Prakt. Chem. 1887, 35, 349–356.
[2] C. F. Koelsch, J. Am. Chem. Soc. 1943, 65, 437–439.
[3] J. Leonard, E. Diez-Barra, S. Merino, Eur. J. Org.
Chem. 1998, 2051–2061.
[4] B. E. Rossiter, N. M. Swingle, Chem. Rev. 1992, 92,
771–806.
[5] D. Enders, K. Lꢂttgen, A. A. Narine, Synthesis 2007,
959–980.
[6] S. Hoz, Acc. Chem. Res. 1993, 26, 69–74.
[7] H. K. Oh, I. K. Kim, H. W. Lee, I. Lee, J. Org. Chem.
2004, 69, 3806–3810.
2006, 8, 2433–2436.
[17] P. Bernal, J. Tamariz, Tetrahedron Lett. 2006, 47, 2905–
2909.
[18] BatchCAD v8, http://www.aspentech.com/.
[19] H. Li, J. Wang, L. Zu, W. Wang, Tetrahedron Lett. 2006,
47, 2585–2589.
[20] J. Barton, R. Rogers, Chemical Reaction Hazards, Insti-
tute of Chemical Engineering, Rugby, Warwickshire,
UK, 1993.
[21] S. Walton, A. Whiting, Org. Process Res. Dev. 2006, 10,
846–846.
[22] S. K. Burr, Top. Curr. Chem. 2007, 274, 125–171.
[23] P. Brownbridge, S. Warren, J. Chem. Soc. Perkin Trans.
1 1976, 2125–2132.
[24] H. Kwart, N. A. Johnson, J. Am. Chem. Soc. 1977, 99,
3441–3449.
[25] H. Kwart, J. Stanulonis, J. Am. Chem. Soc. 1976, 98,
4009–4010.
[26] H. Kwart, T. J. George, J. Am. Chem. Soc. 1977, 99,
5214–5215.
[27] F. M. Menger, J. H. Smith, J. Am. Chem. Soc. 1969, 91,
4211–4216.
[28] S. Perisanu, R. Vilcu, Rev. Roum. Chim. 1992, 37, 197–
203.
[29] a) D. Bhar, S. Chandrasekaran, Tetrahedron 1997, 53,
11835–11842; b) F. F. Khouri, M. E. Kaloustian, J. Am.
Chem. Soc. 1986, 108, 6683–6695.
[8] T. C. Wabnitz, L.-Q. Yu, J. B. Spencer, Chem. Eur. J.
2004, 10, 484–493.
[9] T. C. Wabnitz, J. B. Spencer, Org. Lett. 2003, 5, 2141–
2144.
[10] F. Robert, J. Heritier, J. Quiquerez, H. Simian, I.
Blank, J. Agric. Food Chem. 2004, 52, 3525–3529.
[11] S. Widder, C. S. Luntzel, T. Dittner, W. Pickenhagen, J.
Agric. Food Chem. 2000, 48, 418–423.
[12] H. Wakabayashi, M. Wakabayashi, W. Eisenreich, K.-
H. Engel, J. Agric. Food Chem. 2003, 51, 4349–4355.
[13] Harsnet, Thematic network on hazard assessment of
highly reactive systems, http://www.iqs.url.es/harsnet. In-
dustrial and Materials Technologies Programme of the
European Commission; Project BET2-0572.
[14] C. LeBlond, J. Wang, R. D. Larsen, C. J. Orella, A. L.
Forman, R. N. Landau, J. Laquidara, J. R. Sowa, D. G.
Blackmond, Y. K. Sun, Thermochim. Acta 1996, 289,
189–207.
Adv. Synth. Catal. 2010, 352, 1818 – 1825
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1825