Competing Knorr and Fischer-Fink pathways to pyrroles in neutral aqueous solution

Article abstract of DOI:10.1016/j.tet.2012.05.080

A proposed chemical model for the prebiogenesis of tetrapyrrole macrocycles relies on the condensation of a 3-alkyl-substituted 2,4-diketone and an α-aminoketone to form a pyrrole equipped for subsequent self-condensation leading to porphyrinogens. The co

Full text of DOI:10.1016/j.tet.2012.05.080

Tetrahedron 68 (2012) 6957e6967
Contents lists available at SciVerse ScienceDirect
Tetrahedron
journal homepage: www.elsevier.com/locate/tet
Competing Knorr and FischereFink pathways to pyrroles in neutral aqueous
solution
*
Vanampally Chandrashaker, Masahiko Taniguchi, Marcin Ptaszek, Jonathan S. Lindsey
Department of Chemistry, North Carolina State University, Raleigh, NC 27695, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 5 May 2012
Received in revised form 18 May 2012
Accepted 21 May 2012
Available online 28 May 2012
A proposed chemical model for the prebiogenesis of tetrapyrrole macrocycles relies on the condensation
of a 3-alkyl-substituted 2,4-diketone and an a-aminoketone to form a pyrrole equipped for subsequent
self-condensation leading to porphyrinogens. The condensation of acyclic reactants can proceed via
competing Knorr (desired) and FischereFink (undesired) pathways. Here, the Knorr and FischereFink
pathways are quantitated using (1) analogues of those in the proposed prebiotic route wherein the re-
sulting pyrroles are blocked from subsequent condensation, and (2) a set of internal standards for
quantitation of crude reaction mixtures by 1H NMR spectroscopy following extraction into CS2. The re-
action of 1-amino-2-butanone and 3-methyl-2,4-pentanedione at 60 ꢀC in aqueous solution at pH 7 for
24 h afforded the Knorr pyrrole and FischereFink pyrrole in yields of 48% and 7%, respectively. The re-
action in hot aqueous acid (pH 4.6, reflux, 30 min) afforded the Knorr pyrrole in diminished yield (22%)
relative to that of the FischereFink pyrrole (11%). The Knorr pyrrole (4-ethyl-2,3-dimethylpyrrole) is
evanescent whereas the FischereFink pyrrole (2,3,4-trimethyl-5-propanoylpyrrole) is quite stable;
hence, rapid analysis of the reaction mixture (vs workup and isolation) is essential for accurate analysis.
Ó 2012 Elsevier Ltd. All rights reserved.
Keywords:
Prebiotic
Mass balance
Tetrapyrrole
Aminolevulinic acid
Biomimetic
Pyrrole
1. Introduction
undergo self-condensation leading to the formation of uropor-
phyrinogens (use of the plural indicates the presence of up to four
As part of a program aimed at developing a chemical model for
the prebiotic origin of tetrapyrrole macrocycles, we have been in-
vestigating routes that parallel the extant biosynthesis of tetra-
pyrrole macrocycles yet proceed under prebiotic conditions.1e4 In
isomers). On the other hand, the FischereFink pyrrole I-FF is
blocked from subsequent reactions and is expected to constitute
a dead end. Electrospray mass spectral examination of the reaction
mixture indeed showed a readily detectable peak consistent with
the FischereFink pyrrole I-FF. Mass spectral evidence for the Knorr
pyrrole I-Kn was harder to obtain because the latter is equipped
for self-condensation to give oligopyrromethanes and the
porphyrinogen. Nonetheless, a weak peak consistent with I-Kn
could be observed in experiments carried out at room temper-
ature rather than the typical 60 ꢀC. Given that the overall yield
of uroporphyrinogens was 10% beginning with ALA and 1-AcOH
upon reaction at 60 ꢀC, the low intensity of the peak due to the
Knorr pyrrole I-Kn likely reflects subsequent condensation
leading to the porphyrinogens.2
The 10% yield of uroporphyrinogens upon reaction of ALA and 1-
AcOH has multiple possible explanations. A point of reference is
provided by the reaction of PBG, which upon nonenzymic reaction
even in dilute solution (0.26 mM, pH 7.4, 70 ꢀC, 2 h) affords uro-
porphyrinogens in w65% yield.7 Here, the question of interest is the
extent to which the FischereFink pathway provides a competitor to
the Knorr pathway and thereby limits the conversion of 1-AcOH and
ALA in forming uroporphyrinogens. The intermediate nature of I-
Kn owing to subsequent self-condensation, versus the accumula-
tion of I-FF, precluded quantitation of the respective pathways.
the biosynthesis, two molecules of
d-aminolevulinic acid (ALA)
condense under enzymic catalysis to give the pyrrole porphobili-
nogen (PBG). PBG then undergoes tetramerization and cyclization
with rearrangement to give uroporphyrinogen III as the sole mac-
rocyclic product (Scheme 1). While PBG self-condenses in the ab-
sence of enzymes to give all four possible uroporphyrinogen
isomers (of which isomer III is the dominant species), attempts to
achieve the ALA/PBG conversion in the absence of enzymes have
invariably failed.
We eventually examined alternative routes to a pyrrole analo-
gous to PBG that would undergo self-condensation to give a por-
phyrinogen. Two routes were identified.1,2 In one route, ALA
condenses with a dione (1-AcOH) to give a methoxy-substituted
analogue of PBG (I-Kn) (Scheme 1). In the pyrrole-forming step
there are several possible products, including the desired Knorr
product (I-Kn) and the undesired FischereFink product (I-FF). The
terms refer to the original pyrrole syntheses reported by Knorr5 in
1884 and by Fischer and Fink6 in 1944. The I-Kn pyrrole can
* Corresponding author. E-mail address: jlindsey@ncsu.edu (J.S. Lindsey).
0040-4020/$ e see front matter Ó 2012 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2012.05.080

6958
V. Chandrashaker et al. / Tetrahedron 68 (2012) 6957e6967
A
Biosynthesis
H2N
MeO
MeO
H
N
O
PBG synthase
(1) PBG deaminase
Uroporphyrinogen III
H2N
CO2H
(2) Uroporphyrinogen
III synthase
CO2H
CO2H
ALA
PBG
R3
R7
B
Model Abiotic Synthesis
tetramerization
and
R2
R8
H
N
O
NH HN
NH HN
macrocyclization
H2N
CO2H
ALA
Knorr
pathways
R12
CO2H
CO2H
R18
+
I-Kn
R17
Uroporphyrinogens I-IV
R13
O
O
sequence of R2 - R18
O
O
O
H
N
Fischer-Fink
AP-AP-AP-AP-
I
AP-PA-AP-PA- II
AP-AP-AP-PA- III
AP-AP-PA-PA- IV
CO2H
CO2H
1-AcOH
OMe
A = acetic acid
P = propionic acid
CO2H
No
macrocycles
I-FF
Scheme 1. (A) Biosynthetic conversion of ALA to Uroporphyrinogen III. (B) Model abiotic synthesis showing the Knorr versus FischereFink pathways in the formation of
Uroporphyrinogens IeIV. Note the structural similarity of pyrroles PBG and I-Kn.
Rapoport first quantified the Knorr versus FischereFink path-
ways with radiolabeled substrates. Ethyl -aminoacetoacetate
pyrrole (III-FF-1:III-Kn) was 43.8:1.2 with <1% of the second
FischereFink pyrrole (III-FF-2) that forms upon loss of ethanol and
carbon dioxide instead of acetic acid. Regardless of increasing the
steric burden (R¼tert-butyl) or changing the electronic nature
(R¼p-methoxyphenyl), the FischereFink pathway was still domi-
nant. These results unambiguously revealed the predominance of
the FischereFink pathway, yet the amino-containing substrate was
quite different from that in the reaction of ALA and 1-AcOH.
Falk et al. quantified the Knorr versus FischereFink pathways
using gas chromatography (unspecified detector).9 Representative
a
(R¼methyl) with a 14C-substituted keto carbon was condensed
with methyl 4-acetyl-5-oxohexanoate (Scheme 2).8 Competing in-
termediates leading to the Knorr versus FischereFink pathways are
illustrated for clarity as anions II-Kn and II-FF. Two FischereFink
pyrroles (III-FF-1, III-FF-2) and one Knorr pyrrole (III-Kn) were
obtained. Two pyrroles (III-Kn and III-FF) differ only in the pres-
ence of one labeled carbon and were quantified by means of spe-
cific activity. The ratio of one FischereFink pyrrole and the Knorr
Scheme 2. Rapoport scheme for quantitation of Knorr and FischereFink pathways. The atom labels are provided for clarity.

V. Chandrashaker et al. / Tetrahedron 68 (2012) 6957e6967
6959
substrates are shown in Scheme 3. The reaction of a
b-ketoaldehyde
A
O
O
and an -amino- -ketoester affords two possible initial sites of
a
b
V-Kn
V-FF
imine formation; the resulting aldimine and ketimine each can
then proceed along FischereFink or Knorr pathways. When R1¼Et
and R2¼Me, the product distribution was IV-ket-FF¼85% (isolated
55%), IV-ket-Kn¼1%, IV-ald-FF¼14%, and IV-ald-Kn was not
detected. The use of gas chromatography in this work was enabled
in part because each pyrrole bears one ester moiety, which imparts
considerable stability to the pyrrole nucleus.
pH 4.6
30 min
reflux
H
N
O
H
N
2
+
+
CO2H
O
O
H2N
CO2H
CO2H
62% isolated yield
95:5 ratio
ALA
B
O
O
VI-Kn
VI-FF
pH 4.6
30 min
reflux
H
N
O
H
N
2-Me
+
+
CO2H
O
H2N
CO2H
CO2H
ALA
not observed
16.7% yield
Scheme 4. Butler results concerning Knorr and FischereFink pathways with two
diones.
crystalline and hence prone to remain in solution, especially in
contrast with the FischereFink pyrrole (VI-FF), which is a 2-
ketopyrrole.
Third, the reaction conditions of Butler in refluxing acidic so-
lution (pH 4.6) were quite different from those we sought to em-
ploy (more neutral pH at 60 ꢀC).
Still, the question remained as to whether the low yield of
uroporphyrinogen in Scheme 1 stemmed from competing reaction
to give the FischereFink pyrrole or from other reasons, such as poor
Scheme 3. Falk study of competing Knorr and FischereFink pathways.
nucleofugacity of the methoxy group at the a-methyl site of I-Kn
(vs the ammonium group of PBG). The uncertainty concerning the
relative amounts of the Knorr and FischereFink pyrroles (I-Kn vs I-
FF) prompted quantitative studies, which are described in this
paper. The substrates employed are analogues of dione 1-AcOH and
aminoketone ALA wherein the resulting pyrroles are blocked to-
ward subsequent self-condensation (leading to oligopyrro-
methanes and porphyrinogens). We employed 1H NMR
spectroscopy for assessing relative and absolute yields, which
complements the prior methods of Rapoport, Falk, and Butler
wherein radiolabeling, gas chromatography, and isolation were
utilized. The study here thus provides a first step toward a quanti-
tative understanding of the spontaneous formation of tetrapyrrole
macrocycles from acyclic precursors. Such types of structure-
directed transformationsdwhile far less efficient than enzymic
processes17,18dare believed to be central to the prebiotic origin of
molecules on the early Earth.19
A study with ALA itself and diones was carried out by But-
ler,10e12 who sought to understand the intrinsic reactivity of ALA in
the context of the biosynthetic formation of PBG. The results of
Butler that are germane to the work described herein include the
following. The condensation of ALA and 2,4-pentanedione (2)
afforded a solid product (62% yield) that was shown to be com-
posed of the Knorr and FischereFink pyrroles (V-Kn, V-FF) in 95:5
ratio (panel A, Scheme 4). On the other hand, when the 2,4-dione
contained a 3-methyl group (2-Me, panel B), examination of the
isolated solid product showed the presence only of the
FischereFink pyrrole VI-FF (16.7% yield) without any Knorr pyrrole
(VI-Kn). The results of Butler, while quite consistent with the lit-
erature concerning the effect of 3-alkyl groups on pyrrole-forming
reactions,13 are at odds with the success of the reaction of ALA and
1-AcOH (Scheme 1).
We had embarked on the route shown in Scheme 1 fully cog-
nizant of the results of Rapoport, Falk, and Butlerdas well as
a number of related studies14e16 wherein the FischereFink path-
way was predominant over the Knorr pathwaydyet encouraged
given the following considerations.
2. Results
2.1. Quantitation of Knorr and FischereFink pyrroles
First, in the Rapoport study, the dicarbonyl groups in the amino
substrate were expected to facilitate reaction via the FischereFink
pathway, owing to preferred formation of intermediate II-FF versus
II-Kn. The same interpretation applies to the Falk study. By con-
trast, there is only one carbonyl group in ALA.
Second, in the Butler study, the low yield of the FischereFink
pyrrole (VI-FF) and focus solely on isolable solid products left the
mass balance very incomplete. In this regard, we anticipated
that the Knorr pyrrole (VI-Kn, a trialkylpyrrole) might be poorly
The choice of reactants and analytical method is essential for
quantitating the Knorr and FischereFink products. We initially
chose 2-aminoacetophenone (3-Ph) for reaction with 3-methyl-
2,4-pentanedione (2-Me) to have pyrroles (Knorr and
FischereFink) that (i) could be extracted into organic solvents and
(ii) would be amenable to GC and 1H NMR analysis. While the two
pyrroles were readily observed upon GCeMS analysis (see
Supplementary data), attempts to use 1H NMR spectroscopy were

6960
V. Chandrashaker et al. / Tetrahedron 68 (2012) 6957e6967
thwarted by the nearly identical chemical shifts and resonance
patterns of the phenyl protons of the Knorr and the FischereFink
pyrroles. Accordingly, we changed the aminoketone to 1-amino-2-
butanone (3-Et), which contains the reactive motif of ALA but lacks
the carboxylic acid unit. The reaction investigated is shown in
Scheme 5. Authentic samples of the expected reaction products FF-
Et, Kn-Et,20 and Et2-pyrazine21 were prepared as described in the
Appendix.
Dimethyl 5-methoxyisophthalate is water-insoluble yet dissolves
upon organic extraction of the crude reaction mixture. In another
set, a known amount of diethyl carbonate (B) was added at the end
of the reaction immediately before NMR data acquisition. The loss
of diethyl carbonate under the workup conditions was tested and
found to be <4%, an insignificant amount. The 1H NMR signature of
each standard was utilized for quantitative analysis. The yield and
ratio values reported herein are inferred by 1H NMR spectroscopy
through the use of a quantitative internal standard, unless specif-
ically identified as an isolated (i.e., gravimetric) yield.
We also found that the products of the crude reaction mixture
could be examined efficaciously in carbon disulfide, which is a su-
perb solvent for NMR studies.22 On the other hand, early attempts
to use CDCl3 resulted in very low yields of the Knorr pyrrole. We
attribute the latter result to decomposition of the Knorr pyrrole in
the presence of trace acid from the chloroform solvent.
N
N
Et
N
N
Et
[O]
Et
Et
a dihydropyrazine
dimerization
Et2-pyrazine
2.2. Reaction of 3-methyl-2,4-pentanedione and 1-amino-2-
butanone
O
O
O
+
3-Et
2-Me
H2N
2.2.1. Quantitation. The reaction of 3-methyl-2,4-pentanedione (2-
Me) and 1-amino-2-butanone (3-Et) (0.25 mmol each with 50 mM
of each reactant) was carried out in aqueous solution (pH 7, phos-
phate buffer) at 60 ꢀC for 24 h (Scheme 5). The conditions were
identical with those in the abiotic uroporphyrinogen synthesis
(Scheme 1, panel B).2
Et
HN
O
O
Analysis of the crude reaction mixture by GCeMS revealed the
presence of peaks assigned to the Knorr pyrrole (Kn-Et) and
FischereFink pyrrole (FF-Et) (Fig. 1). A semi-preparative scale re-
action of 2-Me (20 mmol) and 3-Et (2.0 mmol) was carried out in
a mixed solvent of aqueous phosphate buffer solution (pH 7) and
ethanol at 60 ꢀC for 24 h. The reaction mixture was extracted with
Et2O and chromatographed on silica (hexanes/ethyl acetate, 4:1).
The only pyrrole isolated was the FischereFink pyrrole (FF-Et),
which was obtained in 6.6% yield. While additional components
were visible upon TLC analysis, no clear fraction corresponding to
the Knorr pyrrole (Kn-Et) was obtained. The Knorr pyrrole (Kn-
Et)dalso termed hemopyrrole20dbears three alkyl groups and is
known to be unstable to air and light.23 Compound FF-Et was
a stable white crystalline solid, which was thoroughly characterized
by NMR spectroscopy and single-crystal X-ray crystallography.
The limited stability of Kn-Et prompted use of 1H NMR spec-
troscopy for quantitation. The 1H NMR assignments of the various
constituents (in CS2 with C6D6 lock and reference) are shown in
Fig. 2 (panel A). The most reliable means of yield estimation
entailed the substantial difference in chemical shifts of the quartets
(2.61, 2.99 ppm) of the Knorr and FischereFink products. Thus, the
reaction of 2-Me and 3-Et (0.25 mmol each with 50 mM of each
reactant) was carried out in aqueous solution (pH 7, phosphate
buffer) at 60 ꢀC for 24 h in the presence of dimethyl 5-
methoxyisophthalate. The crude reaction mixture was extracted
with Et2O, which dissolves the internal standard dimethyl 5-
methoxyisophthalate. The Et2O extract was concentrated, and the
organic residue was taken up in CS2 for 1H NMR analysis. The peaks
due to FF-Et and Kn-Et were readily observed. Signals corre-
sponding to 2-Me also were evident (keto and enol forms), as were
peaks assigned to the pyrazine (Et2-pyrazine) derived from di-
merization of 3-Et followed by oxidation.
Fischer-Fink
pathway
Knorr
pathway
O
H
N
H
H
OH
N
Et
H
OH
Et
O
loss of AcOH
loss of H2O
O
H
N
H
N
FF-Et
Kn-Et
Scheme 5. FischereFink versus Knorr pathways of model reactants.
Two commercially available compounds were employed as in-
ternal standards for 1H NMR spectroscopy (Chart 1). The choice of
standards for a given experiment depended on the location of the
1H NMR resonances of the products and the nature of the workup
procedure (if any). In one set of reactions, a known amount of di-
methyl 5-methoxyisophthalate (A) was added along with reactants.
OMe
The results upon 1H NMR analysis of the reaction of 2-Me and
3-Et are shown in Table 1. The Knorr pyrrole was the dominant
product (48%) versus the FischereFink pyrrole (7%); the ratio was
7:1 (entry 1). The recovered dione 2-Me (3%) was observed
along with a slight amount (3%) of Et2-pyrazine due to self-
condensation of aminoketone 3-Et. The quartet of the pyrazine at
3.07 ppm was close todbut clearly distinct fromdthe low field
FischereFink quartet and therefore could be quantitated by
O
Ha
Ha
Ha
Ha
O
Ha
O
Ha
MeO2C
CO2Me
Ha = 7.94 (m) ppm
Ha = 4.37 (q) ppm
A
B
Chart 1. Standards for quantitative analysis in CS2.

V. Chandrashaker et al. / Tetrahedron 68 (2012) 6957e6967
6961
Fig. 2. 1H NMR spectral data in CS2 (C6D6 was used as external standard, 300 MHz). (A)
1H NMR chemical shifts of key constituents. (B) FischereFink pyrrole FF-Et. (C) Knorr
pyrrole Kn-Et. (D) The crude mixture obtained from the condensation of 2-Me and
3-Et at 60 ꢀC for 24 h.
Fig. 1. GCeMS trace of the crude mixture obtained from the condensation of 2-Me and
3-Et. (A) GC trace with temperature gradient: temp 1, 60 ꢀC (3 min); temp 2, 200 ꢀC
(5 min); rate 10 ꢀC/min, total run-time, 22 min. (B) Mass spectrum at 9.50 min
corresponding to Kn-Et (C8H13N, calcd m/z¼123.1). (C) Mass spectrum at 17.72 min
corresponding to FF-Et (C10H15NO, calcd m/z¼165.1).
(56%) was seven times that of the FischereFink product FF-Et (8%)
(entry4). A reactionwith 5 equivofdione 2-Meaffordedsimilar results,
with a 40% yield of Kn-Et versus 6% yield of FF-Et (entry 5).
From entries 1e5 the combined yields including the pyrazine
and recovered dione only reached 74%. An omission experiment
with dione 2-Me alone (at 60 ꢀC for 24 h) recovered between 30 and
45% (ketoþenol) with respect to the standard diethyl carbonate. On
the other hand, a simple buffer/CS2 extraction of the dione 2-Me
integration (Fig. 2). The combined yield including the recovered
dione was 64%.
An identical experiment gave only slight variation in the results
(entry 2); however, splitting the ether extract sample and delaying
analysis for 1 h resulted in a decline in the Knorr product (to 42% vs
54%) whereas the FischereFink product was unchanged at 6%
(entry 3). The results indicate the instability of the Knorr product
and the urgency for prompt analysis.
(29 mL, 0.25 mmol) in the presence of the standard resulted in re-
To shorten the workup process, quantitation with dimethyl 5-
methoxyisophthalate and Et2O extraction (entries 1e3) was replaced
with diethyl carbonate in CS2 (entries 4 and 5). Thus, after reaction for
24 h and then allowing the mixture to cool to room temperature
covery of only 62% (ketoþenol) of 2-Me. Extrapolation of the
38e55% loss of dione upon workup to that with entry 4 accounts for
approximately 80% of the total mass balance.
(10 min), CS2 (5 mL) and diethyl carbonate (125 mL of a 1.0 M solution in
2.2.2. Kinetic study. The apparent instability of the Knorr pyrrole Kn-
Et prompted a study to monitor the reaction time course. Four re-
actions were setup and quenched after 24 h, 48 h, 72 h, and 96 h, and
CS2, 0.125 mmol) were added, the CS2 fraction was separated, and 1H
NMR analysis was performed. In so doing, the Knorr pyrrole Kn-Et

6962
V. Chandrashaker et al. / Tetrahedron 68 (2012) 6957e6967
Table 1
Knorr versus FischereFink products (corresponding to Scheme 5).a
Entry
Kn-Et %
FF-Et %
2-Me %f
Et2-pyrazine %
Total %
1b
48
54
42
56
40
7
6
6
8
6
3
2
2
4
15
3
2
2
3
1
64
66
54
74
d
2b
3b,c
4d
5d,e
a
Each reaction was carried out with 50 mM of each reactant in aqueous phos-
phate buffer (pH 7) at 60 ꢀC for 24 h unless noted otherwise.
b
Workup entailed extraction with Et2O and recording 1H NMR spectra in CS2
with dimethyl 5-methoxyisophthalate (5 mM) as internal standard.
c
The Et2O extract was stored at room temperature for 1 h before proceeding with
the workup.
d
Workup entailed extraction with CS2 and recording 1H NMR spectra in CS2 with
diethyl carbonate (25 mM) as internal standard.
e
Excess (5 equiv, 250 mM) of 2-Me was used, of which only 15% was recovered
(37 mM).
f
Only the keto (not the enol) form was accounted for in the 1H NMR spectra.
analyzed as described in Table 1, entry 4. The 1H NMR spectra and
accompanying yield data are displayed in Fig. 3 (top panels). The yield
as a function of time (for the four relevant species) is shown in the
graph in Fig. 3 (lower panel). There was no change in the ratio of Knorr
versus FischereFink pyrroles at 48 h; however, thereafter the yield of
the Knorr pyrrole steadily declined to and almost disappeared at the
end of 96 h. Bycontrast, the yield of the FischereFink pyrrole remained
unchanged. The instability of the Knorr pyrrole (after 2 days) versus
the stability of the FischereFink pyrrole is a consequence of absence or
presence of the electron-withdrawing acyl moiety, respectively.
In summary, with respect to the internal standard the following
conclusions can be drawn: (1) The FischereFink pyrrole (FF-Et) is
stable and forms inyields ranging from 6 to 8%. (2) The Knorr pyrrole
(Kn-Et) is the dominant product and forms inyields ranging from 40
to 56%. (3) The observed yield of the Knorr pyrrole (Kn-Et) can vary
due to factors associated with the workup process. (4) The dihy-
dropyrazine (or Et2-pyrazine upon oxidation) derived from di-
merization and dehydration of the aminoketone 3-Et forms in yields
of 1e3%. (5) The unreacted dione 2-Me is largely lost upon workup.
2.3. Reactions (of ALA and diones) and comparison with
Butler
In the studies reported herein, the dominance of the Knorr pyr-
role was at odds with the prevalence of the FischereFink pyrrole
reported by Butler with quite similar reactants: Butler employed
ALAwhereas here we employed 1-amino-2-butanone (3-Et). On the
other hand, Butler employed reflux in a modestly acidic solution (pH
4.6) for short duration (30 min) whereas here we have employed
more neutral pH at 60 ꢀC for 24 h. To reconcile the disparate results,
we made several direct comparisons as outlined in Table 2. The
Butler results from Scheme 4 are shown here for direct comparison.
First, to validate the Butler result (Scheme 4, panel A; Table 2,
entry 1), we carried out the reaction of acetylacetone (2) and ALA in
refluxing aqueous solution (pH 4.6, acetate buffer) for 30 min. The
resulting white precipitate was isolated, dried, and analyzed by 1H
NMR spectroscopy, which showed the presence of the Knorr and
FischereFink pyrroles in 93:7 ratio (entry 2). These results are al-
most identical with those of Butler, who observed a 95:5 ratio
(entry 1). The predominance of the Knorr pyrrole is expected for 3-
unsubstituted 2,4-dione 2.
On the otherhand, Butler reported that the reaction of ALA and 3-
methyl-2,4-pentanedione (2-Me) under identical conditions pro-
vided a precipitate that contained only the FischereFink compound
(16.7% yield; Scheme 4, panel B; Table 2, entry 3).
Fig. 3. AeD: 1H NMR spectral data in CS2 (C6D6 was used for locking and a chemical
shift standard, 300 MHz) of the crude mixture obtained from the condensation (for the
durations indicated) at 60 ꢀC of 2-Me and 3-Et. The low quantity of 2-Me reflects only
partial recovery. E: Time course data derived from the 1H NMR experiments. (For 2-Me,
only the keto tautomer was taken into account for yield calculations.)
Reaction of ALA and 2-Me under neutral conditions (pH 7, 60 ꢀC,
24 h) followed by extraction with CS2 (with diethyl carbonate as
standard) resulted in only the starting material 2-Me (unreacted),

V. Chandrashaker et al. / Tetrahedron 68 (2012) 6957e6967
6963
Table 2
Prior studies of Knorr versus FischereFink under acidic versus neutral conditionsb
Entry
Aminoketone
Dione
Reaction conditions
Knorr product
FischereFink product
H
N
O
H
N
O
O
O
O
CO2H
H2N
CO2H
1,2
Aq acetate, pH 4.6, reflux, 30 min
CO2H
2
62% isolated total (95:5 ratio)a
93:7 ratio (this study)
ALA
H
N
O
H
N
O
O
O
H2N
CO2H
CO2H
3a
Aq acetate, pH 4.6, reflux, 30 min
CO2H
ALA
2-Me
b
O
O
O
H2N
CO2H
4c
Aq phosphate, pH 7, 60 ꢀC, 24 h
Not known; see text
Not known; see text
ALA
2-Me
H
N
O
H
N
O
O
O
H2N
CO2CH3
CO2CH3
5c
Aq phosphate, pH 7, 60 ꢀC, 24 h
CO2CH3
ALA-Me
2-Me
6.5% yield
60% yield
O
H
N
O
O
H
N
O
H2N
6c
Aq acetate, pH 4.6, reflux, 30 min
FF-Et,
3-Et
2-Me
11% yield
Kn-Et,
22% yield
a
Butler studies.10
Isolated yield.
Yields are on the basis of the standard diethyl carbonate.
b
c
whereas the aqueous layer was colored (entry 4). The inability to ex-
tract the pyrroles precluded quantitative analysis. Our inference was
that both Knorr and FischereFink pyrroles were likely formed and
remained in the aqueous solution; therefore, we replaced ALA with the
Butler experiment (entry 3), yet the 2:1 ratio under acidic condi-
tions at reflux is far less than for the same reactants under neutral
conditions as described in Table 1.
Our interpretations of the results are as follows. The reaction of
methyl ester of
d
-aminolevulinic acid (ALA-Me). The reaction of ALA-
dione 2-Me and an
a small amount of the FischereFink product, but the Knorr product
is present in larger quantity. The FischereFink product (an -keto-
trialkylpyrrole) is more stable and more readily isolable than the
Knorr pyrrole (a trialkylpyrrole). It is not surprising that Butler
isolated only the FischereFink pyrrole. On the other hand, the re-
a-aminoketone (ALA or 3-Et) at pH 4.6 affords
Me with 2-Me (pH 7, 60 ꢀC, 24 h) followed by extraction in CS2 and
subsequent 1H NMR analysis gave both the Knorr pyrrole (60% yield)
and the FischereFink pyrrole (6.5% yield), in other words a w9:1 ratio
(entry 5).
a
We carried out the reaction of 2-Me and 3-Et under the Butler
conditions and analyzed the reaction by extraction with CS2 con-
taining diethyl carbonate as standard. We observed the Knorr
pyrrole in 22% yield and the FischereFink pyrrole in 11% yield
(entry 6). The Knorr product was present, not absent, as in the
action of dione 2-Me and an a-aminoketone (ALA-Me or 3-Et) at
neutral pH affords a large amount of the Knorr product and a small
amount of the FischereFink product; yields of 60% and 6.5% ob-
served in Table 2 (entry 5) cohere with those observed in Table 1.

6964
V. Chandrashaker et al. / Tetrahedron 68 (2012) 6957e6967
3. Conclusions
presence of a sodium phosphate buffer (pH 7.0 ꢂ 0.1, stated as pH 7
for simplicity).
The reaction of a 1,3-dicarbonyl compound and an
a-amino-
carbonyl compound provides the basis for the Knorr pyrrole synthesis.
Traditionally, the amine is an -amino- -ketoester and the 1,3-
dicarbonyl compound lacks a substituent at the 2-position. When
a 2-alkyl-1,3-dicarbonyl compound is employed with an -amino-
ketoester, the FischereFink pathway becomes competitive with the
Knorr pathway. Here, we have shown that the use of an -amino-
ketone(lackingasecondcarbonylgroup)uponreactionwitha3-alkyl-
substituted 2,4-dione affords the Knorr pathway predominantly, but
not to the exclusion of, the FischereFink pathway. The route provides
direct access to 2,3,4-trialkylpyrroles, which are highly reactive to-
wardelectrophilicsubstitution. Itwarrants mentionthathadwetaken
the Butlerresults onface valuedwhere the FischereFink pyrrole is the
only observed product upon reaction of ALA and a 3-alkyl-substituted
dione (e.g., 3-methyl-2,4-pentanedione)dthere would have been no
motivation for examining the reaction of ALA and 1-AcOH. Reaction of
the latter has been employed as the basis for a chemical model of the
prebiogenesis of tetrapyrrole macrocycles beginning with acyclic re-
actants in aqueous solution.2e4
4.2. Synthesis
a
b
4.2.1. 2-Benzoyl-3,4,5-trimethylpyrrole (FF-Ph). A mixture of
aqueous sodium phosphate buffer solution (6 mL, 0.2 M, pH 7),
ethanol (20 mL), and water (20 mL) in a Schlenk flask was purged
with argon for 15 min. The flask was placed on dry ice for 30 min. A
sample of 2-Me (2.3 mL, 20 mmol) was added to the frozen mixture
followed by 3-Ph (344 mg, 2.0 mmol). The flask was evacuated and
purged with argon. Then, the reaction mixture was allowed to
warm to room temperature under a slow flow of argon. The re-
action flask was heated to 60 ꢀC for 24 h in the dark. The reaction
mixture was allowed to warm to room temperature. The reaction
mixture was treated with Et2O (50 mL). The contents were trans-
ferred to a separatory funnel and diluted with 20 mL of water. The
Et2O extract (2ꢃ100 mL) was dried (Na2SO4), concentrated and
chromatographed (silica, hexanes/ethyl acetate, 4:1). The title
compound was obtained as a yellow solid (35 mg, 8.2%): mp
a
b-
a
132e133 ꢀC (lit.26 mp 134e135 ꢀC); 1H NMR (300 MHz)
3H), 1.92 (s, 3H), 2.23 (s, 3H), 7.40e7.55 (m, 3H), 7.58e7.65 (m, 2H),
8.95 (br s, 1H); 13C NMR (75 MHz)
9.0, 11.8, 12.2, 118.8, 127.0,
d 1.88 (s,
We began this work with the notion that the low yields (ꢁ10%)
of porphyrinogens observed upon reaction of an
a
-aminoketone
d
and a suitably functionalized 1,3-dione likely stemmed from pre-
dominant formation of the FischereFink pyrrole at the expense of
the Knorr pyrrole. The quantitative approaches developed herein
have enabled delineation of a model reaction thereof, namely the
128.39, 128.41, 129.2, 130.9, 133.5, 140.6, 185.6; ESI-MS obsd
214.1227, calcd 214.1226 [(MþH)þ, M¼C14H15NO].
4.2.2. 3,4,5-Trimethyl-2-propanoylpyrrole (FF-Et). A mixture of
aqueous sodium phosphate buffer solution (6 mL, 0.2 M, pH 7),
ethanol (20 mL), and water (20 mL) in a Schlenk flask was purged
with argon for 15 min. The flask was placed on dry ice for 30 min. A
sample of 2-Me (2.3 mL, 20 mmol) was added to the frozen mixture
followed by 3-Et (246 mg, 2.0 mmol). The flask was evacuated and
purged with argon. Then, the reaction mixture was allowed to
warm to room temperature under a slow flow of argon. The re-
action flask was then heated to 60 ꢀC for 24 h in the dark. The re-
action mixture was allowed to warm to room temperature. The
reaction mixture was treated with Et2O (50 mL). The contents were
transferred to a separatory funnel and diluted with 20 mL of water.
The Et2O extract (2ꢃ100 mL) was dried (Na2SO4), concentrated and
purified using column chromatography (silica, hexanes/ethyl ace-
tate, 4:1). The title compound was obtained as a dark brown solid
reaction of an a-aminoketone (3-Et) and a 3-alkyl-substituted 2,4-
pentanedione (2-Me). The Knorr pyrrole is formed in greater
quantity than the FischereFink pyrrole both at pH 4.6 and at pH 7.
The absolute yield of the Knorr pyrrole is larger at neutral versus
acidic pH: (i) At neutral pH in the reaction with 1-amino-2-
butanone, the yield of the Knorr pyrrole (56%) was nine times
that of the FischereFink pyrrole (8%). (ii) At neutral pH in the re-
action with ALA-Me, the yield of the Knorr pyrrole (60%) was 10
times that of the FischereFink product (6.5%). The dione was stable
under the reaction conditions.
The predominance of the Knorr pyrrole versus FischereFink
pyrrole is a finding that emerged upon direct quantitative analysis
rather than on isolation, given the profound difference in crystal-
linity and stability of the trialkyl (Knorr) pyrrole and the 2-keto
(FischereFink) pyrrole. Extrapolating these results to the analo-
(22 mg, 6.6%): mp 161e163 ꢀC; 1H NMR (CDCl3, 300 MHz)
d 1.19 (t,
gous reaction an
a
-aminoketone (ALA) and a 1,3-dione (1-AcOH)
J¼7.2 Hz, 3H), 1.92 (s, 3H), 2.20 (s, 3H), 2.27 (s, 3H), 2.74 (q, J¼7.2 Hz,
suggests that formation of the FischereFink pyrrole is likely not the
culprit for the ꢁ10% yields of porphyrinogens. The Knorr pyrrole,
while evanescent, also is exceptionally reactive;24 the latter prop-
erty undoubtedly counterbalances the former property and ex-
plains the successful formation of porphyrinogens. A quantitative
understanding of the multistep process illustrated in Scheme 1 will
require examination of the individual steps other than the com-
peting Knorr and FischereFink reactions.
2H), 9.30 (br s, 1H); 1H NMR (CS2 with C6D6 as standard, 300 MHz)
d
1.48 (t, J¼7.2 Hz, 3H), 2.22 (s, 3H), 2.54 (s, 3H), 2.55 (s, 3H), 2.99 (q,
J¼7.2 Hz, 2H), 10.76 (br s, 1H); 13C NMR (CDCl3, 75 MHz)
d 8.8, 9.1,
11.7, 12.2, 33.0, 118.0, 126.5, 127.5, 132.2, 190.4; ESI-MS obsd
166.1228, calcd 166.1226 [(MþH)þ, M¼C10H15NO].
4.2.3. Ethyl
3-ethyl-4,5-dimethylpyrrole-2-carboxylate
(Kn-Et/
Es). Following the procedure of Lash,27 a suspension of diethyl
aminomalonate hydrochloride (12.5 g, 59.0 mmol) in acetic acid
(100 mL) was treated with 3-methyl-2,4-hexanedione (4, 7.5 g,
58 mmol).28 The mixture was refluxed for 3 h. The dark reaction
mixture was allowed to cool to room temperature and poured on
crushed ice. The precipitate thus formed was filtered, washed (cold
water), and recrystallized (ethanol/H2O, 4:1) to give a white solid
(6.8 g, 59%): mp 95e96 ꢀC (lit.20 mp 97e98 ꢀC); 1H NMR (CDCl3,
4. Experimental section
4.1. General section
Dimethyl 5-methoxyisophthalate was used as received from
a commercial supplier. Technical grade 3-methyl-2,4-pentanedione
(2-Me) contains 3,3-dimethyl-2,4-pentanedione as a substantial
impurity, which was difficult to purify by fractional distillation.
Complexation with copper acetate monohydrate and decom-
plexation with an acid wash followed by simple distillation pro-
vided 2-Me in 98% purity as determined by gas chromatography.25
All samples of aminoketones employed here were in the form of
hydrochloride salts. All solvents and reagents were reagent grade.
Water was deionized. All aqueous reactions were carried out in the
300 MHz)
2.18 (s, 3H), 4.29 (q, J¼7.2 Hz, 2H), 2.72 (q, J¼7.2 Hz, 2H), 8.60 (br s,
1H); 13C NMR (CDCl3, 100 MHz)
8.7, 9.1, 11.6, 14.7, 15.3, 18.7, 59.7,
d
1.10 (t, J¼7.2 Hz, 3H), 1.34 (t, J¼7.2 Hz, 3H), 1.93 (s, 3H),
d
116.1, 116.4, 129.9, 134.2, 161.9; ESI-MS obsd 196.1327, calcd
196.1332 [(MþH)þ, M¼C11H18NO2].
4.2.4. 4-Ethyl-2,3-dimethylpyrrole (Kn-Et). Following the pro-
cedure of Thompson,29 a suspension of sodium hydroxide (6 g) in

V. Chandrashaker et al. / Tetrahedron 68 (2012) 6957e6967
6965
ethylene glycol (75 mL) was treated with Kn-Et/Es (3.5 g, 18 mmol).
The reaction mixture was refluxed for 1 h under dim light. The re-
action mixture was allowed to cool to room temperature, diluted
with water (100 mL) and extracted with hexanes (3ꢃ50 mL). The
hexanes fraction was concentrated and distilled to obtain a colorless
oil (1.8 g, 82%), which solidified upon cooling to ꢄ20 ꢀC. The entire
process from reflux to distillation was carried out under a continu-
ous purge of argon. The title compound was an oil at room tem-
possible. The entire duration from initiation of the workup process
to the start of the NMR data acquisition was about 20e30 min.
4.3.3. Time course. Four reactions were setup as described in the
general procedure for pyrrole formation (see Section 4.3.2). Diethyl
carbonate was employed as the standard. The four reactions were
quenched after 24 h, 48 h, 72 h, and 96 h. Workup and 1H NMR
analysis were carried out as described in the general procedure.
perature: 1H NMR (CS2 with C6D6 as standard, 300 MHz)
d 1.42 (t,
J¼7.2 Hz, 3H), 2.15 (s, 3H), 2.41 (s, 3H), 2.61 (q, J¼7.2 Hz, each peak
4.4. X-ray crystallographic analysis
was further split by 0.9 Hz, 2H), 6.47 (d, J¼2.4 Hz,1H), 7.51 (br s,1H);
13C NMR (CS2, 75 MHz)
d 9.9,11.9,15.8, 20.2,112.6,113.7,123.8,125.8;
4.4.1. Data collection and processing. The samples (FF-Et, FF-Ph)
were mounted on a Mitegen polyimide micromount with a small
amount of Paratone N oil. All X-ray measurements were made on
a BrukereNonius Kappa Axis X8 Apex2 diffractometer at 110 K. The
frame integration was performed using SAINT.30 The resulting raw
data were scaled and absorption-corrected by multi-scan averaging
of symmetry equivalent data using SADABS.31
ESI-MS obsd 124.1122, calcd 124.1121 [(MþH)þ, M¼C8H13N].
4.2.5. 2,5-Diethylpyrazine (Et2-pyrazine). A solution of 3-Et
(1.25 mL, 1.0 M in water) was added to aqueous sodium phos-
phate buffer solution (23.75 mL, 0.2 M, pH 7) and the resulting
solution, exposed to air, was heated at 60 ꢀC for 24 h. The cooled
reaction mixture was extracted with CS2 (20 mL). The CS2 fraction
was treated with an aliquot of CS2 (625 mL, 1.0 M solution of diethyl
4.4.2. Structure solution and refinement. The structures were solved
by direct methods using the XS program.32 All non-hydrogen atoms
were obtained from the initial solution. The hydrogen atoms were
introduced at idealized positions and were treated in a mixed
fashion. The N-bound hydrogens were allowed to refine isotropi-
cally while the C-bound hydrogens were allowed to ride on the
parent carbon atom. The structural model was fit to the data using
full matrix least-squares based on F2. The calculated structure fac-
tors included corrections for anomalous dispersion from the usual
tabulation. The structure was refined using the XL program from
the SHELXTL,33 and graphic plots were produced using NRCVAX.
carbonate in CS2). The resulting mixture was analyzed by 1H NMR
spectroscopy, which depicted only the title compound apart from
the standard. The yield was 60% with respect to the known amount
of diethyl carbonate: 1H NMR (CS2, 300 MHz)
d
1.62 (t, J¼7.2 Hz,
6H), 3.07 (q, J¼7.2 Hz, 4H), 8.53 (s, 2H); 13C NMR (CS2, 75 MHz)
d
15.0, 29.7, 143.6, 156.1 (the NMR data are in accord with those
reported21 in CDCl3); ESI-MS obsd 137.1079, calcd 137.1073
[(MþH)þ, M¼C8H12N2].
4.3. Quantitation procedures
4.3.1. Pyrrole formation with dimethyl 5-methoxyisophthalate as
standard. An aqueous sodium phosphate buffer solution (4.75 mL,
0.2 M, pH 7) in a Schlenk flask was frozen on dry ice whereupon
Acknowledgements
This work was supported by a grant from the NSF Chemistry of
Life Processes Program (NSF CHE-0953010).
samples of 2-Me (30
mL, 0.25 mmol), dimethyl 5-
methoxyisophthalate (5.6 mg, 0.025 mmol), and 3-Et (250
mL,
1.0 M sodium phosphate buffer solution) were added in succession.
The flask was evacuated and purged with argon. The reaction
mixture was allowed to warm to room temperature under a slow
flow of argon. The reaction flask was then heated to 60 ꢀC with
stirring for 24 h in the dark. Under a continuous flow of argon, the
reaction mixture was diluted with Et2O (20 mL). The contents were
transferred from the Schlenk flask to a separatory funnel where-
upon water (20 mL) was added. The mixture was extracted with
Et2O (2ꢃ50 mL). The organic extract was combined, dried (Na2SO4),
and concentrated by rotary evaporation as quickly as possible un-
der dim lighting. The resulting concentrate was further diluted with
CS2 (4 mL), and a sample of the resulting solution (0.4 mL) was
transferred to an NMR tube equipped with an insert tube con-
taining C6D6. The 1H NMR spectrum was recorded as quickly as
possible. The entire duration from initiation of the workup process
to the start of the NMR data acquisition was about 1e1.5 h.
Appendix.
FischereFink pyrrole derived from 2-aminoacetophenone
FF-Ph was previously prepared by an alternative route and
characterized by 1H NMR spectroscopy in CCl4 and by mass spec-
trometry.26 Here, the reaction of dione 2-Me and 2-
aminoacetophenone (3-Ph) afforded FF-Ph and the Knorr pyrrole
Kn-Ph (Scheme 6) as determined by GCeMS analysis. A sample of
FF-Ph was isolated in 8.2% yield and characterized by spectroscopy
(1H NMR,13C NMR), ESI-MS, and single-crystal X-raycrystallography.
O
O
O
H2N
+
Ph
3-Ph
4.3.2. Pyrrole formation with diethyl carbonate as standard. An
aqueous sodium phosphate buffer solution (4.75 mL, 0.2 M, pH 7) in
a Schlenk flask was frozen on dry ice whereupon samples of 2-Me
2-Me
(30 mL, 0.25 mmol) and 3-Et (250 mL,1.0 M sodium phosphate buffer
solution) were added in succession. The flask was evacuated and
purged with argon. The reaction mixture was allowed to warm to
room temperature under a slow flow of argon. The reaction flask
was then heated to 60 ꢀC with stirring for 24 h in the dark. Under
a continuous flow of argon, the reaction mixture was diluted with
O
H
N
H
N
CS2 (5 mL) followed by addition of diethyl carbonate (125 mL of
FF-Ph
Kn-Ph
a 1.0 M solution in CS2, 0.125 mmol). A sample of this CS2 solution
(0.4 mL) was transferred to an NMR tube that contained an insert
tube of C6D6. The 1H NMR spectrum was recorded as quickly as
Scheme 6. Reaction of 2-aminoacetophenone affords the FischereFink and Knorr
pyrroles.

6966
V. Chandrashaker et al. / Tetrahedron 68 (2012) 6957e6967
Synthesis of the Knorr pyrrole
O
O
+
H
The Knorr pyrrole Kn-Et was obtained by independent syn-
thesis according to established procedures. An authentic sample of
pure Kn-Et was prepared following a procedure developed by
Lash,27 wherein condensation of 3-methyl-2,4-hexanedione (4)28
and diethyl aminomalonate under reflux in acetic acid provided
the 2-ethoxycarbonyl hemopyrrole (Kn-Et/Es). Saponification20,29
of the latter in ethylene glycol followed by distillation20 provided
the known compound Kn-Et in 82% yield (Scheme 7).
(1) LDA, THF, -78 °C
(2) Acetone, Jones reagent
75%
O
O
4
H2N
CO2Et
CO2Et
Aerobic synthesis of diethylpyrazine
59%
A sample of 3-Et was heated at 60 ꢀC for 24 h exposed to the
open air. The corresponding 2,5-diethylpyrazine (Et2-pyrazine)
was formed in 60% yield (Scheme 8). Et2-pyrazine is a known
compound previously prepared by an alternative route and char-
acterized by 1H and 13C NMR spectroscopy in CDCl3.21 Here, Et2-
pyrazine was characterized by spectroscopy (1H NMR and 13C NMR
in CS2) and by ESI-MS. These results are consistent with the well-
known sensitivity of dihydropyrazines to aerobic oxidation.34e39
AcOH, reflux, 3 h
O
H
N
Kn-Et/Es
O
Ethylene glycol, NaOH,
reflux, 1 h
82%
Single-crystal X-ray structures
H
N
The crystal structure of FF-Et exhibits dimeric molecules due to
a pair of intermolecular hydrogen bonds between the carbonyl ox-
ygen and the pyrrole NH proton (Fig. 4). The intermolecular dis-
tances between the carbonyl oxygen atoms and the pyrrole NH
Kn-Et
ꢀ
protons of FF-Et are 1.994 and 2.020 A (note that the structure of FF-
Scheme 7. Synthesis of Knorr pyrrole Kn-Et.
Et contains two molecules in the asymmetric unit); similar data
were observed for FF-Ph. These data match well with the distances
40e45
ꢀ
(typically,1.99 A) reported for similar 2-acylpyrrole derivatives.
Data for FF-Et and FF-Ph are provided in the Supplementary data.
Supplementary data
60 °C, 24 h,
aq solution, pH 7
N
O
open to air
Analysis of the reaction of 2-aminoacetophenone (3-Ph) and 3-
methyl-2,4-pentanedione (2-Me). X-ray characterization data for
pyrroles FF-Et and FF-Ph. Crystallographic data for FF-Et and FF-Ph
have been deposited with the Cambridge Crystallographic Data
Center under numbers CCDC 881529 and CCDC 881530, re-
spectively. Supplementary data related to this article can be found
online at doi:10.1016/j.tet.2012.05.080.
H2N
60%
N
3-Et
Et2-pyrazine
Scheme 8. Pyrazine formation from an aminoketone upon aerobic oxidation.
Fig. 4. Illustration of intermolecular hydrogen bonding of two molecules of FF-Et (A) and of FF-Ph (B), from single-crystal X-ray determinations.

V. Chandrashaker et al. / Tetrahedron 68 (2012) 6957e6967
6967
24. Nigst, T. A.; Westermaier, M.; Ofial, A. R.; Mayr, H. Eur. J. Org. Chem. 2008,
2369e2374.
References and notes
25. Yuasa, Y.; Kato, Y. Org. Process Res. Dev. 2002, 6, 628e631.
26. Alberola, A.; Andres, J. M.; Gonzalez, A.; Pedrosa, R.; Vicente, M. Heterocycles
1989, 29, 1973e1982.
27. Lash, T. D.; Li, W.; Quizon-Colquitt, D. M. Tetrahedron 2007, 63, 12324e12342.
28. Kalaitzakis, D.; Smonou, I. Eur. J. Org. Chem. 2012, 43e46.
29. Smithen, D. A.; Cameron, T. S.; Thompson, A. Org. Lett. 2011, 13, 5846e5849.
30. Bruker-Nonius. SAINT Version 2009.9; Bruker-Nonius: Madison, WI 53711, USA,
2009.
1. Lindsey, J. S.; Ptaszek, M.; Taniguchi, M. Origin Life Evol. Biospheres 2009, 39,
495e515.
2. Lindsey, J. S.; Chandrashaker, V.; Taniguchi, M.; Ptaszek, M. New J. Chem. 2011,
35, 65e75.
3. Taniguchi, M.; Chandrashaker, V.; Soares, A.; Lindsey, J. S. New J. Chem. 2012, 36,
1057e1069.
4. Soares, A.; Taniguchi, M.; Chandrashaker, V.; Lindsey, J. S. Chem. Sci. 2012, 3,
1963e1974.
5. Knorr, L. Ber. Dtsch. Chem. Ges. 1884, 17, 1635e1642.
6. Fischer, H.; Fink, E. Hoppe-Seylers Physiol. Chem. 1944, 280, 123e126.
7. Frydman, R. B.; Reil, S.; Frydman, B. Biochemistry 1971, 10, 1154e1160.
8. Harbuck, J. W.; Rapoport, H. J. Org. Chem. 1971, 36, 853e855.
9. Falk, H.; Hofer, O.; Lehner, H. Monatsh. Chem. 1973, 104, 925e932.
10. Butler, A. R.; George, S. D. Tetrahedron 1993, 49, 7017e7026.
11. Butler, A. R.; George, S. D. J. Chem. Soc., Perkin Trans. 2 1994, 315e318.
12. Brown, E. H.; Butler, A. R. J. Chem. Res., Synop. 1998, 458e459.
13. Bean, G. P. In Pyrroles I; Jones, R. A., Ed.; John Wiley & Sons: New York, NY, 1990;
pp 187e195.
14. Bullock, E.; Johnson, A. W.; Markham, E.; Shaw, K. B. J. Chem. Soc. 1958,
1430e1440.
15. Battey, P. K.; Crookes, D. L.; Smith, G. F. Can. J. Chem. 1979, 57, 1694e1706.
16. Paine, J. B.; Brough, J. R.; Buller, K. K.; Erikson, E. E. J. Org. Chem. 1987, 52,
3993e3997.
17. Wolfenden, R.; Snider, M. J. Acc. Chem. Res. 2001, 34, 938e945.
18. Stockbridge, R. B.; Lewis, C. A., Jr.; Yuan, Y.; Wolfenden, R. Proc. Natl. Acad. Sci. U.
S.A. 2010, 107, 22102e22105.
19. Eschenmoser, A. Tetrahedron 2007, 63, 12821e12844.
20. Gendler, P. L.; Duhan, H. A.; Rapoport, H. Clin. Chem. 1978, 24, 230e233.
21. Seeman, J. I.; Paine, J. B., III; Secor, H. V.; Im, H.-S.; Bernstein, E. R. J. Am. Chem.
Soc. 1992, 114, 5269e5280.
22. Bothner-By, A. A.; Dadok, J.; Johnson, T. E.; Lindsey, J. S. J. Phys. Chem. 1996, 100,
17551e17557.
23. Lightner, D. A.; Norris, R. D.; Kirk, D. I.; Key, R. M. Experentia 1974, 30, 587e588.
ꢁ
ꢁ
31. Bruker-Nonius. SADABS Version 2009.9; Bruker-Nonius: Madison, WI 53711,
USA, 2009.
32. Bruker-AXS. XS Version 2009.9; Bruker-AXS: Madison, WI 53711, USA, 2009.
33. Bruker-AXS. XL Version 2009.9; Bruker-AXS: Madison, WI 53711, USA, 2009.
€
34. Novo, M.; Huttmann, G.; Diddens, H. J. Photochem. Photobiol., B 1996, 34,
143e148.
35. Elfsson, B.; Wallin, I.; Eksborg, S.; Rudaeus, K.; Ros, A. M.; Ehrsson, H. Eur. J.
Pharm. Sci. 1998, 7, 87e91.
36. Bunke, A.; Zerbe, O.; Schmid, H.; Burmeister, G.; Merkle, H. P.; Gander, B. J.
Pharm. Sci. 2000, 89, 1335e1341.
37. De Blois, A. W.; Grouls, R. J. E.; Ackerman, E. W.; Wijdeven, W. J. A. Lasers Med.
Sci. 2002, 17, 208e215.
38. Hunter, G. A.; Rivera, E.; Ferreira, G. C. Arch. Biochem. Biophys. 2005, 437,
128e137.
39. Donnelly, R. F.; McCarron, P. A.; Woolfson, A. D. Perspect. Med. Chem. 2007, 1,
49e63.
40. Paine, J. B., III; Dolphin, D.; Trotter, J.; Greenhough, T. J. Can. J. Chem. 1985, 63,
2683e2690.
41. Takaya, H.; Kojima, S.; Murahashi, S.-I. Org. Lett. 2001, 3, 421e424.
42. Senge, M. O.; Smith, K. M. Acta Crystallogr. 2005, C61, o537eo541.
43. Salzameda, N. T.; Lightner, D. A. Monatsh. Chem. 2006, 137, 1453e1462.
ꢂ
44. Camarillo, E. A.; Flores, H.; Amador, P.; Bernes, S. Acta Crystallogr. 2007, E63,
o2593eo2594.
45. Ali, A. A.-S.; Cipot-Wechsler, J.; Crawford, S. M.; Selim, O.; Stoddard, R. L.;
Cameron, T. S.; Thompson, A. Can. J. Chem. 2010, 88, 725e735.