An efficient route to emodic amine and analogous O-methyl protected derivatives starting from emodin

Article abstract of DOI:10.1007/s00706-005-0350-7

Emodic amine could be synthesized in a five-step approach in excellent overall yield by following a modified Curtius rearrangement strategy, starting from the naturally occurring emodin. This unique emodin derived 6-amino substituted polyhydroxylated anthraquinone may serve as a promising synthon for a new class of amino functionalized photodynamically active hypericin derivatives. In addition, the partially O-methyl protected 6-amino- and 6-carboxy-anthraquinones could be synthesized in high yields via selective O-methyl ether cleavage from the corresponding tri-O-methyl derivatives. Springer-Verlag 2005.

Full text of DOI:10.1007/s00706-005-0350-7

Monatshefte f€ur Chemie 136, 1629–1639 (2005)
DOI 10.1007/s00706-005-0350-7
An Efficient Route to Emodic Amine
and Analogous O-Methyl Protected
Derivatives Starting from Emodin
ꢀ
Bernd Lackner, Klaus Bretterbauer, and Heinz Falk
Institute of Organic Chemistry, Johannes Kepler University Linz, 4040 Linz, Austria
Received May 20, 2005; accepted May 25, 2005
Published online August 10, 2005 # Springer-Verlag 2005
Summary. Emodic amine could be synthesized in a five-step approach in excellent overall yield by
following a modified Curtius rearrangement strategy, starting from the naturally occurring emodin.
This unique emodin derived 6-amino substituted polyhydroxylated anthraquinone may serve as a
promising synthon for a new class of amino functionalized photodynamically active hypericin deriv-
atives. In addition, the partially O-methyl protected 6-amino- and 6-carboxy-anthraquinones could be
synthesized in high yields via selective O-methyl ether cleavage from the corresponding tri-O-methyl
derivatives.
Keywords. Anthraquinones; Curtius rearrangement; Emodic amine; Selective O-methyl ether cleav-
age; Microwave assisted synthesis.
Introduction
The polyhydroxylated anthraquinone emodin (1, 1,3,8-trihydroxy-6-methyl-9,10-
anthraquinone) is a wide-spread natural product occurring in fungi, lichens, and a
variety of plants (e.g. rhubarb, aloe) which exhibits tumor cell-growth inhibition
[1], antibiotic [2], anti-viral [3], and cytostatic [4] activity. There is a continued
interest in the syntheses of new anthraquinone derivatives, either as promising bio-
logically active compounds or as valuable synthons for the preparation of photo-
dynamically active hypericin derivatives, which is one of the main interests in our
research group [5]. Polyhydroxylated anthraquinones bearing amino groups appear
to be found rather seldom in nature. Thus, to the best of our knowledge 5-amino-
physcion (2, 5-amino-1,8-dihydroxy-3-methoxy-6-methyl-9,10-anthraquinone [6])
isolated by Steglich et al. [7] from the fungal species Dermocybe canaria consti-
tutes the only natural representative of an amino substituted emodin derivative (see
Fig. 1). Further examples for amino substituted anthraquinones are the powerful
ꢀ
Corresponding author. E-mail: heinz.falk@jku.at

1630
B. Lackner et al.
Fig. 1. Structures of emodin (1), 5-aminophyscion (2), mitoxantrone (3), and emodic amine (4)
cytostatic drug mitoxantrone (3) as well as a series of highly cytotoxic and cyto-
static 4-amino analogues of 1 recently synthesized by Teich et al. [8]. In addition,
there exists a wide range of highly functionalized mainly 1,4- or 1,8-bisamino
substituted synthetic dyes [9].
The transformation of the methyl group in position 6 of emodin (1) into the
amino functionality of 1,3,8-trihydroxy-6-amino-9,10-anthraquinone (4, emodic
amine), gaining access to a new class of amino anthraquinones, was targeted (see
Fig. 1). The novel emodic amine (4) might act as a valuable synthon for the
preparation of a promising new class of amino substituted photodynamically active
hypericin derivatives. Furthermore, the syntheses of the corresponding O-methyl
protected analogues of 4 represent a second target of this work.
Results and Discussion
To the best of our knowledge, up to now no efforts have been undertaken to convert
the methyl group of emodin (1) into the amino functionality. In principle, there are
several classical synthesis strategies conceivable allowing access to amino deriv-
atives, like the Curtius [10], Hofmann [11], Lossen [12], and Schmidt [13] rearrange-
ments. Due to the fact that all of these strategies require the corresponding
carboxylic acid as starting material, our first target was the transformation of the
methyl group of 1 to the carboxylic group. Emodin (1), isolated from natural source
(Cortex frangulae) [14], was protected as the 1,3,8-trimethoxy-6-methyl-9,10-
anthraquinone (5) by microwave (MW) assisted [15] or conventional [16] synthesis
both in 98% yield. The oxidation of 5 to 1,3,8-trimethoxy-6-carboxy-9,10-anthra-
quinone (6, tri-O-methyl emodic acid) was performed by using tetrabutylammonium
permanganate (TBAP) [17] in pyridine at 80ꢁC for 2 h in 88% yield. It should be
stressed, that the initiation of this reaction may vary depending on the scale as well
as the moisture content of the oxidizing reagent and therefore the progress of the
reaction has to be controlled by TLC to avoid overoxidation. In contrast to Di
Napoli’s multistep synthesis [18] of 6, which is based on a total synthesis anthra-
quinone ring closure strategy, we were able to obtain 6 in an efficient two step syn-
thesis starting from 1 with an overall yield of 86% (see Fig. 2).
The conversion of tri-O-methyl emodic acid (6) to 1,3,8-trimethoxy-6-amino-
9,10-anthraquinone (8, tri-O-methyl emodic amine), via the carbamic ester 7 was
performed by a modified one-step Curtius rearrangement [19]. For this purpose
diphenylphosphoryl azide (DPPA), which was first reported by Shioiri et al. [20],
constitutes a convenient reagent for the conversion of carboxylic acids to the

Synthesis of Emodic Amine by Curtius Rearrangement
1631
Fig. 2. Synthesis of tri-O-methyl emodic amine (8) from emodin (1): (a) cf. Ref. [15]: Me2SO4=
K2CO3=tetrabutylammonium bromide, MW (600 W), 75ꢁC, 20 min; (b) TBAP=pyridine, 80ꢁC, 2 h;
(c) dry dioxane=TEA=DPPA=TMSE, 80ꢁC, 15 h; (d) MeOH=KOH (1 M)=H2O, reflux, 12 h
corresponding carbamates followed by the in situ treatment with 2-(trimethylsilyl)
ethanol (TMSE) to the intermediate carbamic esters. In analogy to this route, 6 was
treated with triethylamine (TEA) and DPPA in dry dioxane at 80ꢁC for 2.5 h. After
addition of TMSE and heating at 80ꢁC for further 12 h the generated carbamate was
converted in situ to the intermediate carbamic ester 1,3,8-trimethoxy-6-[(trimethyl-
silylethoxycarbonyl)amino]-9,10-anthraquinone (7) in quantitative yield. Finally, 7
was hydrolyzed with KOH=MeOH=H2O under reflux for 12 h to obtain the tri-O-
methyl protected amine 8 in 92% yield based on 6 after column chromatography.
It is noteworthy, that the main impurity constitutes an unpolar yellow colored
fraction consisting of unhydrolyzed 7, which could be recycled. Interestingly
enough, full conversion of 7 to 8 could not be achieved at any time and it should
be noted, that this behavior seems to be a characteristic of this reaction. Accord-
ingly, tri-O-methyl emodic amine (8) could be obtained from emodin (1) in a four-
step synthesis with an overall yield of 79% (see Fig. 2). It might be also noted,
that the application of the Hofmann rearrangement was put aside due to the rather
low yield of 49% obtained in the synthesis of 1,3,8-trimethoxy-6-carbamoyl-9,10-
anthraquinone (9) starting from 6. The possible syntheses of 8 via a Lossen or
Schmidt strategy, which seems also feasible was not performed because of the
known disadvantages, like low yield protocols and unpleasant reagents [12, 13].
Thus, the Curtius rearrangement turned out to be the most efficient method for the
preparation of 8.
Selective O-Methyl Ether Cleavage
To reach our main target, which constitutes the synthesis of emodic amine (4), full
O-methyl deprotection of 8 has to be performed. In analogy to Steglich et al. [21]

1632
B. Lackner et al.
Fig. 3. Selective deprotection of tri-O-methyl emodic amine (8) to emodic amine (4), di-O-methyl
emodic amine (10), and mono-O-methyl emodic amine (11): (a) pyridinium chloride, 150ꢁC, 6 h; (b)
pyridinium chloride, MW (600 W), reflux, 2 min; (c) BBr3=CH2Cl2, ꢂ7ꢁC to 1ꢁC, 1 h; (d) HBr=
AcOH, reflux, 1 h; (e) HBr, reflux, 30 min
and previous results of our group [22], the total deprotection of 1,3,8-trimethoxy
anthraquinones by means of BBr3 seemed to be suitable for this purpose. In this
specific case, the treatment of 8 with an excess of BBr3 starting from 0ꢁC to room
temperature within 2 h, followed by 2 h reflux afforded 4 in yields <26%, with
1,3-dimethoxy-8-hydroxy-6-amino-9,10-anthraquinone (10, di-O-methyl emodic
amine) as the main product. Neither modifications concerning the molar excess
of reagent nor the temperature as well as reaction time could improve this result.
Furthermore, Brockmann’s deprotection [23] of methyl ethers using KI=H3PO4
under rather harsh conditions caused only destruction of 8. Finally, application
of Hassall’s deprotection [24] using a pyridinium chloride melt at 150ꢁC for 6 h
turned out to be the method of choice for the complete O-methyl ether cleavage of
8 to 4 in 96% yield (see Fig. 3).
In addition, an improved variation of Hassall’s method [24] could be achieved
by using pyridinium chloride in combination with microwave assisted synthesis.
Thus, a mixture of 8 and pyridinium chloride was molten in the microwave unit
(600 W) within 1 min and kept at gentle reflux (225ꢁC) for further 1 min to obtain 4
in 98% yield. All in all, the promising synthon and interesting amino analogue of
emodin (1), the target compound 4, could be obtained from 1 in an efficient five-
step synthesis with an excellent overall yield of 78%.
Beside the synthesis of tri-O-methyl emodic amine (8) and its full deprotection
to 4 mentioned above, we also investigated the selective mono- and dideprotection
of 8 to the corresponding di-O-methyl emodic amine (10) as well as 1,8-dihydroxy-3-
methoxy-6-amino-9,10-anthraquinone (11, mono-O-methyl emodic amine), shown
in Fig. 3. In analogy to our recently published improved procedure [22] for the selec-
tive mono-ether cleavage of 1,3,8-trimethoxyanthraquinones, 8 was converted to 10
by means of BBr3=CH2Cl2 (ꢂ7ꢁC to 1ꢁC, 1 h) in 91% yield. Finally, the mono-O-

Synthesis of Emodic Amine by Curtius Rearrangement
1633
Fig. 4. Selective deprotection of tri-O-methyl emodic acid (6) to di-O-methyl emodic acid (12) and
mono-O-methyl emodic acid (13): (a) BBr3=CH2Cl2, ꢂ7ꢁC to 1ꢁC, 1 h; (b) HBr=AcOH, reflux, 1.5 h;
(c) HBr, reflux, 1 h
methyl protected amine 11 was prepared in high yield via selective methyl ether
cleavage of 8 using either the system HBr=AcOH under reflux for 1 h (89%) or
HBr under reflux for 30 min (98%). Both partially O-methyl protected emodic amines
10 and 11 could be synthesized in five steps with excellent overall yields (>72%)
starting from 1.
For the realization of the abovementioned efficient route to emodic amine (4)
and its O-methyl protected derivatives 8, 10, and 11, tri-O-methyl emodic acid (6)
served as a key intermediate. In addition, 6 also turned out to act as a valuable
starting material for the first synthesis of 1,3-dimethoxy-8-hydroxy-6-carboxy-
9,10-anthraquinone (12, di-O-methyl emodic acid) by using BBr3=CH2Cl2
(ꢂ7ꢁC to 1ꢁC, 1 h) in 90% yield (see Fig. 4). The synthesis of 1,8-dihydroxy-3-
methoxy-6-carboxy-9,10-anthraquinone (13, mono-O-methyl emodic acid) has
already been reported by Bloomer et al. [25]. However, these authors followed
an inconvenient multistep cycloaddition approach. Starting from 6, we were now
able to efficiently synthesize 13 via a selective O-methyl ether cleavage, using
either the system HBr=AcOH under reflux for 1 h (93%) or HBr under reflux in
1 h (81%) in excellent yields. Accordingly, we also developed a high yield protocol
for the three step syntheses of the partially O-methyl protected emodic acid deriv-
atives 12 and 13, starting from the natural compound emodin (1) in overall yields
of >78%.
Conclusion
An efficient route to emodin derived 6-amino anthraquinones was achieved.
This was realized following a Curtius rearrangement strategy in excellent over-
all yields starting from the natural product emodin (1). The main target emodic
amine (4) might act as a promising precursor for an interesting new class of
amino functionalized hypericin derivatives. In addition, the polyhydroxylated
amino anthraquinone 4 and its analogous O-methyl protected derivatives 8, 10,
and 11 might possess bioactive relevance. Compounds 4 and 7–12 (see Table 1)
could be synthesized for the first time and were fully characterized on the basis of
1
their IR, UV=Vis, mass, H and 13C NMR spectra, particularly by 2D NMR
measurements including HSQC, HMBC, and NOESY experiments. In contrast
to inconvenient multistep cycloaddition approaches described in literature, com-
pounds 6 and 13 could also be efficiently synthesized in excellent yields starting
from 1.

1634
B. Lackner et al.
Table 1. Overview concerning novel emodin derived anthraquinone derivatives
Entry
R1
R2
R3
R4
Yield
%
t=T
mp=ꢁC
>250e
Rf
4
H
H
NH2
H
96a,
98b
88
6 h=150ꢁCa,
0.42g
MW=600 W=2 minb
2 h=80ꢁC
6
7
CH3
CH3
CH3
CH3
COOH
CH3
CH3
242–244
>180e
0.36h
0.85i
c
NHCOO(CH2)2–
–Si(CH3)3
NH2
–
15 h=80ꢁC
8
9
CH3
CH3
CH3
H
CH3
CH3
CH3
CH3
CH3
CH3
H
92d
49
12 h=80ꢁC
253–255
>315e
0.52g
0.43j
0.73g
0.79g
CONH2
NH2
5 h=5ꢁC to reflux
1 h=ꢂ7 to 1ꢁC
1 h=refluxa,
10
11
91
257–258
266–268
NH2
H
89a,
98b
90
0.5 h=refluxb
1 h=ꢂ7 to 1ꢁC
1.5 h=refluxa,
1 h=refluxb
12
13
CH3
H
CH3
CH3
COOH
COOH
H
H
>278e
>298e, f
0.10g
0.08g
93a,
81b
Method A (see experimental part); b method B (see experimental part); c intermediate; d based on 6; e decomposition;
Ref. [25] 312ꢁC (sealed tube); g CHCl3:MeOH ¼ 10:1; h CHCl3:EtOH ¼ 1:2; i CHCl3:EtOH ¼ 10:1; j CHCl3:AcOEt ¼
a
f
10:1
Experimental
All solvents were of p.a. quality and dried by conventional means if necessary. Reagents were
supplied by commercial sources and were used without further purification. Melting points were
1
measured on a Kofler melting point microscope (Reichert, Vienna). H NMR spectra were re-
corded on a Bruker Avance DRX 500 MHz spectrometer using a TXI cryoprobe with z-gradient
coil. All NMR experiments were performed with DMSO-d6 as the solvent at 30ꢁC. 13C NMR
spectra and 2D NMR experiments were performed using standard pulse sequences as provided by
the manufacturer. Typical 90ꢁ hard pulse durations were 8.2 ꢀs (1H) and 16.6 ꢀs (13C), 90ꢁ pulses
in decoupling experiments were set to 67 ꢀs. HSQC and HMBC experiments were optimized for
coupling constants of 145 Hz for single quantum correlations and 10 Hz for multi-bond correla-
tions. NOESY mixing time was set to 400 ms. IR, UV=Vis, and mass spectra were recorded using
the Bruker Tensor 27, Varian Cary 100 Bio UV=Vis, Hewlett Packard 5989 quadrupole, and Fisons
MD 800 instruments. Microwave assisted synthesis was performed on an MLS-ETHOS 1600
microwave unit with Terminal 320 from MLS-Milestone. 1,3,8-Trihydroxy-6-methyl-9,10-anthra-
quinone (1, emodin) was isolated from Cortex frangulae according to Ref. [14]. 1,3,8-Trimethoxy-
6-methyl-9,10-anthraquinone (5, tri-O-methyl emodin) was prepared by means of microwave
assisted synthesis according to Ref. [15]. Column chromatography was carried out on silica gel
0.060–0.200 mm (pore diameter 6 nm). All novel compounds were judged to be pure (>97%) by
1
means of their H NMR spectra and chromatography.
1,3,8-Trihydroxy-6-amino-9,10-anthraquinone (4, C14H9NO5)
Method A: A mixture of 6.0 mg (0.0192 mmol) 8 and 500 mg (4.33 mmol) pyridinium chloride was
molten at 150ꢁC under Ar and kept at this temperature for 6 h. After cooling the solid was taken up

Synthesis of Emodic Amine by Curtius Rearrangement
1635
with distilled H2O and extracted with AcOEt. The organic layer was washed twice with distilled H2O,
dried over Na2SO4, filtered, evaporated to dryness, and chromatographed (CHCl3:MeOH¼ 10:1) to
afford 5.0mg (0.0184 mmol, 96%) 4 as a red solid.
Method B (microwave assisted synthesis): A mixture of 6.0 mg (0.0192mmol) 8 and 1.0 g
(8.65 mmol) pyridinium chloride was irradiated in the microwave unit for 2 min at 600 W under
Ar. The mixture was molten within the first minute followed by gentle reflux (t ꢃ 225ꢁC) for the second
minute. After cooling the work-up was performed according to method A to afford 5.1 mg (0.0188mmol,
98%) 4 as a red solid. Mp >250ꢁC (decomp.); TLC: Rf ¼ 0.27 (CHCl3:AcOEt ¼ 1:1), Rf ¼ 0.42
1
(CHCl3:MeOH ¼ 10:1); H NMR (500 MHz): ꢁ ¼ 12.46 (s, 1-OH), 12.29 (s, 8-OH), 11.08 (sbr, 3-
OH), 7.07 (s, ar-H4), 7.01 (s, ar-H5), 6.94 (sbr, 6-NH2), 6.55 (s, ar-H2), 6.23 (s, ar-H7) ppm; 13C
NMR (125MHz): ꢁ ¼ 186.8 (C9), 182.0 (C10), 164.4 (C8), 164.2 (C3), 163.8 (C1), 156.7 (C6), 134.9
(C4a), 134.5 (C10a), 108.8 (C9a), 108.2 (C5), 108.1 (C2 and C4), 104.9 (C8a), 102.4 (C7) ppm; ESI-MS
(MeOHþ 1% NH3; c ꢃ 1 mg ꢄ cmꢂ3, negative ion mode): m=z ¼ 270 ([M–H]ꢂ); IR (KBr): ꢂꢀ¼ 3469,
3371, 3237, 2924, 2853, 1727, 1622, 1468, 1402, 1324, 1291, 1267, 1214, 1157, 1113, 1026, 1004, 913,
849, 773, 665 cmꢂ1; UV-Vis (CHCl3, c ¼ 5.53 ꢄ 10ꢂ5 mol ꢄ dmꢂ3): ꢃmax(") ¼ 258 (9660), 303 (13770),
478 (2600) nm (dm3 ꢄ molꢂ1 ꢄ cmꢂ1); UV-Vis (MeOH, c ¼ 5.53 ꢄ 10ꢂ5 mol ꢄ dmꢂ3): ꢃmax(") ¼ 204
(12310), 224 (12770), 258 (12180), 307 (14510), 385 (1680), 491 (3160) nm (dm3 ꢄ molꢂ1 ꢄ cmꢂ1).
1,3,8-Trimethoxy-6-carboxy-9,10-anthraquinone (6)
A solution of 1.548 g (4.80 mmol) tetrabutylammonium bromide in 3 cm3 distilled H2O was added
dropwise to a stirred solution of 0.759g (4.80mmol) finely ground KMnO4 in 9 cm3 distilled H2O.
The resulting purple highly viscous mixture was stirred for 1 h at room temperature and the produced
tetrabutylammonium permanganate (TBAP) was filtered, washed twice with distilled H2O, and quickly
treated with freshly dried diethyl ether. Within 1 h a solution of 1.497 g (4.14mmol) TBAP in 22 cm3
pyridine was added to a solution of 251.0mg (0.810mmol) 5 in 7 cm3 pyridine at 80ꢁC under Ar. The
reaction was controlled by TLC, and after 125 min the mixture was cooled and 2.5 g (13.15mmol) of
ground Na2S2O5 were added. After stirring for 15min on an ice bath, the mixture was evaporated to
dryness and the residue was taken up in 20cm3 distilled H2O. The brown suspension was acidified to
pH¼ 1 by adding concentrated HCl and the precipitate was centrifuged, washed three times with
distilled H2O, dried under vacuum over P2O5, and chromatographed (CHCl3:EtOH¼ 3:1) to afford
115.1 mg (0.336mmol, 42%) 6. A further crop of 127.0 mg (0.371 mmol, 46%) 6 could be isolated by
extraction of the second and third yellow colored aqueous washing solutions with AcOEt to yield a
total amount of 242.1 mg (0.707mmol, 88%) 6 as a yellow solid. Mp 242–244ꢁC; TLC: Rf ¼ 0.36
1
(CHCl3:EtOH¼ 1:2), Rf ¼ 0.0 (CHCl3:AcOEt¼ 1:2); H NMR (500 MHz): ꢁ ¼ 13.60 (sbr, 6-COOH),
8.19 (s, ar-H5), 7.91 (s, ar-H7), 7.22 (d, J ¼ 2.1Hz, ar-H4), 7.02 (d, J ¼ 2.1 Hz, ar-H2), 3.98 (s, 8-
OCH3), 3.96 (s, 3-OCH3), 3.92 (s, 1-OCH3) ppm; 13C NMR (125MHz): ꢁ ¼ 182.5 (C10), 179.5 (C9),
165.7 (6-COOH), 163.1 (C3), 161.2 (C1), 158.9 (C8), 135.6 (C6), 135.1 (C10a or C4a), 134.2 (C4a or
C10a), 126.1 (C8a), 118.6 (C5), 118.5 (C7), 117.6 (C9a), 105.1 (C2), 102.5 (C4), 56.41 (8-OCH3),
56.36 (1-OCH3), 55.88 (3-OCH3) ppm; NCI-MS (solid probe, CH4): m=z ¼ 342 ([M]ꢂ); IR (KBr):
ꢂꢀ¼ 3486, 3091, 2922, 2850, 1723, 1655, 1598, 1562, 1460, 1413, 1331, 1235, 1071, 1012, 948, 873,
750 cmꢂ1; UV-Vis (DMSO, c ¼ 1.24 ꢄ 10ꢂ5 mol ꢄ dmꢂ3): ꢃmax(") ¼ 276 (85890), 405 (3950) nm
(dm3 ꢄ molꢂ1 ꢄ cmꢂ1); UV-Vis (CHCl3): ꢃmax(A) ¼ 274 (100), 330 (33), 402 (27) nm (rel. int.).
1,3,8-Trimethoxy-6-[(trimethylsilylethoxycarbonyl)amino]-9,10-anthraquinone
(7, C23H27NO7Si)
To a solution of 48.0 mg (0.140 mmol) 6 dissolved in 1.5 cm3 dry 1,4-dioxane at 80ꢁC under Ar,
25.6mg (0.25 mmol) triethylamine were added and stirred for 15min. After addition of 64.0mg
(0.230 mmol) diphenylphosphoryl azide the reaction mixture was stirred for further 2.5h followed
by the addition of 54.0 mg (0.460 mmol) 2-(trimethylsilyl)ethanol. The resulting mixture was kept at
80ꢁC for 12 h, cooled, and evaporated to dryness to obtain the intermediate carbamic ester 7 in
quantitative yield as a yellow solid. At that point it should be mentioned, that the isolation of 7 was

1636
B. Lackner et al.
performed once for full analytical and spectroscopic characterization. From the synthesis point of
view, 7 hydrolyses to 8 even during the work-up (e.g. evaporation of the solvent at elevated tempera-
ture) and therefore 7 acts only as an intermediate in the conversion of 6 to 8. Mp >180ꢁC (decomp.);
TLC: Rf ¼ 0.85 (CHCl3:EtOH¼ 10:1), Rf ¼ 0.43 (CHCl3:AcOEt¼ 1:1); 1H NMR (500MHz): ꢁ ¼ 7.97
(s, ar-H7), 7.42 (s, ar-H5), 7.33 (s, ar-H4), 6.96 (s, 6-NHCO-), 6.79 (s, ar-H2), 4.32 (t, J ¼ 8.3 Hz,
–OCH2–), 4.03 (s, 8-OCH3), 3.98 (s, 1-OCH3), 3.96 (s, 3-OCH3), 1.09 (t, J ¼ 8.3 Hz, –CH2Si–), 0.10
(s, –Si(CH3)3) ppm; 13C NMR (125 MHz): ꢁ ¼ 184.1 (C10), 181.0 (C9), 163.9 (C3), 162.1 (C1), 161.7
(C8), 153.4 (–NHCO–), 143.3 (C6), 136.5 (C4a), 135.6 (C10a), 119.2 (C8a), 118.7 (C9a), 107.9
(C5), 107.3 (C7), 105.8 (C2), 102.3 (C4), 64.50 (–O–CH2–), 57.75 (1-OCH3 or 8-OCH3), 56.72
(8-OCH3 or 1-OCH3), 56.07 (3-OCH3), 17.93 (–CH2–Si–), ꢂ1.27 (–Si(CH3)3) ppm; NCI-MS (solid
probe, CH4): m=z ¼ 457 ([M]ꢂ); IR (KBr): ꢂꢀ¼ 3337, 3086, 2926, 2854, 1728, 1663, 1586, 1531,
1457, 1336, 1255, 1221, 1141, 1061, 1010, 983, 956, 925, 892, 857, 759 cmꢂ1; UV-Vis (CHCl3,
c ¼ 1.01ꢄ 10ꢂ5 mol ꢄ dmꢂ3): ꢃmax(") ¼ 287 (17440), 416 (2190) nm (dm3 ꢄ molꢂ1 ꢄ cmꢂ1).
1,3,8-Trimethoxy-6-amino-9,10-anthraquinone (8, C17H15NO5)
The intermediate carbamic ester 7 was dissolved in 100 cm3 1 M KOH in methanol containing
100 mm3 distilled H2O and refluxed under Ar for 12h. The conversion was controlled by TLC.
After cooling the mixture was neutralized with 2 M HCl on an ice bath under stirring followed by
evaporation of the solvent. The residue was dissolved in CHCl3 and washed twice with distilled H2O.
The organic layer was dried (Na2SO4), filtered, evaporated, and chromatographed (CHCl3:
MeOH¼ 10:1) to afford 40.4 mg (0.129mmol, 92% based on 6) 8 as a red solid. Mp 253–255ꢁC;
TLC: Rf ¼ 0.64 (CHCl3:EtOH¼ 5:1), Rf ¼ 0.52 (CHCl3:MeOH ¼ 10:1), Rf ¼ 0.08 (CHCl3:AcOEt¼
1:1); 1H NMR (500MHz): ꢁ ¼ 7.14 (d, J ¼ 2.5Hz, ar-H4), 6.93 (d, J ¼ 2.5Hz, ar-H2), 6.89 (d,
J ¼ 2.2 Hz, ar-H5), 6.53 (d, J ¼ 2.2 Hz, ar-H7), 6.33 (sbr, 6-NH2), 3.91 (s, 3-OCH3), 3.86 (s, 1-OCH3),
3.78 (s, 8-OCH3) ppm; 13C NMR (125 MHz): ꢁ ¼ 184.0 (C10), 178.5 (C9), 162.6 (C3), 161.4 (C8), 160.9
(C1), 153.8 (C6), 135.5 (C4a or C10a), 135.3 (C10a or C4a), 118.0 (C9a), 112.4 (C8a), 105.2 (C2), 103.4
(C5), 101.9 (C4), 101.5 (C7), 56.25 (1-OCH3), 55.71 (3-OCH3), 55.57 (8-OCH3) ppm; ESI-MS
(MeOH:DMSO¼ 5:1 þ 2% CF3COOH, c ꢃ 1 mgꢄ cmꢂ3, positive ion mode): m=z ¼ 314 ([M þ H]þ);
NCI-MS (solid probe, CH4): m=z ¼ 313 ([M]ꢂ); IR (KBr): ꢂꢀ¼ 3476, 3352, 3228, 2924, 2852, 1637,
1604, 1558, 1440, 1353, 1320, 1268, 1149, 1022, 976, 945, 881, 834, 662 cmꢂ1; UV-Vis (DMSO,
c ¼ 1.51ꢄ 10ꢂ5 mol ꢄ dmꢂ3): ꢃmax(") ¼ 266 (7660), 304 (10700) nm (dm3 ꢄ molꢂ1 ꢄ cmꢂ1); UV-Vis
(CHCl3, c ¼ 1.21 ꢄ 10ꢂ5 mol ꢄ dmꢂ3): ꢃmax(") ¼ 294 (18060), 439 (1400) nm (dm3 ꢄ molꢂ1 ꢄ cmꢂ1).
1,3,8-Trimethoxy-6-carbamoyl-9,10-anthraquinone (9, C18H15NO6)
To a stirred suspension of 60.9mg (0.178mmol) 6 in 40cm3 dry benzene at 5–10ꢁC under Ar, a
solution of 200 mm3 (2.32 mmol) oxalyl chloride in 10 cm3 benzene was added within 30min.
After warming up to room temperature within 30min the mixture was refluxed for 3 h, followed by
cooling and evaporation of the solvent. The yellow acid chloride was dissolved in 50cm3 benzene
and cooled to 5–10ꢁC. A solution of 215mm3 (3.33mmol) NH3 (32% aq) in 3 cm3 distilled H2O
was added slowly under stirring within 15 min. After further 15min of stirring and warming up to room
temperature, the mixture was stirred for another 30min. The yellow precipitate was filtered, washed
three times with distilled H2O, and dried under vacuum over P2O5 to afford 29.6mg (0.087 mmol,
49%) 9. Mp >315ꢁC (decomp.); TLC: Rf ¼ 0.83 (CHCl3:MeOH¼ 5:2), Rf ¼ 0.43 (CHCl3:EtOH¼
10:1); 1H NMR (500 MHz): ꢁ ¼ 8.35 (sbr, 6-CONHꢄ), 8.17 (d, J ¼ 1.2Hz, ar-H5), 7.90 (d, J ¼ 1.2 Hz,
ar-H7), 7.68 (sbr, 6-CONHꢅ), 7.21 (d, J ¼ 2.4 Hz, ar-H4), 7.01 (d, J ¼ 2.4 Hz, ar-H2), 3.96 (s, 8-OCH3),
3.95 (s, 3-OCH3), 3.91 (s, 1-OCH3) ppm; 13C NMR (125MHz): ꢁ ¼ 182.8 (C10), 179.7 (C9), 166.1
(6-CONH2), 163.6 (C3), 161.1 (C1), 158.8 (C8), 138.7 (C6), 135.6 (C4a or C10a), 134.0 (C10a or
C4a), 125.0 (C8a), 117.6 (C9a), 117.4 (C7), 117.2 (C5), 105.0 (C2), 102.4 (C4), 56.49 (1-OCH3 or 8-
OCH3), 56.39 (8-OCH3 or 1-OCH3), 55.93 (3-OCH3) ppm; ESI-MS (MeOH:DMSO¼ 3:1 þ 2%
CF3COOH, c ꢃ 1 mgꢄ cmꢂ3, positive ion mode): m=z ¼ 342 ([M þ H]þ); IR (KBr): ꢂꢀ¼ 3389, 3149,
3025, 2955, 2925, 2854, 1736, 1667, 1627, 1597, 1564, 1493, 1466, 1453, 1410, 1359, 1326, 1301,

Synthesis of Emodic Amine by Curtius Rearrangement
1637
1242, 1199, 1163, 1126, 1074, 1013, 948, 868, 836, 753, 698 cmꢂ1; UV-Vis (DMSO, c ¼
2.23ꢄ 10ꢂ5 mol ꢄ dmꢂ3): ꢃmax(") ¼ 274 (14710), 401 (4080) nm (dm3 ꢄ molꢂ1 ꢄ cmꢂ1); UV-Vis (CHCl3,
c ¼ 6.89ꢄ 10ꢂ5 mol ꢄ dmꢂ3): ꢃmax(") ¼ 277 (4230), 406 (1220) nm (dm3 ꢄ molꢂ1 ꢄ cmꢂ1).
1,3-Dimethoxy-8-hydroxy-6-amino-9,10-anthraquinone (10, C16H13NO5)
A stirred solution of 6.9 mg (0.0220 mmol) 8 in 1.5 cm3 CH2Cl2 was flushed with Ar and cooled to
ꢂ7ꢁC. To this solution 8.7 mm3 (0.0903 mmol) BBr3 were added through a septum by means of a
syringe and the mixture was stirred for 1 h (Tend ¼ 1ꢁC). The reaction mixture was poured into ice=H2O
and extracted with CHCl3. The organic layer was dried (Na2SO4), filtered, and evaporated to dryness
to yield 6.0 mg (0.0200 mmol, 91%) 10 as a red solid. Mp 257–258ꢁC; TLC: Rf ¼ 0.45
1
(CHCl3:AcOEt¼ 1:1), Rf ¼ 0.73 (CHCl3:MeOH¼ 10:1); H NMR (500 MHz): ꢁ ¼ 13.58 (s, 8-OH),
7.26 (d, J ¼ 1.9 Hz, ar-H4), 6.97 (d, J ¼ 1.9Hz, ar-H2), 6.92 (d, J ¼ 1.7 Hz, ar-H5), 6.61 (sbr, 6-NH2),
6.24 (d, J ¼ 1.7 Hz, ar-H7), 3.95 (s, 3-OCH3), 3.92 (s, 1-OCH3) ppm; 13C NMR (125MHz): ꢁ ¼ 184.3
(C9), 182.7 (C10), 164.6 (C8), 164.0 (C3), 162.5 (C1), 155.4 (C6), 136.5 (C4a), 133.4 (C10a), 114.4
(C9a), 106.6 (C8a), 106.0 (C5), 104.7 (C2), 104.2 (C4), 103.3 (C7), 56.44 (1-OCH3), 55.92 (3-OCH3)
ppm; ESI-MS (MeOH:DMSO¼ 6:1 þ 1% HCOOH, c ꢃ 1 mgꢄ cmꢂ3, positive ion mode): m=z ¼ 300
([M þ H]þ); IR (KBr): ꢂꢀ¼ 3458, 3357, 3237, 2923, 2852, 1730, 1671, 1618, 1595, 1557, 1495, 1452,
1424, 1405, 1348, 1325, 1298, 1266, 1244, 1220, 1190, 1149, 1061, 1032, 1009, 948, 888, 842, 812,
753 cmꢂ1; UV-Vis (CHCl3, c ¼ 1.20ꢄ 10ꢂ5 mol ꢄ dmꢂ3): ꢃmax(") ¼ 255 (15050), 300 (20830), 465
(2260) nm (dm3 ꢄ molꢂ1 ꢄ cmꢂ1); UV-Vis (MeOH, c ¼ 1.20ꢄ 10ꢂ5 mol ꢄ dmꢂ3): ꢃmax(") ¼ 221 (21380),
241 (13880), 256 (14210), 305 (16820), 491 (1700) nm (dm3 ꢄ molꢂ1 ꢄ cmꢂ1).
1,8-Dihydroxy-3-methoxy-6-amino-9,10-anthraquinone (11, C15H11NO5)
Method A: To a refluxing solution of 6.9 mg (0.0220 mmol) 8 in 3 cm3 of glacial acetic acid under Ar,
350 mm3 HBr (47% aq) were added and heated for 60min. The mixture was cooled, poured into
ice=H2O, and neutralized with saturated NaHCO3 solution. After extraction with CHCl3 the organic
layer was dried (Na2SO4), filtered, evaporated to dryness, and chromatographed (CHCl3:MeOH¼
10:1) to yield 5.6mg (0.0196 mmol, 89%) 11 as a red solid.
Method B: A solution of 12.0 mg (0.0383 mmol) 8 in 4 cm3 HBr (47% aq) was refluxed for 30min
under Ar to yield 10.7mg (0.0375 mmol, 98%) of 11 after work-up according to method A. Mp
266–268ꢁC; TLC: Rf ¼ 0.65 (CHCl3:AcOEt¼ 1:1), Rf ¼ 0.79 (CHCl3:MeOH¼ 10:1); 1H NMR
(500 MHz): ꢁ ¼ 12.56 (s, 8-OH), 12.25 (s, 1-OH), 7.16 (d, J ¼ 2.2 Hz, ar-H4), 7.04 (d, J ¼ 2.0 Hz, ar-
H5), 7.01 (sbr, 6-NH2), 6.82 (d, J ¼ 2.2 Hz, ar-H2), 6.25 (d, J ¼ 2.0 Hz, ar-H7), 3.91 (s, 3-OCH3) ppm;
13C NMR (125MHz): ꢁ ¼ 186.7 (C9), 181.7 (C10), 164.9 (C3), 164.5 (C8), 163.7 (C1), 156.9 (C6),
134.6 (C4a), 134.5 (C10a), 109.7 (C9a), 108.4 (C5), 106.9 (C4), 106.7 (C2), 105.0 (C8a), 102.3 (C7),
56.09 (3-OCH3) ppm; ESI-MS (MeOH:DMSO¼ 6:1 þ 1% HCOOH, c ꢃ 1 mgꢄ cmꢂ3, positive ion
mode): m=z ¼ 286 ([M þ H]þ); IR (KBr): ꢂꢀ¼ 3486, 3440, 3381, 3362, 3246, 2954, 2924, 2852,
1731, 1631, 1609, 1598, 1572, 1481, 1462, 1442, 1421, 1386, 1326, 1313, 1299, 1271, 1242, 1217,
1190, 1167, 1156, 1113, 1039, 837, 760, 720 cmꢂ1; UV-Vis (CHCl3, c ¼ 1.20ꢄ 10ꢂ5 mol ꢄ dmꢂ3):
ꢃ
max(") ¼ 256 (15140), 303 (21030), 475 (3780) nm (dm3 ꢄ molꢂ1 ꢄ cmꢂ1); UV-Vis (MeOH,
c ¼ 1.20ꢄ 10ꢂ5 mol ꢄ dmꢂ3): ꢃmax(") ¼ 223 (23050), 240 (14470), 257 (15100), 308 (18160), 493
(3050) nm (dm3 ꢄ molꢂ1 ꢄ cmꢂ1).
1,3-Dimethoxy-8-hydroxy-6-carboxy-9,10-anthraquinone (12, C17H12O7)
A stirred solution of 6.4 mg (0.0187 mmol) 6 in 1.5 cm3 CH2Cl2 was flushed with Ar and cooled to
ꢂ7ꢁC. To this solution 7.3 mm3 (0.0757 mmol) BBr3 were added through a septum by means of a
syringe and the mixture was stirred for 1 h (Tend ¼ 1ꢁC). The reaction mixture was poured into ice=H2O
and extracted with AcOEt. The organic layer was dried (Na2SO4), filtered, and evaporated to dryness
to yield 5.5 mg (0.0167mmol, 90%) 12 as a yellow solid. Mp >278ꢁC (decomp.); TLC: Rf ¼ 0.47
(CHCl3:MeOH ¼ 3:2), Rf ¼ 0.10 (CHCl3:MeOH ¼ 10:1); 1H NMR (500MHz): ꢁ ¼ 13.62 (sbr, 6-
COOH), 13.07 (s, 8-OH), 8.08 (s, ar-H5), 7.71 (s, ar-H7), 7.33 (s, ar-H4), 7.06 (s, ar-H2), 4.00 (s,

1638
B. Lackner et al.
3-OCH3), 3.99 (s, 1-OCH3) ppm; 13C NMR (125MHz): ꢁ ¼ 186.2 (C9), 181.5 (C10), 165.5 (6-
COOH), 165.4 (C3), 163.3 (C1), 161.3 (C8), 136.7 (C4a or C10a), 136.6 (C10a or C4a), 132.7
(C6), 124.1 (C7), 119.1 (C8a), 118.0 (C5), 114.0 (C9a), 104.9 (C4), 104.5 (C2), 56.69 (1-OCH3),
56.24 (3-OCH3) ppm; ESI-MS (MeOH:DMSO¼ 6:1 þ 1% NH3, c ꢃ 1 mgꢄ cmꢂ3, negative ion mode):
m=z ¼ 327 ([M ꢂ H]ꢂ); IR (KBr): ꢂꢀ¼ 3131, 2925, 2854, 1730, 1670, 1631, 1593, 1545, 1479, 1464,
1438, 1396, 1370, 1338, 1282, 1248, 1208, 1187, 1170, 1109, 1051, 1030, 986, 881, 771 cmꢂ1
;
UV-Vis (CHCl3, c ¼ 9.29 ꢄ 10ꢂ5 mol ꢄ dmꢂ3): ꢃmax(") ¼ 241 (10050), 278 (9160), 425 (4070) nm
(dm3 ꢄ molꢂ1 ꢄ cmꢂ1); UV-Vis (MeOH, c ¼ 5.57 ꢄ 10ꢂ5 mol ꢄ dmꢂ3): ꢃmax(") ¼ 205 (16100), 228
(17410), 274 (10300), 423 (4200) nm (dm3 ꢄ molꢂ1 ꢄ cmꢂ1).
1,8-Dihydroxy-3-methoxy-6-carboxy-9,10-anthraquinone (13)
Method A: To a refluxing solution of 6.2 mg (0.0181 mmol) 6 in 3 cm3 of glacial acetic acid under Ar,
500 mm3 HBr (47% aq) were added and heated for 90min. The mixture was cooled, poured into
ice=H2O, and neutralized with saturated NaHCO3 solution. After extraction with CHCl3 the organic
layer was evaporated to dryness to yield 5.3 mg (0.0169 mmol, 93%) 13 as a yellow solid.
Method B: A solution of 5.0 mg (0.0146mmol) 6 in 1.75cm3 HBr (47% aq) was refluxed for
60min under Ar to yield 3.7 mg (0.0118mmol, 81%) 13 after work-up according to method A.
Mp >298ꢁC (decomp.) (Ref. [25] 312ꢁC (sealed tube)); TLC: Rf ¼ 0.44 (CHCl3:MeOH¼ 3:2),
1
Rf ¼ 0.08 (CHCl3:MeOH¼ 10:1); H NMR (500 MHz): ꢁ ¼ 13.72 (sbr, 6-COOH), 12.10 (s, 1-OH),
12.02 (s, 8-OH), 8.12 (s, ar-H5), 7.75 (s, ar-H7), 7.23 (s, ar-H4), 6.91 (s, ar-H2), 3.95 (s, 3-OCH3) ppm;
13C NMR (125MHz): ꢁ ¼ 189.6 (C9), 180.7 (C10), 166.4 (C3), 165.4 (6-COOH), 164.6 (C1), 161.1
(C8), 137.7 (C6 or C10a), 134.8 (C4a), 133.6 (C10a or C6), 124.2 (C7), 118.8 (C5), 118.5 (C8a), 110.2
(C9a), 107.9 (C4), 106.7 (C2), 56.45 (3-OCH3) ppm; ESI-MS (MeOH:DMSO¼ 6:1 þ 1% NH3,
c ꢃ 1 mgꢄ cmꢂ3, negative ion mode): m=z ¼ 313 ([M–H]ꢂ); IR (KBr): ꢂꢀ¼ 3067, 2956, 2925, 2854,
1698, 1626, 1611, 1571, 1481, 1434, 1397, 1313, 1276, 1250, 1211, 1165, 1094, 1029, 1016, 980, 918,
906, 789, 772, 667 cmꢂ1; UV-Vis (CHCl3, c ¼ 8.00 ꢄ 10ꢂ5 mol ꢄ dmꢂ3): ꢃmax(") ¼ 250 (7830), 275
(7260), 441 (4600) nm (dm3 ꢄ molꢂ1 ꢄ cmꢂ1); UV-Vis (MeOH, c ¼ 7.73ꢄ 10ꢂ5 mol ꢄ dmꢂ3):
ꢃ
max(") ¼ 206 (10670), 228 (12450), 250 (7170), 272 (7320), 436 (4210) nm (dm3 ꢄ molꢂ1 ꢄ cmꢂ1).
Acknowledgements
The cryogenic 500 MHz probe was purchased from FWF project P15380 (project leader: Prof.
Dr. N. M€uller). We are grateful to Prof. Dr. C. Klampfl (Institute of Analytical Chemistry) and
Dr. C. Schwarzinger (Institute for Chemical Technology of Organic Materials) for recording the
mass spectra.
References
[1] Kuo YC, Sun CM, Ou JC, Tsai WJ (1997) Life Sci 61: 2335
[2] Ubbink-Kok T, Anderson JA, Konings WN (1986) Antimicrob Agents Chem 30: 147
[3] Cohen PA, Hudson JB, Toweres GHN (1996) Experientia 52: 180
[4] Jing X, Ueki N, Cheng J, Imanishi H, Hada T (2002) Jpn J Cancer R 93: 874
[5] Falk H (1999) Angew Chem 111: 3306; Angew Chem Int Ed 38: 3116
[6] For most of the compounds within this work a numbering and nomenclature not strictly conform
1
to the IUPAC rules was chosen to provide a more illustrative comparison of H and 13C NMR
spectroscopical data
[7] Keller G, Steglich W (1987) Phytochemistry 26: 2119
€
[8] Teich L, Daub KS, Krugel V, Nissler L, Gebhardt R, Eger K (2004) Bioorg Med Chem 12: 5961
[9] Muller E, Bayer O (1979) Methoden der Organischen Chemie (Houben-Weyl), 4th ed, vol VII=3c
€
[10] Curtius T (1890) Ber Dtsch Chem Ges 23: 3023

Synthesis of Emodic Amine by Curtius Rearrangement
1639
[11] Hofmann AW (1881) Ber Dtsch Chem Ges 14: 2725
[12] Lossen W (1872) Justus Liebigs Ann Chem 161: 347
[13] Schmidt KF (1923) Angew Chem 36: 511
[14] Falk H, Meyer J, Oberreiter M (1993) Monatsh Chem 124: 339
[15] Lackner B, Popova Y, Etzlstorfer C, Smelcerovic AA, Klampfl CW, Falk H (2005) Monatsh
Chem 136: 777
[16] Falk H, Tran TNH (1996) Monatsh Chem 127: 717
[17] Owton WM, Brunavs M, Miles MV, Dobson DR, Stegghs DJ (1995) J Chem Soc Perkin Trans 1,
Org Bio-Org Chem 7: 931
[18] Di Napoli G (1998) Eur Pat Appl EP 0 822 178 A1
[19] Capson TL, Poulter CD (1984) Tetrahedron Lett 25: 3515
[20] Shioiri T, Ninimiya K, Yamada S (1972) J Am Chem Soc 94: 6203
[21] Steglich W, Reininger W (1970) J Chem Soc Chem Commun 3: 178
€
[22] Waser M, Lackner B, Zuschrader J, Muller N, Falk H (2005) Tetrahedron Lett 46: 2377
[23] Brockmann H; Kluge F, Muxfeldt H (1957) Chem Ber 90: 2302
[24] Hassall CH, Morgan BA (1973) J Chem Soc Perkin Trans 1, 2853
[25] Bloomer JL, Stagliano KW, Gazillo JA (1993) J Org Chem 58: 7906