Preparation and Characterization of a Small Library of Thermally-Labile End-Caps for Variable-Temperature Triggering of Self-Immolative Polymers

Article abstract of DOI:10.1021/acs.joc.8b00135

The reaction between furans and maleimides has increasingly become a method of interest as its reversibility makes it a useful tool for applications ranging from self-healing materials, to self-immolative polymers, to hydrogels for cell culture and for the preparation of bone repair. However, most of these applications have relied on simple monosubstituted furans and simple maleimides and have not extensively evaluated the potential thermal variability inherent in the process that is achievable through simple substrate modification. A small library of cycloadducts suitable for the above applications was prepared, and the temperature dependence of the retro-Diels-Alder processes was determined through in situ ¹H NMR analyses complemented by computational calculations. The practical range of the reported systems ranges from 40 to >110 °C. The cycloreversion reactions are more complex than would be expected based on simple trends expected based on frontier molecular orbital analyses of the materials.

Full text of DOI:10.1021/acs.joc.8b00135

Subscriber access provided by UNIVERSITY OF TOLEDO LIBRARIES
Preparation and Characterization of a Small Library of Thermally-Labile
End-Caps for Variable-Temperature Triggering of Self-Immolative Polymers
S. Maryamdokht Taimoory, S. Iraj Sadraei, Rose Fayoumi, Sarah Nasri, Matthew Revington, and John F. Trant
J. Org. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.joc.8b00135 • Publication Date (Web): 28 Mar 2018
Downloaded from http://pubs.acs.org on March 28, 2018
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 33
The Journal of Organic Chemistry
1
2
3
4
5
Preparation and Characterization of a Small Library of ThermallyꢀLabile Endꢀ
Caps for VariableꢀTemperature Triggering of SelfꢀImmolative Polymers
6
7
8
9
S. Maryamdokht Taimoorya, S. Iraj Sadraeia, Rose Fayoumia, Sarah Nasria, Matthew
Revingtona, John F. Tranta,b*
a Department of Chemistry and Biochemistry, University of Windsor, 401 Sunset Ave.
Windsor, Ontario N9B 3P4
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
b Canadian Centre for Alternatives to Animal Methods, University of Windsor, 401
Sunset Ave. Windsor, Ontario N9B 3P4
j.trant@uwindsor.ca; Fax: 1ꢀ519ꢀ973ꢀ7098
Abstract:
The reaction between furans and maleimides has become increasingly a target of
interest as its reversibility makes it a useful tool for applications ranging from selfꢀ
healing materials, to selfꢀimmolative polymers, to hydrogels for cell culture and for the
preparation of bone repair. However, most of these applications have relied on simple
monosubstituted furans and simple maleimides and have not extensively evaluated the
potential thermal variability inherent in the process that is achievable through simple
substrate modification. A small library of cycloadducts suitable for the above applications
was prepared, and the temperature dependence of the retroꢀDielsꢀAlder processes was
determined through in situ 1H NMR analyses complemented by computational
calculations. The practical range of the reported systems ranges from 40 ºC to >110 ºC.
The cycloreversion reactions are more complex than would be expected based on simple
trends expected based on FMO analyses of the materials.
Keywords: DielsꢀAlder, Selfꢀimmolative polymers, hydrogels, furanꢀmaleimide Dielsꢀ
Alder, kinetic analysis, Density Functional Theory
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 2 of 33
1
2
3
4
1. Introduction
5
6
7
8
The DielsꢀAlder reaction is potentially the most studied, and certainly one of the most
used reactions, in synthetic chemistry over the past 90 years.1 It has remained one of the
most efficient carbonꢀcarbon bond forming reactions for much of that period. The
mechanism of the transformation is well established, as are the pathways that deviate
from the ideal pericyclic transformation. The reaction has drawn renewed interest as an
example of “Click” Chemistry,2 as it is often highly exergonic through the conversion of
3carbonꢀcarbon π bonds into 2 σ and 1 π bond, is completely atom efficient, and proceeds
without the formation of byꢀproducts.
The maleimideꢀfuran DielsꢀAlder reaction is of particular interest as it allows for
reversibility: the aromatic stabilization of furan nearly balances the favourability of the σꢀ
bond forming forward reactions.3 The reaction is exergonic, but is reversible at elevated
temperatures and higher dilution. This has allowed for a number of innovative
applications in materials science and especially in selfꢀhealable systems,4 in
thermoresponsive polymers,5 and for the generation 3ꢀdimensional hydrogel matrices6 for
tissue culture,7 and bone repair.8
Gillies and Trant recently reported the application of the DielsꢀAlder reaction
between furan and maleimide as a thermallyꢀsensitive endꢀcap for selfꢀimmolative
polymers (SIPs).9 SIPs are materials that depolymerize in a headꢀtoꢀtail fashion upon the
cleavage of an endꢀcap. This scission is dependent on the structure of the endꢀcap, and
various stimuli can be applied to initiate polymer degradation. Endꢀcaps sensitive to
reduction, oxidation, pH, light, and enzymes have all been developed;10 but heat is
particularly attractive for biomedical applications.9, 11 Hyperthermal therapy is the use of
localized heat to kill cancer cells, especially by using either external radiofrequency for
tumours near accessible body cavities to stimulate nanoparticles, or direct heating using
intraluminal or interstitial probes for tumours deeper in the body.12 Thermally sensitive
endꢀcaps could provide access to an alternative hyperthermic treatment modality by
allowing for localized release of drugs from selfꢀimmolative nanoparticles.
The system studied in this report relies on the natural instability of 2ꢀ
methylsubsituted furan rings (Figure 1).13 This reactivity can be masked while the furan
ring is incorporated into the 7ꢀoxanorbornene construct with maleimide. Two simple
systems were examined, a benzyl maleimide and a propargyl maleimide that allowed for
the introduction of additional functionality, such as poly(ethylene glycol). Both materials
were thermally similar, and degraded with similar profiles: 40 hours at 60 ºC provides the
plateau for the depolymerization of the polymer;9 as these polymers degrade rapidly at
that temperature, the rate limiting step is endꢀcap removal.14 Ideally, the technology
would not be limited to this temperature range. It would be useful to have more and less
stable systems able to respond to different temperature signals. This would provide a
library of endꢀcaps useable for a variety of different applications.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Page 3 of 33
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 1. A) Observed instability of 2ꢀbromomethylfuran 1, and its analogy to a selfꢀ
immolativeꢀpolymer conjugated methylfuran derivative 3. B) Use of the masked furan 4
as an endꢀcap for a selfꢀimmolative polymer, 5 as reported by Fan, Trant, and Gillies
(Reference 8). Following endꢀcapping, the cycloadduct undergoes a retroꢀDiels Alder
when exposed to heat to liberate the unstable furan derivative and the accompanying
maleimide. This work focuses on understanding the role of new functional groups
introduced on either the furan or maleimide have on the facility of the retroꢀDiels Alder
reaction.
Herein, we report our studies into the establishment of a small catalogue of
cycloadduct systems as useful tools for our programs developing biocompatible and
microꢀlithography selfꢀimmolative systems, and our studies in developing nextꢀ
generation hydrogel matrix crosslinking moieties for 3ꢀD cell culture and organoid
scaffolding. As the normalꢀdemand DielsꢀAlder reaction is highly sensitive to the energy
difference between the frontier molecular orbitals of the diene (HOMO) and dienophile
(LUMO), we tuned the reactivity of the systems by introducing electron withdrawing and
donating substituents to both the 2ꢀhydroxymethylfuran (4ꢀposition) and maleimide (Nꢀ
alkylation) functionalities. Although the forward reactions proceeded as expected, there is
no clear experimental correlation between the FMO gap and the facility of the retroꢀDiels
Alder reactions. Consequently, we have carried out a combined experimental
investigation into the reaction kinetics of the cycloreversion reactions and have
accompanied it with a detailed computational analysis of the complete reaction pathway
to better characterize these complex systems and understand the factors responsible for
the differential reactivity.
2. Results and Discussion
Synthesis of the endꢀcap library, and experimental analysis of the DielsꢀAlder
reaction between αꢀfurfuryl alcohol and Nꢀsubstituted maleimides.
The forward normal demand DielsꢀAlder reaction is well understood. There is a
clear correlation between the energy gap between the HOMO of the diene and LUMO of
the dienophile, and the rate of reaction. Electron rich dienes raise the energy of the
HOMO, and electron poor dienophiles lower the energy of the LUMO (see Supporting
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 4 of 33
1
2
3
4
5
6
7
8
9
Information for the FMOs of the studied systems; Figure S1). However, this picture is
complicated by the presence of potential secondary steric interactions that can either
stabilize or destabilize a given transition state.15 The ratio of the two potential products,
the endo and exo derivatives, is dependent on the precise experimental conditions; there
are few general rules regarding this distribution, and the reactions must be studied on an
individual basis. Our current research efforts require ready access to a diverse library of
cycloadducts. Consequently, we prepared four different 4ꢀsubstituted αꢀfurfuryl alcohols
(the hydroxymethyl group is required for selfꢀimmolative activity via conjugation to an
SIP) and three maleimides (Table 1). The particular choice of substrates was determined
by balancing electronic diversity with simple synthetic access: any viable SIP endcap will
require rapid access to multigram scale amounts of the materials. Each diene was treated
with each dienophile in turn to yieldtwelveꢀdifferent potential cycloadducts.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 1. The results of the DielsꢀAlder cycloadditions between 4ꢀsubstituted αꢀfurfuryl
alcohols and Nꢀsubstituted maleimides. Products are numbered by combining the
numbers of the two coupling partners for clarity and easy consistency and subscripts are
used to define the two possible stereoisomeric cycloadducts: for example, the
cycloadduct between αꢀfurfuryl alcohols 1 and Nꢀ(pꢀmethoxyphenyl)ꢀmaleimide 5 is
cycloadduct 15endo. Yields refer to isolated yields of both isomers. Where the isomers
were not readily separable, the yield refers to the combined isolated yield, and the
individual bracketed percentages refer to the conversion to each isomer based on crude
1H NMR based on residual furan. Temperature refers to the reaction temperature at which
the cycloaddition was conducted and is a surrogate for reaction facility.
40 ºC
60 ºC
35 ºC
15endo 65%
15exo 15%
Inseparable (52% total) 17endo 70%
16endo (70%)
16exo (23%)
80 ºC
17exo 23%
75 ºC
70 ºC
25endo <5%
25exo 63%
26endo <5%
26exo 31%
27endo 69%
27exo 21%
75 ºC
35endo 61%
35exo <5%
50 ºC
36endo 26%
36exo <5%
70 ºC
Inseparable (31% total)1
37endo (70%)
37exo (23%)
ACS Paragon Plus Environment

Page 5 of 33
The Journal of Organic Chemistry
1
2
3
4
5
6
60 ºC
45endo 25%
45exo <5%
80 ºC
46endo 30%
46exo <5%
65 ºC; 60%
7
8
9
1 Yields are moderate compared to conversion for the 3X series due to debromination side products.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Most of the cycloadditions proceeded smoothly at temperatures between ambient
and 60 ºC, but for more sluggish reactions, higher temperature reactions were conducted
in a sealedꢀtube. In cases in which Rf difference between the nonꢀpolar endo and the more
polar exo stereoisomers was > 0.05, we were able to isolate each isomer using flash
chromatography. In some cases, it proved not possible to easily separate the isomers
under all tested conditions. Although preparative TLC was able to separate every
mixture, this is not a readily scalable alternative: as the purpose of the study is to generate
a useable materials science tool, these are not promising systems. Most reactions were
carried out under kinetic control and the reactions favor the less polar endo cycloadduct.
However, several reactions did not progress unless heated significantly, and in these
cases the thermodynamically more stable exo cycloadduct predominated. This was
especially true, as expected, for the challenging electronꢀpoor nitroꢀsubstituted dienes.
The syntheses of all adducts proved successful except for the DielsꢀAlder reaction
between the methoxyꢀsubstituted hydroxymethylfuran 4 and the benzylꢀsubstituted
maleimide 7. At temperatures below 60 ºC, no significant reaction was observed;
however, at 70 ºC, the simple cycloadduct was never isolated. Instead the aromatized
derivative 47 in Table 1 was isolated in 60 % yield from the residual starting materials.
This likely arose from an initial elimination of the bridging ether moiety to provide a
diolefin that readily underwent a second elimination of water to generate the aromatic 47.
This type of product has been previously observed occasionally from DielsꢀAlder
reactions of furans with maleimides, but has normally required the addition of acid or
base catalysts or reagents,16 the lack of additives in this reaction suggests that a certain
fine electronic balance is reached.
The stereochemistry of the products was assigned based on analogy with our
previously published system,7b, 9 using the coupling constants between the bridgehead
proton on the unsubstituted system. However, in the disubstituted cases, there is no
bridgehead proton making a case based on a Karplus relationship impossible. Instead, we
had noted that for these systems, the endo olefinic peaks are always observed upfield of
the exo olefinic peaks. Assignments based on this observation were consistent with the
expected reaction outcome (majority endo in kineticallyꢀcontrolled experiments, majority
exo in our thermodynamically controlled ones), and we are confident in our assignment.
However, to better support this conclusion, we obtained crystal structures of both 27endo
and 27exo. These crystal structures also provided an opportunity to support the accuracy
of our independently generated computational models. The structures were optimized
computationally from simple drawn structures, they were not derived from the crystals.
Figure 2 highlights and compares several structural features of both the endo and exo
adducts as determined by single crystal Xꢀray and our DFT calculations. The calculations
and crystal structure results are nearly identical (<1% discrepancy) in all cases except for
the angle β formed between the hydroxymethyl and the cyclohexene ring in 27endo
.
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 6 of 33
1
2
3
4
5
6
7
8
9
However, angle β, in both the calculated and experimental systems for 27exo are
consistent with the calculated value for 27endo, the discrepancy may result from a crystalꢀ
packing feature rather than due to an error in the calculation of the preferred
conformation. Our calculated minimum energy structures appear to be excellent models
for this system.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Endo adduct
Exo adduct
X-ray
Calculated
X-ray
Calculated
a/Å
b/Å
c/Å
d/Å
α/°
β/°
γ/°
1.51
1.53
1.51
1.54
1.51
1.54
1.52
1.55
2.32
2.33
2.30
2.32
2.12
2.13
2.11
2.13
104.8
111.8
105.6
99.9
104.9
102.2
105.3
99.8
104.2
101.2
104.2
99.9
104.2
101.7
104.9
99.6
δ/°
Figure 2. A comparison of selected structural parameters extracted from both the Xꢀray
structures and computationallyꢀdetermined geometries of 27endo and 27exo. Ellipsoid
structures provided as Figures S2 and S3.
Although the forward reaction is not central to our study, we wanted to ensure
that our experimental and computational approaches are well validated; furthermore,
many of these particular reactions are not reported in the literature. Consequently, we reꢀ
ran three of the DielsꢀAlder reactions in NMR tubes at 40ºC to monitor and
experimentally determine the reaction kinetics of our system. Our kinetic studies
1
involved acquiring a H spectrum (n=8) every five minutes for up to four hours at a
constant temperature. The conversion to each cycloadduct at any given time point could
be determined from the ratios of the integrals (the solvent peak was used as the internal
standard; Figure 3; see Supporting Information for the H NMR spectra of the other
1
systems showing the conversion to the cycloadduct as a function of time Figure S4).
ACS Paragon Plus Environment

Page 7 of 33
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3. Sample kinetic experiment with selected spectra showing the cycloaddition
between αꢀfurfuryl alcohol 1 and Nꢀ(pꢀmethoxyphenyl)ꢀmaleimide 5 at 40 . Selected
timepoints shown only. The ratio of starting material (ꢀ) to the exo (ꢀ) and endo (ꢀ)
products was determined using the relative integrations of the identified signals. Multiple
peaks were used to obtain a more accurate measurement of the true value of the ratio, and
were found to always be in close agreement.
The conversion profile of the cycloaddition between αꢀfurfuryl alcohol with Nꢀ
benzylꢀand Nꢀ(methoxyphenyl)ꢀmaleimide is provided as Figure 4. The reaction is not
carried out at a high concentration and we do not expect complete conversion. The
reaction favours the endo product in both cases, and after 4 hours the rate of change
slows markedly as the systems approach equilibrium. The kinetic parameters can be
readily determined from the initial data points. The DielsꢀAlder reaction is well
established to be a bimolecular process, and to confirm the quality of the data, we
generated a plot of the conversion data as a function of time. In all cases, the reaction
profile better matched a secondꢀorder process as expected, with the slight observed
deviation from linearity possibly arising from competition from the retroꢀDielsꢀAlder
process (Figure S5). The calculated rate constants at 40 ºC are consistently within the
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 8 of 33
1
2
3
4
range of those reported for other furanꢀmaleimide cycloadditions.10,23–26
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 4. Conversion to the exoꢀ and endoꢀcycloadducts from the reaction between (a) αꢀ
furfuryl alcohol 1 and benzylꢀmaleimide 7; and (b) αꢀfurfuryl alcohol 1 and 4ꢀ
methoxyphenylꢀmaleimide 5 at 40 °C.
Computational Analysis
To understand the critical retroꢀDielsꢀAlder system, we require a comprehensive
analysis of the complete reaction profile: the products of the retroꢀcycloaddition are the
starting furan and maleimide. The forward reaction was also expected to provide insight
into the relative challenge of our proposed systems, and perhaps help explain the
temperatureꢀdependent decrease in endoꢀexo selectivity for the various reactions.
The rate constant of the DielsꢀAlder (DA) reaction between αꢀfurfuryl alcohol 1 and
the OMeꢀsubstituted maleimide, 5 was 6.8x10ꢀ3 L molꢀ1 sꢀ1; while that of the DA reaction
between the same diene and the less electrophilic Bnꢀmaleimide, 7, had a lower value of
1.7*10ꢀ3 L molꢀ1 sꢀ1. These were among the more facile reactions: we found that the
DielsꢀAlder reactions were sluggish for substituted furans, the increased steric bulk
appears to interfere with the ease of accessing the transition state more than any benefit
that could be found from improved electronic effects on the FMO gap. We expected the
ACS Paragon Plus Environment

Page 9 of 33
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
electron rich 4 to provide the most facile reactions: but at ambient temperature the
reaction did not progress, and heating was required for all three cycloadditions. As
mentioned earlier, in the case of cycloadduct 47, we were never able to isolate the desired
cycloadduct: cycloaddition was immediately followed by aromatization. This aromatic
product has been previously observed in similar cycloadditions,16a but we did not observe
it as a product in any of the other reactions even in trace amounts (>0.5%). Thermal endꢀ
cap design is not as simple as expected.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Consequently, we used a quantum mechanical analysis to investigate the various
possible reaction pathways for the twentyꢀfour cycloadditions (endo and exo directedꢀ
processes for each of the twelve combinations) described above. All feature almost the
same reaction profile: only 1 transition state, using a one step synchronous/asynchronous
mechanism is involved (all twoꢀstep mechanisms were found to be significantly higher in
energy). Likewise, there is no computational evidence for the formation of the required
Michaelꢀaddition intermediate that is sometimes proposed for DielsꢀAlder processes.17
The computed reaction energies and activation barriers for both endo and exo adducts are
tabulated in the supporting information (Table S1), but two representative reaction
profiles are provided as Figure 5 (27 and 16, see supporting information Figures S5ꢀS14
for all other cases).
G‡ = 0.2 kcal/mol
11.9
12.1
OH
O
O
OH
R2
O
O
O
N
R1
0.0HO
O
N
NO2
+
R1
N
R2
O
O
O
-18.9
-20.7
Grxn = 1.8 kcal/mol
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 10 of 33
1
2
3
4
5
6
G‡ = 3.1 kcal/mol
12.0
8.9
7
8
9
NO2
0.0
OH
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
OH
HO
O
O
O
O
+
O
N
O
O
N
R2
O
R1
R2
N R1
O
-18.7
-19.7
Grxn = 1.0 kcal/mol
Figure 5. M06ꢀ2X/6ꢀ311+G(d,p) calculated potential energy surface of the DielsꢀAlder
reaction between A) 2 and 7; B) 1 and 6.
In all cases, our calculations showed that the endoꢀTSs are energetically preferred
to the exoꢀTSs (M06ꢀ2X/6311+G(d,p) calculations Table S1, Figure S6 for the potential
energy surfaces of the other reactions). Secondary orbital overlap is considered the prime
factor responsible for lowering the energy of the endo transition states in all cases despite
their being more sterically crowded. The maleimides are particularly good dienophiles in
this respect and we were expecting excellent endo selectivity. As expected the exoꢀ
products prove to be the thermodynamically favoured adduct in all substituted cases
except for the methoxyfuranꢀincorporating cycloadducts 45 and 46; Table S2. Discussion
of the structural origin of this unexpected observation is beyond the scope of this current
disclosure, and will be discussed in context in a computationalꢀfocused followꢀup to this
report.
This level of theory, as supported by the obtained Xꢀray structures, provides an
excellent estimation of the preferred geometry, and the precise calculation of the energies
of the products and transition states at this level of theory has been found to be sufficient
for many similar systems.18
The differing stabilities of the transition states are highly dependent on the precise
nature of the interactions of the substituents. Generally, increasing electron density on the
furan ring and decreasing it on the maleimide decreases the FMO gap between the two
partners as expected (Figure S1; see supporting information for calculated values of FMO
gap). As we wanted to limit the impact of steric effects on reactivity, we chose to
functionalize the nitrogen of the maleimide rather than the olefin system directly. This
same substitution pattern is also what we expect to use for introducing these cycloadducts
as ligating endꢀcaps to conjugate an SIP to a copolymer as has been done previously.9
However, this does mean that the substitution has only a small effect on the electronics of
the maleimide system. Consequently, in most cases the substitution and steric demands of
the furyl group determine the bulk of the observed variance. These effects are
ACS Paragon Plus Environment

Page 11 of 33
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
considerable, and although a complete analysis of the minutiae of these interactions is
unnecessary for the current discussion, we observed that the substituent can be engaged
in nonꢀcovalent interactions to organize the system, and can impart a significant degree of
asymmetry to the transition state, sufficient to change the mechanism from a synchronous
oneꢀstep DielsꢀAlder to an asynchronous reaction. This effect is most pronounced for the
exo cycloadducts of the unsubstituted (15ꢀ17) and methoxyꢀfuran systems (45 and 46).
The monoꢀsubstituted furan, αꢀfurfuryl alcohol 1, would have the fewest negative
steric interactions, and we thus expected to see the best endo selectivity for these systems
as the secondary orbital overlap stereoelectronic effect should dominate. However, we
observed significant amounts of the exo cycloadduct even though we ran the experiments
under pseudoꢀkinetic control at a low temperature. The computational calculations help
explain this result.
The endo pathways (TS related to compounds 15endo, 16endo, 17endo) pass through
a mildly asymmetric transition state with only a small difference in the length of the two
new σ bonds (ꢁꢁd ≈ 0.1Å, Figure S7). The exo transition states (TS related to
compounds 15exo, 16exo, 17exo) however is lower in energy than expected, and a major
reason is that these are highly asynchronous geometries (ꢁꢁd > 0.5Å). This means that
there is significant charge transfer in the transition state, and this is associated with lower
energy barriers. Consequently, the high levels of exo product is due to a difference in
mechanism between the two stereoisomeric possibilities. To the best of our knowledge,
this discrepancy has not been noted for furans, although the high levels of exo product
from furanꢀmaleimide cycloadditions is well established. However this mechanistic
difference: an asynchronous endo pathway and a highly asymmetric exo mechanism has
been reported for reactions between 2,3ꢀdibromoꢀ1,3ꢀbutadiene and maleic anhydride.19
Curiously, all the transition states involving the methoxyꢀsubstituted furfuryl
alcohols (45endo, 45endo, 46exo, 46exo) also exhibit this asymmetric character, partially
explaining the low energy barriers calculated for these processes. Because of the high
degree of asymmetry, we also actively identified the lowest energy pathways that would
lead through a true Michaelꢀadditionꢀcyclization discrete twoꢀstep pathway: the logical
extreme of a highly asynchronous mechanism. However, in all cases, these transition
states were found to be far higher in energy for the initial addition than the pericyclic
mechanisms, and were consequently dismissed as unlikely in this case.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 12 of 33
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 6. Illustration of the degree of asymmetry in the asynchronous transition states
leading to 46 compared to the synchronicity of those leading to 36.
A prime example of this asynchronous process is provided in Figure 6 for the
transition states leading to 46. The formation of 46endo, proceeds through a clear twoꢀ
stage oneꢀstep process: the 2 σ bonds are generated in one step but in a defined sequence.
The bond to the 2ꢀposition (hydroxymethyl group) is largely formed already at the
transition state, while the bond to the 5ꢀposition is almost an Ångstrom longer. The
asymmetry is very similar in the exo transition state as well. This is contrast to the
bromoꢀsubstituted derivative 36. In both the endo and exo transition states, the bondꢀ
formation is nearly simultaneous. These are also both far higher in energy than the ones
leading to 46, and is consistent with the experimental results. The asynchonicity can be
analyzed using an intrinsic reaction coꢀordinate path (Figure 7). The presence of an
inflection point in an asynchronous pathway (where the first σ bond is nearly fully
formed at 1.67 Å, while the second σ bond is just beginning to form, with a length of 2.48
Å) indicates that the process passes through this more challenging asymmetric species
before the geometry relaxes to the product. (see Supporting Information Figure S7 for
comparison of these values for compounds 15ꢀ17 which feature a mildly asynchronous
endo transitions states and very highly asynchronous exo transition states).
All the reactions encompass some subtleties like this: differential densities of the
FMOs on the furyl substituents, higher dipole moment and in turn lower activation
barriers, and the presence of a complex balance of stabilizing and destabilizing nonꢀ
covalent interactions between diene and dienophile. Although fascinating, for
conciseness, a full discussion of the precise factors leading to the stabilities of the
transition states and the products will be instead disseminated in a forthcoming report.
ACS Paragon Plus Environment

Page 13 of 33
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 7. Intrinsic reaction coꢀordinate of the cycloaddition between methoxyꢀsubstituted
furan and methoxyꢀsubstituted maleimide endoꢀTS45. The mechanism is highly
asynchronous and approaches a oneꢀstep asynchronous mechanism; the presented
structure is at the inflection point of the curve as the second σꢀbond is being formed (halfꢀ
pathway point).
Spectroscopic study of the retroꢀDiels–Alder reaction between furfuryl alcohol and
maleimide derivatives
With a full theoretical description of the reactions in hand, we investigated the
practicalities of using these potential endꢀcaps as thermally responsive materials. Ideally,
we would have a library of compounds with varying thermal stability that are easy to
prepare, isolate and handle. These furanꢀmaleimide constructs are notable due to their
tendency for cycloreversion to the diene and dienophile. This is driven by the aromatic
stability of the furan, which is sufficient to significantly stabilize the diene relative to
nonꢀaromatic dienes favouring the reverse reaction, but not so much that the aromatic
stabilizing energy cannot be overcome to allow for the forward reaction.
Computationally we can predict the stability of the different systems (Table S2) using
the M06ꢀ2X/6ꢀ311+G(d,p) level of theory. These calculations again strongly impress the
greater importance of the furan than maleimide substitution on determining the reaction.
The calculations suggest that the methoxyꢀsubstituted systems should be by far the most
labile and susceptible to cycloreversion, followed by the unsubstituted furans, with both
the nitro and bromoꢀsubstituted systems being considerably more thermally stable (with
the bromo systems being the most stable of all). In all but one case (46) the endo
cycloadduct should be more likely to cyclorevert than the exo, and, again in general, the
nitrophenylꢀmaleimide is predicted to be the least stable of the three, followed by the
methoxyꢀsubstituted phenylmaleimide, with the benzyl maleimide being the most stable
of all the systems. Again, the methoxy furan series (45ꢀ46)endo and (45ꢀ46)exo breaks
these patterns due to the unusual stability of 46endo. However, to confirm the accuracy of
1
these models we used H NMR studies to monitor the cycloreversion reactions and
determine the kinetic behavior of the materials.
Our first experiment was to determine the reproducibility of our experiment, and the
depolymerization of 15endo at 70 ºC was studied in triplicate. The experimental setꢀup is
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 14 of 33
1
2
3
4
5
6
7
8
9
simple: 3ꢀ10 mg of cycloadduct are dissolved in 750 ꢂL of either deuterated acetonitrile
or DMSO and transferred to an NMR tube. The NMR probe is brought up to the reaction
temperature in the presence of an NMR tube containing the same solvent. At time 0, the
sample is inserted into the instrument. Due to the volume of solvent in an NMR tube, the
time to heat to the temperature of the probe was considered to be negligible. An
automated sequence then collects spectra consisting of 4 scans at regular intervals (every
1.0 minute for the first ten minutes, then every 5 minutes for the next 230 minutes) over a
total of 4 hours. A script then uses the solvent peak as an internal control, and integrates
the same region in every spectrum (the exact peaks differ depending on the
diene/dienophile pair and are selected based on the isolation of the resonances) to
produce a data table providing the conversion of the reaction at each time point. For
15endo, three independent trials from two different synthetic batches were tabulated, the
values averaged over the three trials and plotted as Figure S8. The variance is minimal;
consequently, all other experiments were only done as single trials (See Figure 8 for a
typical profile, compound 15endo, and Figure S9 for additional examples). All
cycloadducts were studied at 70 ºC (Figure 9). Those that proceeded rapidly at this
temperature were also investigated at lower temperatures, while those that were more
sluggish were examined at higher temperatures; several examples incorporating furan 1
demonstrate this point (Figure 10). The initial rates of each reaction (defined as the slope
of the conversion curve from 0ꢀ10 minutes), and the time to 25% conversion are provided
in Table 2 as two different surrogates for thermal stability. Many of the reactions did not
proceed to completion as the reaction reaches equilibrium in the NMR tube. Furthermore,
although isolated, many of the exo cycloadducts were not studied in this manner. It is
well established that the exo adducts are more stable than their endo counterparts, and
there was no lack of temperatureꢀresistant examples in the library.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 2. Initial rate (mol•Lꢀ1•sꢀ1x10ꢀ8) and time to 25 % conversion for the various
cycloadducts at the different temperatures. If there is no conversion, the values are stated
to be <1x10ꢀ8 mol•Lꢀ1•sꢀ1 and >105 s for the two parameters respectively.
40 ºC
31
60 ºC
274
2860 s
70 ºC
744
593 s
1,350
231 s
598
85 ºC
100 ºC
110 ºC
15
35,700 s†
57
16
12,600 s
17
889 s
<1
77
92
11,400 s
112
4,270 s
355
2,400 s
169
3,500 s
25
>105 s
<1
48,800 s†
26
>105 s
<1
27endo
27exo
35
>105 s
<1
36
9,920 s
591
>105 s
<1
281
1580 s
>105 s
16
818 s
36
ACS Paragon Plus Environment

Page 15 of 33
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
24,200 s†
6
381
1,390 s
37
45
46
60,500 s†
209
8,990 s
554
632 s
869
425 s
3170
117 s
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
†The value is extrapolated from the data, as conversion did not cross 25% over the
reaction period.
1
Figure 8. Example retroꢀDiels Alder reaction profile, as monitored by H NMR, of
cycloadduct 15endo
.
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 16 of 33
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 9. The 1H NMRꢀmeasured conversion of the cycloꢀreversion of all compounds
at 70 ºC. Compounds 25, 26, 27 all showed no integrable formation of product over the
reaction period and are superimposed on the baseline.
The results obtained from these kinetic analyses are intriguing and it is hard to
discern any obvious pattern. The unsubstituted furan (15ꢀ17) shows the most facile
cycloreversion behavior (<70 ºC), while the substituted systems, are all more thermally
resistant. The nitroꢀsubstituted electron poor systems (25ꢀ27) are the most thermally
resistant (>100ºC), while the moderately electron poor 2ꢀbromo derivatives (35ꢀ37;
~100ºC) are intermediate in reactivity. The methoxyꢀsubstituted systems (45 and 46) are
almost as reactive as the unsubstituted systems (~70ꢀ85 ºC). In terms of the maleimides,
the more electron poor the maleimide, in general, the easier the cycloreversion, although
the effect due to varying the substituent is far smaller than that observed for the furan.
This is presumably due to the distant functionalization.
ACS Paragon Plus Environment

Page 17 of 33
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 10. Representative conversion curves for the retroꢀDielsꢀAlder reactions
of the cycloadducts to illustrate the range of potential triggering temperatures. All three
examples incorporate the same maleimide building block and vary only in the identity of
the furan moiety. Temperature labels indicate the constant temperature at which the H
NMR experiments were run to obtain the data.
1
Most of the reactions do not proceed to completion under the concentrations
examined (5.3 mg/mL). This is expected for an equilibrium process. This is not a
functional concern for these endꢀcaps when they will be employed with SIPs as 5.3
mg/mL is far above any conceivable operational concentration envisioned for any
biomedical application.
All materials were stored at ꢀ20ºC for more than six months, and no observable
decomposition was observed. These are stable endꢀcaps that can be prepared in bulk in
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 18 of 33
1
2
3
4
5
6
7
8
9
advance for functionalization of SIPs as required. Furthermore, we propose that only a
few examples from this library are sufficient to provide the required thermal flexibility
for most applications. Construct 16 is suitable for nearꢀphysiological temperature
applications and will cleave when exposed to physiological environments. Cycloadduct
46 is cleavable at elevated temperatures around 70 ºC, and may prove appropriate for
localized hyperthermic treatments; construct 35 is appropriate for polymers requiring
depolymerization at around 100 ºC; and 27exo provides a thermally stable derivative
appropriate for depolymerization above 110 ºC.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Conclusion
To support our goal of developing new practical classes of endꢀcaps for selfꢀ
immolative materials, we have further examined both the forward and reverse Dielsꢀ
Alder reaction between hydroxymethylfurans and maleimides. For utility, the forward
process must provide materials that are readily separable and easily provide substrates in
good yield. For most of these cycloadducts, the endo and exo systems were readily
separable and in all cases the materials were stable when stored at ꢀ20ºC or ꢀ80 ºC for a
minimum of six months, with no evidence of cylcoreversion. The cycloreversion
chemistry did not follow a simple trend as steric effects appear to play an equally
important role in determining reactivity as the wellꢀestablished electronic effects on
pericyclic kinetics. Computational studies support the experimental results and provide
significant additional insight into the factors responsible for the deviations from pure
electronic control of reaction kinetics and thermodynamics. This simple library of
compounds successfully provides materials that are labile at 40 ºC, and others that are
still reasonably stable even at 110 ºC: a useful operational range for biomedical or
industrial applications. We propose that four different members of this library are
sufficient for most applications over this range: 16 for nearꢀphysiological temperatures,
46 for elevated temperatures around 70 ºC, 35 for temperatures around 100 ºC, and 27exo
for higher temperatures. Our applications of these systems, detailed computationallyꢀ
derived details of the mechanisms of reaction, and a broadening of the members of this
library will follow in due course.
Experimental Section
General Experimental Protocols
Solvents were purchased from Caledon Labs (Caledon, Ontario), SigmaꢀAldrich
(Oakville, Ontario) or VWR Canada (Mississauga, Ontario). Other chemicals were
purchased from SigmaꢀAldrich, AK Scientific, Oakwood Chemicals, Alfa Aesar or Acros
Chemicals and were used without further purification unless otherwise noted. Anhydrous
toluene, tetrahydrofuran (THF), diethyl ether and N,Nꢀdimethylformamide (DMF) were
obtained from an Innovative Technology (Newburyport, USA) solvent purification
system based on aluminium oxide columns. CH2Cl2, pyridine, acetonitrile, N,Nꢀ
diisopropylethylamine (DIPEA) and NEt3 were freshly distilled from CaH2 prior to use.
Purified water was obtained from a Millipore deionization system. All heated reactions
were conducted using oil baths on IKA RET Basic stir plates equipped with a P1000
temperature probe. Thin layer chromatography was performed using EMD aluminumꢀ
ACS Paragon Plus Environment

Page 19 of 33
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
backed silica 60 F254ꢀcoated plates and were visualized using either UVꢀlight (254 nm),
KMnO4, vanillin, Hanessian’s stain, or Dragendorff’s stain. Preparative TLC was done
using glassꢀbacked silica plates (Silicycle) of either 250, 500, 1000 or 2000 ꢂm thickness
depending on application. Column chromatography was carried out using standard flash
technique with silica (SiliaflashꢀP60, 230ꢀ400 mesh Silicycle) under compressed air
pressure. Standard workꢀup procedure for all reactions undergoing an aqueous wash
involved back extraction of every aqueous phase, a drying of the combined organic
phases with anhydrous magnesium sulphate, filtration either using vacuum and a
sinteredꢀglass frit or through a glassꢀwool plug using gravity, and concentration under
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
reduced pressure on a rotary evaporator (Buchi or Synthware). H NMR spectra were
obtained at 300 MHz or 500 MHz, and 13C NMR spectra were obtained at 75 or 125 MHz
on Bruker instruments. NMR chemical shifts (δ) are reported in ppm and are calibrated
against residual solvent signals of CHCl3 (δ 7.26), DMSOꢀd5 (δ 2.50), acetoneꢀd5 (δ
2.05), or methanolꢀd3 (δ 3.31). HRMS were conducted on a Waters XEVO G2ꢀXS TOF
instrument with an ASAP probe in CI mode.
Computational Methodology
All calculations were performed using Gaussian 09, Revision D.01.2120 on Sharcnet.
Geometry optimizations and frequency calculations (to confirm that each stationary point
is either a firstꢀorder saddle point or a minimum) were carried out using the B3LYP
density functional with the 6ꢀ311++G(d,p) basis set.21 As the M06ꢀ2X functional is good
to reproduce the free energies of cycloaddition reactions better than other functionals,
singleꢀpoint calculations were performed using the M06ꢀ2X/6ꢀ311+G(d,p) level of theory
and basis set.22 The selfꢀconsistent reaction field (SCRF) using IEFPCM solvation
model23 were used to consider the solvation effects of acetonitrile at 308K for both
geometry optimizations and singleꢀpoint calculations. The HOMO and LUMO molecular
orbitals and their associated energies were calculated at the B3LYP/6ꢀ311++G(d,p) level
of theory.
Details regarding the kinetic studies
Kinetic studies were carried out in an NMR tube. All samples were prepared in the
same way. Forward DielsꢀAlder reactions: 5.65 mg of furan (5.0 ꢂL, 0.057 mmol) and an
equimolar amount of the maleimide was added directly to 750 ꢂL of deuterated solvent
(either acetonitrileꢀd3 or more often DMSOꢀd6 for any reaction 70 ºC or above). For
reverse DielsꢀAlder reactions, 16.0 mg of cycloadduct was added to 3.0 mL of deuterated
solvent in a vial. The solution was then partitioned between 4 NMR tubes and stored at ꢀ
20 ºC until used for an experiment. The experiments were carried out at the indicated
temperatures on a Bruker 300 1H NMR spectrometer equipped with a variable
temperature probe. A blank NMR tube containing the solvent but no analyte was inserted
into the probe and the spectrometer was allowed to equilibrate to the indicated
temperature for 10 minutes. The tube was then switched for the sample, and the
experiment was run with a spectrum being collected every minute for the first 11
minutes, and then every 5 minutes thereafter for a total of four hours using the
multi_zgvd script in the Bruker Topspin suite (8 scans per spectrum). Following data
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 20 of 33
1
2
3
4
5
6
7
8
9
collection, the first and last spectra of the series were examined. If there was no change
(and no appearance of starting material or product as determined by comparison of the
spectra with those of previously isolated samples), no further action was taken, and
conversion was determined to effectively be 0 %. If any integration was noted in the final
spectrum at the frequency corresponding to the peaks of the relevant reaction products,
then all spectra in the series were integrated over identical frequency ranges. The data
was exported to .txt files, imported into excel and converted to reaction conversions.
Initial rates were calculated based on the first four data points. In all observed cases, this
region was effectively linear. The time required to obtain 25% conversion was
determined by a simple linear interpolation between the relevant data points. This value is
arbitrary, but a halfꢀlife was considered to be a non accurate measurement as many of the
cycloreversions plateau before 50% conversion due to the equilibrium kinetics of the
system. The 25% conversion point was generally observed in the early stages of the
cycloreversion and was selected as a meaningful surrogate for our own use of these
systems, and will potentially play the same role for other researchers.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Specific Experimental Protocols
The experimental schemes associated with the known furans and maleimides (1ꢀ7) are
available in the supporting information. Intermediates in the synthesis not referenced in
the main text are numbered as SX, and are illustrated in Schemes S1ꢀS4.
General procedure for the synthesis of maleimic acids 5a, 6a, 7a:
All compounds were synthesized using literature methodologies24 with yields ranging
from 90% to 91%. Briefly, maleic anhydride (S1) (500 mg, 5.2 mmol) and equimolar
amounts of the required amine were combined in CHCl3 (6 mL) and stirred for 45
minutes as a precipitate formed. This precipitate was then filtered and washed with cold
(4 ºC) water. The analytical data for these maleimic acids have been previously
published, and the data was consistent with the published spectra.24ꢀ25
General procedure for synthesis of maleimides 5, 6 and 7.
The maleamic acids (5a, 6a or 7a, 4.7 mmol) were dissolved in 5 mL of acetic anhydride
along with sodium acetate (100 mg, 1.2 mmol). The mixture was heated for 2ꢀ4 h at 100
ºC (exact reaction time depended on the substituent), until the reaction was determined to
be complete by TLC. The solution was then cooled, diluted with water, then extracted
repeatedly with ethyl acetate. The combined organics were dried with magnesium sulfate,
and filtered and concentrated in the usual fashion. The solid residue was then redissolved
in a minimum amount of THF, and precipitated through the dropwise addition of iceꢀcold
ether. The solid was recovered, resuspended in additional minimal THF, and precipitated
by addition into cold water. A final filtration provided the desired maleimides in 73% to
80% yield. Spectroscopic data is consistent with previous reports.25
Nꢀ(pꢀmethoxyphenyl)ꢀMaleimide 5
ACS Paragon Plus Environment

Page 21 of 33
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
Prepared as per the general procedure above using 12.0 g of maleic anhydride (24ꢀfold
scale of the general protocol, 125 mmol). Synthesis of 5a proceeded for 45 minutes,
providing 26.4g of the maleimic acid in 90% crude yield as a yellow powder. The ring
closing (using 130 mL of acetic anhydride and 5.8 g of sodium acetate) provided a dark
yellow amorphous solid, that after recrystallization as in the general protocol, was
recovered as 17.8 g of bright yellow needles in 73% yield; 66% yield overall from the
maleic anhydride.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Yellow needles. 1H NMR (300 MHz, CDCl3) δppm: 7.27ꢀ7.23 (m, 2H), 7.03ꢀ6.98 (m, 2H),
6.85 (s, 2H), 3.80 (s, 3H). Spectral data are consistent with previously published
spectra.26
Nꢀ(pꢀnitrophenyl)ꢀMaleimide 6
Prepared as per the general procedure above using 10.0 g of maleic anhydride (20ꢀfold
scale of the general protocol). Synthesis of 6a proceeded for 2 hours, providing 23.65g of
the maleimic acid in 91% crude yield as brown crystals. The ring closing (using 115 mL
of acetic anhydride and 5.1 g of sodium acetate) provided a dark yellow amorphous solid,
that after recrystallization as in the general protocol, was recovered as 17.1g of a pale
yellow powder in 78% yield; 71% yield overall from the maleic anhydride.
1
Yellow powder, H NMR (300 MHz, CDCl3) δppm: 8.35ꢀ8.32 (m, 2H), 7.70ꢀ7.60 (m,
2H,), 6.94 (s, 2H). Spectral data are consistent with previously published spectra.26
NꢀbenzylꢀMaleimide 7
Prepared as per the general procedure above using 1 g of maleic anhydride (2ꢀfold scale
of the general protocol). Synthesis of 7a proceeded for 45 minutes, providing 2.1 g of the
maleimic acid in 93% crude yield; the ring closing to 7 required only 1 hour. The crude
mixture was first purified by flash chromatography (7:3 hexanesꢀethyl acetate), and the
fractions containing the product were combined, concentrated, and then recrystallized
from 2ꢀpropanol and water to provide 1.45 g, an 80% yield of the title compound, as
white crystals in 75% overall yield after vacuum drying.
1
White crystals. H NMR (300 MHz, CDCl3) δppm: 7.33ꢀ7.22 (5H, m), 6.69 (2H, s), 4.66
(2H, s). Spectral data are consistent with previously published spectra.24
5ꢀNitrofurfural diacetate, S3 Prepared according to a modified version of the protocol
of Hou.27 A mixture of 8.6 mL concentrated HNO3 and 0.06 mL concentrated H2SO4 was
slowly added into 90 mL of acetic anhydride while stirring at a temperature of 0 °C. This
was followed by the slow addition of 10.4 mL of furfural, S2, into the acid mixture with
stirring and temperature remaining at 0 °C. The mixture was left to stir at this same
temperature for 1 hour. At this time, 80 mL of water was added and the mixture was left
to stir at room temperature for an additional 30 minutes, over which time a white
precipitate formed. A 10% NaOH solution (10 g of NaOH in 100 mL of water) was then
added to the mixture until the pH rose to 2.5. The mixture was then heated in a water bath
at 55 °C for 1 hour. After cooling, the precipitate was filtered and washed with water
prior to being recrystallized from anhydrous ethanol and dried to provide 5.2 g of white
crystals in a 75% yield. The material was then used without further purification. Rf= 0.37
(6:4, hexanesꢀethyl acetate).
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 22 of 33
1
2
3
4
5
6
7
8
9
5ꢀNitrofurfural, S4 Prepared according to a modified version of the protocol of Hou.27
Previously prepared 5ꢀnitrofufural diacetate (S3, 5.2 g, 21.4 mmol) was added to 52 mL
of 50% H2SO4 and the resulting mixture was heated using a heat gun (Wagner model#
283022 HT 775, 540 ºC) for 2 minutes. After cooling, the hydrolysate was extracted via
ethyl acetate and the organic layer was washed with water, dried with magnesium sulfate
and then filtered and concentrated. A simple distillation provided, after cooling of the
distillate, 2.5g of the title compounds, S4, as a yellowꢀbrownish solid in 83% yield.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
Yellowꢀbrownish solid; Rf= 0.43 (1:1, hexanesꢀethyl acetate); H NMR (300 MHz,
CDCl3) δppm: 9.85 (s, 1H), 7.43 (d, J = 3.82 Hz, 1H), 7.36 (d, J = 3.87 Hz, 1H). These
obtained values are in agreement with previously reported spectroscopic data.28
5ꢀNitroꢀ2ꢀfuranmethanol, 2: Prepared according to a modified version of the protocol of
Emami.29 5ꢀnitrofurfural (S4, 2.22 g, 15.7 mmol) was dissolved in 47 mL of absolute
methanol and the solution was cooled to 0 °C. Then NaBH4 (0.65 g, 0.017 mol) was
slowly added and the solution was stirred for another 30 minutes. Once the reaction was
complete, the solvent was removed under reduced pressure and the residue was then
dissolved in a minimum amount of water. This solution was extracted with diethyl ether
(3 x 10 mL). The combined organic phases were then washed with water, dried with
magnesium sulfate, filtered and concentrated. This provided 0.78g of the title compound
as a yellow oil in a moderate 35% yield.
Yellow oil; Rf= 0.38 (1:1, hexanesꢀethyl acetate); 1H NMR (300 MHz, CDCl3) δppm: 7.28
(1 H, d, J = 3.71), 6.55 (1 H, d, J = 3.58), 4.70 (2 H, s), 2.68 (1 H, bs). These obtained
values are in agreement with previously reported spectroscopic data.30
Methyl 5ꢀbromoꢀ2ꢀfuroate S6 Prepared according to the approach of Torii and coꢀ
workers.31 Bromine (6.07 g, 0.038 mol) was carefully added (dropwise over a period of
15 minutes) to a solution of methyl furoate (S5, 3.2 g, 0.025 mol) stirred at 50°C under an
argon atmosphere in a flameꢀdried roundꢀbottom flask. The resulting dark
orange/brownish solution was additionally stirred for another 15 minutes at 50°C. The
reaction mixture was then poured into cold water (10 mL) and extracted with ethyl
acetate (2 x 50 mL). The combined extracts were washed with water (1 x 10 mL) and
brine (1 x 10 mL) prior to being dried with magnesium sulfate and concentrated. The
final product was purified by flash chromatography (10:1, hexanesꢀethyl acetate) to
obtain 4.5 g of S6 in 85% yield. The spectral data was consistent with literature reports.31
Rf= 0.17 (7:3, hexanesꢀethyl acetate).
5ꢀBromoꢀ2ꢀfuranmethanol 3 Prepared according to a modified version of the approach
of Aggarwal.32 A stirred solution of methyl 5ꢀbromoꢀ2ꢀfuroate (S6, 6.7 g, 4.6 mmol) in
anhydrous THF (111 mL) was cooled to 0°C. LiAlH4 (1.4 g, 5.1 mmol) was carefully
added to the reaction mixture, which was then stirred for a period of 15 minutes. Then,
the reaction mixture was warmed to room temperature over a period of 45 minutes prior
to a standard Fieser quench and filtration;33 the THF was mostly removed via rotary
evaporation. The resulting crude was diluted with ethyl acetate (200 mL) and washed
with water (50 mL) and brine (50 mL). It was then dried over magnesium sulfate and
concentrated and the final product, 3.9 g of a colorless oil, was obtained after flash
ACS Paragon Plus Environment

Page 23 of 33
The Journal of Organic Chemistry
1
2
3
4
5
chromatography (5:1, hexanesꢀethyl acetate) in 65% yield. Spectral data is consistent
with the published data.32 Rf= 0.50 (8:2, hexanesꢀethyl acetate).
6
7
8
9
Methyl 5ꢀmethoxyꢀ2ꢀfuroate S7 Prepared according to a modified version of the
protocol of Torii.31 Methyl 5ꢀbromoꢀ2ꢀfuroate (S6, 2.2 g, 10.7 mmol) was added to a
solution of sodium (0.6 g) and sodium iodide (0.03 g) in 30 ml of absolute methanol. The
solution was refluxed for 7 hours, then poured into cold water (100 mL). The mixture
was then extracted thrice with diethyl ether, and the combined organics were dried over
magnesium sulfate, filtered and concentrated. Chromatography of the resulting residue
(8:2, hexanesꢀethyl acetate) provided 0.8g of the title compound as an oily product in
48% yield.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Clear oil; Rf= 0.50 (8:2, hexanesꢀethyl acetate); 1H NMR (300 MHz, CDCl3) δppm: 7.13
(d, J = 3.69 Hz, 1H), 5.33 (d, J = 3.60 Hz, 1H), 3.96 (s, 3H), 3.86 (s, 3H). Spectral data is
in agreement with published data.34
5ꢀmethoxyꢀ2ꢀfuranmethanol 4 Prepared according to a modified version of the protocol
of Manly.35 A solution of methyl 5ꢀmethoxyꢀ2ꢀfuroate (S7, 0.45 g) in 2 mL of dry ether
was slowly added to a fastꢀstirring solution of LiAlH4 (0.14 g) in 4.5 mL of dry ether.
After a 1.5 hour reflux, 0.3 mL of water was carefully added prior to 1.9 mL of sodium
hydroxide. The reaction mixture was diluted with ether, and the phases separated. The
aqueous layer was then extracted several times with ether, and the combined ether
extracts were dried and evaporated to provide 150 mg of the title compound as a colorless
liquid in 40% yield. No further purification was required.
Colourless oil; Rf= 0.21 (3:7, hexanesꢀethyl acetate); 1H NMR (300 MHz, CDCl3) δppm
:
6.09 (d, J = 3.30 Hz, 1H), 5.03 (d, J = 3.30 Hz, 1H), 4.37 (s, 2H), 3.78 (s, 3H), 2.89 (bs,
1H). Spectral data is in agreement with published data.34
racꢀ(3aR,4R,7S,7aS)ꢀ4ꢀ(hydroxymethyl)ꢀ2ꢀ(4ꢀmethoxyphenyl)ꢀ3a,4,7,7aꢀtetrahydroꢀ
1Hꢀ4,7ꢀepoxyisoindoleꢀ1,3(2H)ꢀdione
(15endo
)
and
rac-(3aS,4R,7S,7aR)ꢀ4ꢀ
(hydroxymethyl)ꢀ2ꢀ(4ꢀmethoxyphenyl)ꢀ3a,4,7,7aꢀtetrahydroꢀ1Hꢀ4,7ꢀepoxyisoindoleꢀ
1,3(2H)ꢀdione (15exo) Nꢀ(pꢀmethoxyphenyl)ꢀMaleimide 5 (0.5 g, 2.67 mmol) and 2ꢀ
(hydroxymethyl) furan 1 (0.31 g, 3.2 mmol) were dissolved in anhydrous acetonitrile
under a nitrogen atmosphere in a flameꢀdried flask equipped with a magnetic stirringꢀbar.
The reaction was stirred at 35 ºC for 18 hours. When TLC indicated there was no longer
starting material present, the solvent was removed, and the reaction was concentrated
under reduced pressure for 1 hour. Endo and exo cycloadducts was separated by column
(6:4 to 4:6, hexanesꢀethyl acetate). We obtained two fractions, one contained 690 mg of
pure endo material (85% yield), while the second (<5%) contained 26 mg of a mixture of
the exo and endo isomers. Because these compounds have an inherently unstable nature
at ambient temperatures, they are stored at ꢀ20ºC until required.
racꢀ(3aR,4R,7S,7aS)ꢀ4ꢀ(hydroxymethyl)ꢀ2ꢀ(4ꢀmethoxyphenyl)ꢀ3a,4,7,7aꢀtetrahydroꢀ
1Hꢀ4,7ꢀepoxyisoindoleꢀ1,3(2H)ꢀdione (15endo) White solid. M.P.=128ꢀ134 ºC (CD3CN,
1
decomposition); Rf= 0.30 (2:8, hexanesꢀethyl acetate); H NMR (300 MHz, CD3CN):
δppm 7.17ꢀ7.12 (m, 4H), 6.71 (dd, J = 5.0, 1.5 Hz, 1H), 6.62 (dd, J = 5.8, 1.6 Hz, 1H),
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 24 of 33
1
2
3
4
5
6
7
8
5.45 (d, J = 6.7 Hz, 1H), 4.32 (dd, J = 12.94, 5.67 Hz, 1H), 4.20 (dd, J = 12.82, 6.37 Hz,
1H), 3.95 (s, 3H), 3.93ꢀ3.85 (m, 1H), 3.66 (d, J = 7.67 Hz, 1H), 3.31 (t, J = 5.88 Hz, 1H,
OH); 13C NMR (75 MHz, CD3CN): δppm 175.0, 174.7, 159.7, 135.7, 135.4, 128.0, 124.9,
114.0, 92.7, 79.5, 60.5, 55.3, 48.1, 45.7; HRMS (CI): Calculated for [M]+ (C16H15NO5):
301.0950. Found: 301.0944.
9
rac-(3aS,4R,7S,7aR)ꢀ4ꢀ(hydroxymethyl)ꢀ2ꢀ(4ꢀmethoxyphenyl)ꢀ3a,4,7,7aꢀtetrahydroꢀ
1Hꢀ4,7ꢀepoxyisoindoleꢀ1,3(2H)ꢀdione (15exo) Rf= 0.19 (1:9, hexanesꢀethyl acetate); H
1
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
NMR (300 MHz, CDCl3): δppm 7.07ꢀ6.89 (m, 4H), 6.65 (d, J = 5.71 Hz, 1H), 6.60ꢀ6.55
(m, 1H), 5.36 (d, J = 1.65 Hz, 1H), 4.17ꢀ4.12 (m, 2H), 3.82 (s, 3H), 3.13 (d, J = 6.57 Hz,
1H), 3.09 (d, J = 6.58 Hz, 1H), 2.84 (bs, 1H, OH). NOTE: Provided for convenience
only. The exo product was not isolated from residual endo material.
racꢀ(3aR,4R,7S,7aS)ꢀ4ꢀ(hydroxymethyl)ꢀ2ꢀ(4ꢀnitrophenyl)ꢀ3a,4,7,7aꢀtetrahydroꢀ1Hꢀ
4,7ꢀepoxyisoindoleꢀ1,3(2H)ꢀdione
(16endo
)
and
racꢀ(3aS,4R,7S,7aR)ꢀ4ꢀ
(hydroxymethyl)ꢀ2ꢀ(4ꢀnitrophenyl)ꢀ3a,4,7,7aꢀtetrahydroꢀ1Hꢀ4,7ꢀepoxyisoindoleꢀ
1,3(2H)ꢀdione (16exo) Nꢀ(pꢀnitrophenyl)ꢀMaleimide 6 (0.5 g, 2.30 mmol) and 2ꢀ
(hydroxymethyl) furan 1 (0.273 g, 2.80 mmol) were dissolved in anhydrous acetonitrile
under a nitrogen atmosphere in a flameꢀdried flask equipped with a magnetic stirringꢀbar.
The reaction was stirred at 40 ºC. After 4.5 hours, a new polar spot formed but starting
material was still present and the reaction was left to stir at the same temperature
overnight. When TLC indicated there was no longer starting material present, the solvent
was removed, and the reaction was concentrated under reduced pressure. A column with
(7:3 to 3:7, hexanesꢀethyl acetate) was used to separate the components. The pure endo
product was separated (100 mg, 13%), while a mixture of the endo and exo products (300
mg, 39 % yield) was obtained in a second fraction. Ultimately, there was a yield of 52%
endo/exo cycloadduct mixture. Since this mixture has an inherently unstable nature at
ambient temperatures, it is stored at ꢀ20ºC until required.
racꢀ(3aR,4R,7S,7aS)ꢀ4ꢀ(hydroxymethyl)ꢀ2ꢀ(4ꢀnitrophenyl)ꢀ3a,4,7,7aꢀtetrahydroꢀ1Hꢀ
4,7ꢀepoxyisoindoleꢀ1,3(2H)ꢀdione (16endo) Light yellow solid; M.P.=145ꢀ147 ºC
(CD3CN, decomposition); Rf= 0.30 (3:7, hexanesꢀethyl acetate); 1H NMR (300 MHz,
CD3CN): 8.29 (d, J = 8.62 Hz, 2H), 7.42 (d, J = 8.60 Hz, 2H), 6.60 (d, J = 5.76 Hz, 1H),
6.49 (d, J = 5.69 Hz, 1H), 5.34 (d, J = 5.50 Hz, 1H), 4.18 (dd, J = 12.90, 5.80 Hz, 1H),
4.06 (dd, J = 12.89, 6.30 Hz, 1H), 3.81 (dd, J = 7.60, 5.61 Hz, 1H), 3.59 (d, J = 7.75 Hz,
1H), 3.24 (t, J = 6.08 Hz, 1H, OH); 13C NMR (75 MHz, CDCl3): δppm 174.1, 173.8,
147.4, 137.6, 135.8, 135.6, 127.7, 124.4, 92.9, 79.6, 60.3, 48.3, 45.9; HRMS (CI):
Calculated for [M]+ (C15H12N2O6): 316.0695. Found: [M]+ 316.0697.
racꢀ(3aS,4R,7S,7aR)ꢀ4ꢀ(hydroxymethyl)ꢀ2ꢀ(4ꢀnitrophenyl)ꢀ3a,4,7,7aꢀtetrahydroꢀ1Hꢀ
4,7ꢀepoxyisoindoleꢀ1,3(2H)ꢀdione (16exo) White creamy solid; Rf= 0.25 (3:7, hexanesꢀ
ethyl acetate); 1H NMR (300 MHz, CDCl3): δppm 8.32 (d, J = 9.16 Hz, 2H), 7.58 (d, J =
9.05 Hz, 2H), 6.67ꢀ6.60 (m, 2H), 5.43 (d, J = 1.22 Hz, 1H), 4.19 (d, J = 7.29 Hz, 2H),
3.24 (d, J = 6.53 Hz, 1H), 3.19 (d, J = 6.70 Hz, 1H), 2.54 (t, J = 7.21 Hz, 1H, OH)
NOTE:
Provided for convenience only. The exo product was not isolated from residual endo
material.
ACS Paragon Plus Environment

Page 25 of 33
The Journal of Organic Chemistry
1
2
3
4
5
racꢀ(3aR,4R,7S,7aS)ꢀ2ꢀbenzylꢀ4ꢀ(hydroxymethyl)ꢀ3a,4,7,7aꢀtetrahydroꢀ1Hꢀ4,7ꢀ
epoxyisoindoleꢀ1,3(2H)ꢀdione
(17endo
)
and
rac-(3aS,4R,7S,7aR)ꢀ2ꢀbenzylꢀ4ꢀ
(hydroxymethyl)ꢀ3a,4,7,7aꢀtetrahydroꢀ1Hꢀ4,7ꢀepoxyisoindoleꢀ1,3(2H)ꢀdione (17exo).
Made according to the protocol of Gillies, Trant and Fan.9 Nꢀbenzyl maleimide24 (7, 2.0
g, 10.7 mmol) and 2ꢀ(hydroxymethyl) furan (1, 931 ꢂL, 1.05 g, 10.7 mmol) were
dissolved in anhydrous acetonitrile under a nitrogen atmosphere in a flameꢀdried flask
equipped with a magnetic stirringꢀbar. The reaction was stirred at 35 ºC for 14 hours.
When TLC indicated the reaction had reached equilibrium, the solvent was removed, and
the reaction was concentrated under reduced pressure for 1 hour. Crude NMR indicated a
ratio of (1:0.4:0.3) of endoꢀexoꢀunreacted maleimide. The crude material was then
purified by flash chromatography (6:4 to 4:6, hexanesꢀethyl acetate) to provide 1.61g (53
% yield) of the endo and 677 mg (22 % yield) of the exo product for a combined 75 %
isolated yield. Due to the inherent thermal instability, the material is stored at ꢀ20ºC until
needed.
racꢀ(3aR,4R,7S,7aS)ꢀ2ꢀbenzylꢀ4ꢀ(hydroxymethyl)ꢀ3a,4,7,7aꢀtetrahydroꢀ1Hꢀ4,7ꢀ
epoxyisoindoleꢀ1,3(2H)ꢀdione (17endo) Clear oil; Rf= 0.27 (6:4, hexanesꢀethyl acetate);
1H NMR (300 MHz, CDCl3): δppm 7.31ꢀ7.26 (m, 5H), 6.15 (dd, J = 5.8, 1.5 Hz, 1H),
6.06 (d, J = 5.8 Hz, 1H), 5.26 (dd, J = 5.5, 1.6 Hz, 1H), 4.47 (s, 2H), 4.25 (d, J = 12.2
Hz, 1H), 4.15 (d, J = 12.2 Hz, 1H), 3.63 (dd, J = 7.6, 5.5 Hz, 1H), 3.40 (d, J = 7.6 Hz,
1H), 2.11 (s, 1H). Spectral data is consistent with the published data.9
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
rac-(3aS,4R,7S,7aR)ꢀ2ꢀbenzylꢀ4ꢀ(hydroxymethyl)ꢀ3a,4,7,7aꢀtetrahydroꢀ1Hꢀ4,7ꢀ
epoxyisoindoleꢀ1,3(2H)ꢀdione (17exo) Colourless solid; Rf= 0.14 (6:4, hexanesꢀethyl
acetate); 1H NMR (300 MHz, CDCl3): δppm 7.33ꢀ7.26 (m, 5H), 6.61 (d, J = 5.7 Hz, 1H),
6.54 (dd, J = 5.7, 1.5 Hz, 1H), 5.28 (d, J = 1.7 Hz, 1H), 4.66 (bs, 2H), 4.09 (dd, J = 12.2,
8.8 Hz, 1H), 4.03 (dd, J = 12.2, 6.3 Hz, 1H), 3.02 (d, J = 6.5 Hz, 1H), 2.99 (d, J = 6.5
Hz, 1H), 2.76 (bt, J = 7.4 Hz, 1H, OH). Spectral data is consistent with the published
data.9
racꢀ(3aR,4R,7R,7aS)ꢀ4ꢀ(hydroxymethyl)ꢀ2ꢀ(4ꢀmethoxyphenyl)ꢀ7ꢀnitroꢀ3a,4,7,7aꢀ
tetrahydroꢀ1Hꢀ4,7ꢀepoxyisoindoleꢀ1,3(2H)ꢀdione (25endo) and racꢀ(3aS,4R,7R,7aR)ꢀ4ꢀ
(hydroxymethyl)ꢀ2ꢀ(4ꢀmethoxyphenyl)ꢀ7ꢀnitroꢀ3a,4,7,7aꢀtetrahydroꢀ1Hꢀ4,7ꢀ
epoxyisoindoleꢀ1,3(2H)ꢀdione (25exo) Nꢀ(pꢀmethoxyphenyl)ꢀMaleimide 5 (0.349 g, 1.72
mmol) and 5ꢀnitroꢀ2 furanmethanol 2 (0.295 g, 2.06 mmol) were dissolved in anhydrous
acetonitrile under a nitrogen atmosphere in a flameꢀdried flask equipped with a magnetic
stirringꢀbar. The reaction was stirred at 50ꢀ75 ºC for 8 hours, at which time TLC analysis
(3:7, hexanesꢀethyl acetate) indicated the formation of a new, polar spot (Rf = 0.13). The
reaction would not proceed to completion, and although we observed a new spot
consistent with the endo product form, it was negligible and was never able to be isolated
by chromatography. The reaction was stopped by cooling the mixture, and the solvent
was removed. Column chromatography (6:4 to 2:8 to 0.5:9.5, hexanesꢀethyl acetate) was
carried out, and provided 405 mg of the exo product in 63 % yield. The material was kept
at ꢀ20ºC until required.
racꢀ(3aS,4R,7R,7aR)ꢀ4ꢀ(hydroxymethyl)ꢀ2ꢀ(4ꢀmethoxyphenyl)ꢀ7ꢀnitroꢀ3a,4,7,7aꢀ
tetrahydroꢀ1Hꢀ4,7ꢀepoxyisoindoleꢀ1,3(2H)ꢀdione (25exo) White spongy solid;
1
M.P.=173ꢀ174 ºC (CDCl3, decomposition); Rf= 0.13 (3:7, hexanesꢀethyl acetate); H
NMR 1H NMR (300 MHz, CDCl3): δppm 7.19 (d, J = 9.03 Hz, 2H), 6.98 (d, J = 8.98 Hz,
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 26 of 33
1
2
3
4
5
6
7
8
9
2H), 6.96 (d, J = 5.31 Hz, 1H), 6.90 (d, J = 5.58 Hz, 1H), 4.22 (d, J = 2.65 Hz, 1H), 4.19
(d, J = 1.56 Hz, 1H), 3.83 (s, 3H), 3.77 (d, J = 6.64 Hz, 1H), 3.42 (d, J = 6.64 Hz, 1H),
2.71 (t, J = 7.46 Hz, 1H, OH); C NMR (300 MHz, MeOD): δppm 172.8, 171.6, 160.2,
141.5, 134.8, 128.2, 124.7, 114.4, 112.3, 91.8, 59.2, 55.1, 52.6, 50.9. HRMS (CI):
Calculated for [M]+ (C16H14N2O7): 346.0801, [M + H]+ (C16H15N2O7): 347.0874. Found
[M + H]+: 347.0877.
13
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
racꢀ(3aR,4R,7R,7aS)ꢀ4ꢀ(hydroxymethyl)ꢀ7ꢀnitroꢀ2ꢀ(4ꢀnitrophenyl)ꢀ3a,4,7,7aꢀ
tetrahydroꢀ1Hꢀ4,7ꢀepoxyisoindoleꢀ1,3(2H)ꢀdione (26endo) Nꢀ(pꢀnitrophenyl)ꢀMaleimide
2 (0.254 g, 1.16 mmol) and 5ꢀnitroꢀ2ꢀfuranmethanol 6 (0.200 g, 1.40 mmol) were
dissolved in anhydrous acetonitrile under a nitrogen atmosphere in a flameꢀdried flask
equipped with a magnetic stirringꢀbar. The reaction was stirred at 40 ºC for 4 hours, then
at 60 ºC for an additional 4 hours, at which point only starting material was observed by
TLC (3:7, hexanesꢀethyl acetate). The temperature was accordingly increased to 80 ºC,
and was left to stir at this temperature overnight (12 hours). TLC indicated moderate
conversion, and the solvent was removed, and the reaction mixture was concentrated.
Column purification was then performed (6:4 to 0.5:9.5, hexanesꢀethyl acetate). 162 mg
of the polar spot was obtained in 31% yield (Rf = 0.16) and was determined to be exo
product. A negligible amount of endo material was observed by crude NMR but was not
able to be readily isolated from the starting materials. Product was stored at ꢀ20ºC until
required.
racꢀ(3aS,4R,7R,7aR)ꢀ4ꢀ(hydroxymethyl)ꢀ7ꢀnitroꢀ2ꢀ(4ꢀnitrophenyl)ꢀ3a,4,7,7aꢀ
tetrahydroꢀ1Hꢀ4,7ꢀepoxyisoindoleꢀ1,3(2H)ꢀdione (26exo) Offꢀwhite spongy solid;
1
M.P.=170ꢀ171.3 ºC (CD3CN, decomposition); Rf= 0.16 (3:7, hexanesꢀethyl acetate); H
NMR (300 MHz, CD3CN): δppm 8.34 (d, J = 9.05 Hz, 2H), 7.54 (d, J = 9.11 Hz, 2H),
6.93 (d, J = 5.73 Hz, 1H), 6.90 (d, J = 5.85 Hz, 1H), 4.26 (dd, J = 13.21, 6.39 Hz, 1H),
4.05 (dd, J = 13.32, 5.80 Hz, 1H), 3.94 (d, J = 6.57 Hz, 1H), 3.46 (d, J = 6.70 Hz, 1H),
13
3.44 (t, J = 5.93 Hz, 1H, OH); C NMR (100 MHz, CDCl3): δppm 171.7, 170.7, 147.6,
141.3, 137.0, 134.8, 127.5, 124.6, 112.0, 91.7, 59.1, 52.7, 51.0. HRMS (CI): Calculated
for [M]+ (C15H11N3O8): 361.0546; [M + H]+ (C15H12N3O8): 362.0624. Found [M]+:
362.0614.
racꢀ(3aR,4R,7R,7aS)ꢀ2ꢀbenzylꢀ4ꢀ(hydroxymethyl)ꢀ7ꢀnitroꢀ3a,4,7,7aꢀtetrahydroꢀ1Hꢀ
4,7ꢀepoxyisoindoleꢀ1,3(2H)ꢀdione (27endo
)
and racꢀ(3aS,4R,7R,7aR)ꢀ2ꢀbenzylꢀ4ꢀ
hydroxymethyl)ꢀ7ꢀnitroꢀ3a,4,7,7aꢀtetrahydroꢀ1Hꢀ4,7ꢀepoxyisoindoleꢀ1,3(2H)ꢀdione
(27exo) Nꢀbenzyl maleimide 7 (0.322 g, 1.72 mmol) and 5ꢀnitroꢀ2ꢀfuranmethanol 2 (0.295
g, 2.06 mmol) were dissolved in anhydrous acetonitrile under a nitrogen atmosphere in a
flameꢀdried flask equipped with a magnetic stirringꢀbar. The reaction was stirred at 65ꢀ70
ºC for 16 hours. At that point, TLC analysis (2:8, hexanesꢀethyl acetate) showed the
formation a new nonpolar spot (endo; Rf = 0.27) and the formation of a polar spot (exo;
Rf = 0.14). At this point, the reaction was concentrated under reduced pressure, and
purified by flash chromatography (3:7 to 1:9, hexanesꢀethyl acetate). The first fraction
contained 438 mg of product (69% yield) endo product and 129 mg (21% yield) of the
exo product. Both compounds were stored at ꢀ20ºC until required.
racꢀ(3aR,4R,7R,7aS)ꢀ2ꢀbenzylꢀ4ꢀ(hydroxymethyl)ꢀ7ꢀnitroꢀ3a,4,7,7aꢀtetrahydroꢀ1Hꢀ
4,7ꢀepoxyisoindoleꢀ1,3(2H)ꢀdione (27endo) White crystal; M.P.= 99.5ꢀ107 ºC (CDCl3,
ACS Paragon Plus Environment

Page 27 of 33
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
decomposition); Rf= 0.27 (2:8, hexanesꢀethyl acetate); 1H NMR (300 MHz, CDCl3): δppm
7.35ꢀ7.27 (m, 5H), 6.46 (d, J = 5.75 Hz, 1H), 6.23 (d, J = 5.75 Hz, 1H), 4.55 (d, J =
13.96 Hz, 1H), 4.50 (d, J = 13.94 Hz, 1H), 4.29 (dd, J = 13.13, 6.18 Hz, 1H), 4.19 (dd, J
= 13.14, 6.94 Hz, 1H), 3.96 (d, J = 7.95 Hz, 1H), 3.84 (d, J = 7.95 Hz, 1H), 2.16 (t, J =
6.60 Hz, 1H, OH); 13C NMR (300 MHz, CDCl3): δppm 172.6, 170.7, 137.2, 134.9, 132.7,
129.3, 128.8, 128.6, 112.4, 91.7, 60.8, 57.6, 51.0, 48.1, 43.0. HRMS (CI): Calculated for
[M]+ (C16H14N2O6): 330.0852; [M+ H]+ (C16H15N2O6): 331.0930. Found [M + H]+:
331.0939.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
racꢀ(3aS,4R,7R,7aR)ꢀ2ꢀbenzylꢀ4ꢀhydroxymethyl)ꢀ7ꢀnitroꢀ3a,4,7,7aꢀtetrahydroꢀ1Hꢀ
4,7ꢀepoxyisoindoleꢀ1,3(2H)ꢀdione (27exo) White crystal; Rf= 0.14 (2:8, hexanesꢀethyl
acetate); 1H NMR (300 MHz, CDCl3): δppm 7.38ꢀ7.27 (m, 5H), 6.89 (d, J = 5.62 Hz, 1H),
6.83 (d, J = 5.52 Hz, 1H), 4.71 (d, J = 14.23 Hz, 1H), 4.63 (d, J = 14.31 Hz, 1H), 4.19ꢀ
3.99 (m, 2H), 3.64 (d, J = 6.48 Hz, 1H), 3.26 (d, J = 6.47 Hz, 1H), 2.75ꢀ2.68 (m, 1H,
OH); 13C NMR (300 MHz, CDCl3): δppm 173.4, 170.6, 141.3, 135.4, 134.9, 129.1, 128.7,
128.6, 111.4, 9.4, 60.3, 51.7, 50.9, 43.5.
racꢀ(3aR,4S,7R,7aR)ꢀ4ꢀbromoꢀ7ꢀ(hydroxymethyl)ꢀ2ꢀ(4ꢀmethoxyphenyl)ꢀ3a,4,7,7aꢀ
tetrahydroꢀ1Hꢀ4,7ꢀepoxyisoindoleꢀ1,3(2H)ꢀdione (35endo) and racꢀ(3aS,4S,7R,7aS)ꢀ4ꢀ
bromoꢀ7ꢀ(hydroxymethyl)ꢀ2ꢀ(4ꢀmethoxyphenyl)ꢀ3a,4,7,7aꢀtetrahydroꢀ1Hꢀ4,7ꢀ
epoxyisoindoleꢀ1,3(2H)ꢀdione (35exo) Nꢀ(pꢀmethoxyphenyl)ꢀMaleimide 5 (0.8 g, 3.90
mmol) and 5ꢀBromoꢀ2ꢀfuranmethanol 3 (0.60 g, 3.00 mmol) were dissolved in anhydrous
acetonitrile under a nitrogen atmosphere in a flameꢀdried flask equipped with a magnetic
stirringꢀbar. The reaction was stirred at room temperature for 3 hours, then the progress
of reaction was monitored every 2 hours and the temperature was increased from 35 ºC to
75 ºC over 6 hours, and held at the highest temperature for an additional 10 hours. The
reaction was concentrated under reduced pressure and purified by column (6:4 to 2:8,
hexanesꢀethyl acetate) to separate the starting material from 851 mg of the nonꢀpolar
endo (Rf = 0.50, 6:4, hexanesꢀethyl acetate) product (61% yield) and 63 mg of a polar exo
(Rf = 0.17, 6:4, hexanesꢀethyl acetate) product in < 5% yields. The mass balance included
some debrominated material, 15 and starting materials. These products stored at ꢀ20ºC
until required.
racꢀ(3aR,4S,7R,7aR)ꢀ4ꢀbromoꢀ7ꢀ(hydroxymethyl)ꢀ2ꢀ(4ꢀmethoxyphenyl)ꢀ3a,4,7,7aꢀ
tetrahydroꢀ1Hꢀ4,7ꢀepoxyisoindoleꢀ1,3(2H)ꢀdione (35endo) Amorphous White solid. Rf=
0.50 (6:4, hexanesꢀethyl acetate); 1H NMR (300 MHz, CD3CN): δppm 7.06ꢀ6.89 (m, 4H),
6.64 (d, J = 5.54 Hz, 1H), 6.53 (d, J = 5.44 Hz, 1H), 4.14 (dd, J = 13.38, 6.13 Hz, 1H),
4.02 (dd, J = 12.83, 6.25 Hz, 1H), 3.87 (d, J = 7.52 Hz, 1H), 3.79 (s, 3H), 3.74 (d, J =
13
8.01 Hz, 1H), 3.35 (t, J = 5.99 Hz, 1H, OH); C NMR (75 MHz, CD3CN): δppm 174.9,
174.0, 161.4, 140.9, 138.3, 129.7, 126.0, 115.9, 93.5, 89.8, 61.5, 58.2, 56.8, 46.7. HRMS
(CI): Calculated for [M]+ (C16H14BrNO5): 379.0055. Found: 379.0046.
racꢀ(3aS,4S,7R,7aS)ꢀ4ꢀbromoꢀ7ꢀ(hydroxymethyl)ꢀ2ꢀ(4ꢀmethoxyphenyl)ꢀ3a,4,7,7aꢀ
tetrahydroꢀ1Hꢀ4,7ꢀepoxyisoindoleꢀ1,3(2H)ꢀdione (35exo) White solid; Rf= 0.17 (6:4,
1
hexanesꢀethyl acetate); H NMR (300 MHz, CDCl3): δppm 7.20 (d, J = 8.64 Hz, 1H),
6.97 (d, J = 8.67 Hz, 1H), 6.64 (d, J = 7.74 Hz, 1H), 6.62 (d, J = 7.89 Hz, 1H), 4.01 (d, J
= 11.56 Hz, 1H), 3.84 (d, J = 11.22 Hz, 1H), 3.83 (s, 3H), 3.36ꢀ3.33 (m, 2H); 13C NMR
(75 MHz, CDCl3): δppm 171.8, 170.5, 159.8, 142.1, 138.7, 127.6, 123.9, 114.5, 89.4, 89.0,
55.5, 54.8, 51.5, 27.4.
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 28 of 33
1
2
3
4
5
6
7
8
9
racꢀ(3aR,4S,7R,7aR)ꢀ4ꢀbromoꢀ7ꢀ(hydroxymethyl)ꢀ2ꢀ(4ꢀnitrophenyl)ꢀ3a,4,7,7aꢀ
tetrahydroꢀ1Hꢀ4,7ꢀepoxyisoindoleꢀ1,3(2H)ꢀdione (36endo) Nꢀ(pꢀnitrophenyl)ꢀMaleimide
6 (990 mg, 5.6 mmol) and 5ꢀBromoꢀ2ꢀfuranmethanol 3 (1.03 g, 4.7 mmol) were
dissolved in anhydrous acetonitrile under a nitrogen atmosphere in a flameꢀdried flask
equipped with a magnetic stirringꢀbar. The reaction was stirred at 50 ºC for 72 hours until
further conversion was no longer noted by TLC (6:4 hexanesꢀethyl acetate). The reaction
mixture was concentrated, and the residue purified by column chromatography (6:4,
hexanesꢀethyl acetate) in two columns successively (the first column provided product
contaminated with the starting materials). Crude 1H NMR had shown the presence of
both endo and exo derivatives, albeit in low conversion. The small amount of exo was
estimated to be well under 3% of the mass balance and was not isolated. The
chromatography provided 530 mg of the title compound as a colourless solid in 26%
yield.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
racꢀ(3aR,4S,7R,7aR)ꢀ4ꢀbromoꢀ7ꢀ(hydroxymethyl)ꢀ2ꢀ(4ꢀnitrophenyl)ꢀ3a,4,7,7aꢀ
tetrahydroꢀ1Hꢀ4,7ꢀepoxyisoindoleꢀ1,3(2H)ꢀdione (36endo) Amorphous white solid; Rf=
1
0.20 (6:4, hexanesꢀethyl acetate); H NMR (300 MHz, CDCl3): δppm 8.31 (d, J = 8.75
Hz, 1H), 7.43 (d, J = 8.48 Hz, 1H), 6.65 (d, J = 5.46 Hz, 1H), 6.51 (d, J = 5.48 Hz, 1H),
4.36 (d, J = 12.88 Hz, 1H), 4.24 (d, J = 13.02 Hz, 1H), 4.00 (d, J = 7.90 Hz, 1H), 3.91
(d, J = 8.00 Hz, 1H); 13C NMR (75 MHz, CDCl3): δppm 171.9, 170.7, 147.4, 140.2,
136.0, 127.1, 126.8, 124.6, 91.9, 87.6, 61.0, 56.8, 48.4. HRMS (CI): Calculated for [M]+
(C15H11BrN2O6): 393.9800, [M + H]+ (C15H12BrN2O6): 394.9879. Found: 394.9872.
racꢀ(3aR,4S,7R,7aR)ꢀ2ꢀbenzylꢀ4ꢀbromoꢀ7ꢀ(hydroxymethyl)ꢀ3a,4,7,7aꢀtetrahydroꢀ
1Hꢀ4,7ꢀepoxyisoindoleꢀ1,3(2H)ꢀdione (37endo) and racꢀ(3aS,4S,7R,7aS)ꢀ2ꢀbenzylꢀ4ꢀ
bromoꢀ7ꢀ(hydroxymethyl)ꢀ3a,4,7,7aꢀtetrahydroꢀ1Hꢀ4,7ꢀepoxyisoindoleꢀ1,3(2H)ꢀ
dione (37exo) Nꢀbenzyl maleimide 7 (0.70 g, 3.8 mmol) and 5 Bromoꢀ2ꢀfuranmethanol 3
(0.80 g, 4.5 mmol) were dissolved in anhydrous acetonitrile under a nitrogen atmosphere
in a flameꢀdried flask equipped with a magnetic stirringꢀbar. The reaction was stirred at
room temperature for 7 hours, a TLC (7:3, hexanesꢀethyl acetate) showing starting
material was left, so the temperature was increased slowly (55 ºC to 70 ºC) and the
progress of reaction was monitored every 2 hours. The reaction was left for an additional
14 hours at 70 ºC when TLC showed some conversion. The solvent was removed, and the
reaction mixture was concentrated. After column chromatography purification of reaction
mixture (7:3, hexanesꢀethyl acetate), 403 mg of an inseparable mixture of the endo and
exo (2:1) cycloadducts as a white solid was obtained with an overall yield of 31%. In
none of the more than 10 solvent mixtures examined was separation observed. The
mixture was kept at ꢀ20ºC until required.
racꢀ(3aR,4S,7R,7aR)ꢀ2ꢀbenzylꢀ4ꢀbromoꢀ7ꢀ(hydroxymethyl)ꢀ3a,4,7,7aꢀtetrahydroꢀ
1Hꢀ4,7ꢀepoxyisoindoleꢀ1,3(2H)ꢀdione (37endo) Rf= 0.29 (7:3, hexanesꢀethyl acetate); H
1
NMR (300 MHz, CDCl3): δppm 7.37ꢀ7.27 (m, 5H), 6.12 (d, J = 5.51 Hz, 1H), 5.97 (d, J
= 5.24 Hz, 1H), 4.52ꢀ4.46 (m, 2H), 4.24 (dd, J = 12.40, 1.86 Hz, 1H), 4.13 (dd, J =
12.44, 2.95 Hz, 1H), 3.79ꢀ3.12 (m, 2H), 3.25ꢀ3.20 (m, 1H, OH); 13C NMR (75 MHz,
CDCl3): δppm 172.9, 172.2, 138.6, 135.2, 135.0, 129.1, 128.6, 128.2, 90.6, 89.1, 68.9,
56.0, 51.6, 47.96, 42.5.
ACS Paragon Plus Environment

Page 29 of 33
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
racꢀ(3aS,4S,7R,7aS)ꢀ2ꢀbenzylꢀ4ꢀbromoꢀ7ꢀ(hydroxymethyl)ꢀ3a,4,7,7aꢀtetrahydroꢀ1Hꢀ
4,7ꢀepoxyisoindoleꢀ1,3(2H)ꢀdione (37exo) Rf= 0.28 (7:3, hexanesꢀethyl acetate); 1H
NMR (300 MHz, CDCl3): δppm 7.37ꢀ7.27 (m, 5H), 6.12 (d, J = 5.51 Hz, 1H), 5.97 (d, J
= 5.24 Hz, 1H), 4.71 (d, J = 14.35 Hz, 1H), 4.66 (d, J = 14.37 Hz, 1H), 3.93 (d, J =
11.41 Hz, 1H), 3.65 (dd, J = 14.15, 7.79 Hz, 1H), 3.77ꢀ3.73 (m, 1H); 3.25ꢀ3.20 (m, 1H,
OH); 13C NMR (75 MHz, CDCl3): δppm 172.0, 170.9, 139.2, 135.1, 135.0, 128.7, 128.3,
128.0, 90.5, 88.8, 69.1, 54.8, 48.04, 42.9, 27.3.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Endo/exo mixture HRMS (CI): Calculated for [M]+ (C16H14BrNO4): 363.0106, [M +
H]+ (C16H15BrNO4): 364.0184. Found: 364.0176.
rac-(3aR,4R,7R,7aS)ꢀ4ꢀ(hydroxymethyl)ꢀ7ꢀmethoxyꢀ2ꢀ(4ꢀmethoxyphenyl)ꢀ3a,4,7,7aꢀ
tetrahydroꢀ1Hꢀ4,7ꢀepoxyisoindoleꢀ1,3(2H)ꢀdione (45endo) and rac-(3aR,4R,7R,7aS)ꢀ4ꢀ
(hydroxymethyl)ꢀ7ꢀmethoxyꢀ2ꢀ(4ꢀmethoxyphenyl)ꢀ3a,4,7,7aꢀtetrahydroꢀ1Hꢀ4,7ꢀ
epoxyisoindoleꢀ1,3(2H)ꢀdione (45exo) Nꢀ(pꢀmethoxyphenyl)ꢀMaleimide 5 (0.32 g, 2.5
mmol) and 5ꢀmethoxyꢀ2ꢀfuranmethanol 4 (0.43 g, 2.1 mmol) were dissolved in
anhydrous acetonitrile under a nitrogen atmosphere in a flameꢀdried flask equipped with
a magnetic stirringꢀbar. The reaction was stirred at 45 ºC for 4 hours; TLC analysis (2:8,
hexanesꢀethyl acetate) indicated that a new nonpolar spot (Rf = 0.50) had formed with
some starting material remaining. The temperature was then increased to 55ꢀ60 ºC and
the reaction stirred for 14 hours. TLC showed a new polar spot (Rf = 0.25) with no furan
remaining. The solvent was then removed under reduced pressure. Column purification
(4:6 to 1.5:8.5, hexanesꢀethyl acetate) was carried out to obtain 191 mg of the endo
product (25% yield) and 26 mg (4% yield) of the exo product. Due to the fact that these
compounds have an inherently unstable nature at ambient temperatures, they are stored at
ꢀ20ºC until required.
rac-(3aR,4R,7R,7aS)ꢀ4ꢀ(hydroxymethyl)ꢀ7ꢀmethoxyꢀ2ꢀ(4ꢀmethoxyphenyl)ꢀ3a,4,7,7aꢀ
tetrahydroꢀ1Hꢀ4,7ꢀepoxyisoindoleꢀ1,3(2H)ꢀdione (45endo) White solid; Rf= 0.50 (2:8,
hexanesꢀethyl acetate); 1H NMR (300 MHz, CD3CN): δppm 7.03 (d, J = 9.10 Hz, 2H),
6.98 (d, J = 9.14 Hz, 2H), 6.63 (d, J = 5.82 Hz, 1H), 6.58 (d, J = 5.83 Hz, 1H), 4.10 (dd,
J = 12.91, 5.81 Hz, 1H), 3.99 (dd, J = 12.91, 6.13 Hz, 1H), 3.81 (s, 3H), 3.71 (d, J = 7.87
13
Hz, 1H), 3.56 (s, 3H), 3.51 (d, J = 7.90 Hz, 1H), 3.20 (t, J = 6.02, Hz, 1H, OH); C
NMR (125 MHz, CDCl3): δppm 174.5, 173.8, 159.8, 137.7, 134.9, 128.2, 128.2, 124.9,
114.4, 86.8, 60.7, 55.3, 54.1, 50.2, 48.9. HRMS (CI): Calculated for [M]+ (C17H17NO6):
331.1056, [M + H]+: 332.1134. Found [M + H]+: 332.1127
rac-(3aR,4R,7R,7aS)ꢀ4ꢀ(hydroxymethyl)ꢀ7ꢀmethoxyꢀ2ꢀ(4ꢀmethoxyphenyl)ꢀ3a,4,7,7aꢀ
tetrahydroꢀ1Hꢀ4,7ꢀepoxyisoindoleꢀ1,3(2H)ꢀdione (45exo) White solid; Rf= 0.25 (2:8,
hexanesꢀethyl acetate); 1H NMR (300 MHz, CDCl3): δppm 7.21 (d, J = 8.94 Hz, 2H), 6.97
(d, J = 9.01 Hz, 2H), 6.79 (d, J = 5.66 Hz, 1H), 6.56 (d, J = 5.66 Hz, 1H), 4.14 (d, J =
12.65 Hz, 1H), 4.08 (d, J = 12.71 Hz, 1H), 3.82 (s, 3H), 3.63 (s, 3H), 3.27 (d, J = 6.59
Hz, 1H), 3.21 (d, J = 6.57 Hz, 1H).
racꢀ(3aR,4R,7R,7aS)ꢀ4ꢀ(hydroxymethyl)ꢀ7ꢀmethoxyꢀ2ꢀ(4ꢀnitrophenyl)ꢀ3a,4,7,7aꢀ
tetrahydroꢀ1Hꢀ4,7ꢀepoxyisoindoleꢀ1,3(2H)ꢀdione (46endo) and racꢀ(3aS,4R,7R,7aR)ꢀ4ꢀ
(hydroxymethyl)ꢀ7ꢀmethoxyꢀ2ꢀ(4ꢀnitrophenyl)ꢀ3a,4,7,7aꢀtetrahydroꢀ1Hꢀ4,7ꢀ
epoxyisoindoleꢀ1,3(2H)ꢀdione (46exo) Nꢀ(pꢀnitrophenyl)ꢀMaleimide 6 (458 mg, 2.1
mmol) and 5ꢀmethoxyꢀ2ꢀfuranmethanol 4 (320 mg g, 2.5 mmol) were dissolved in
ACS Paragon Plus Environment