Photochemical Reaction of o-Phthalaldehyde with Fullerene C60: The Stimulus for the Phthalide Additions to C60

Article abstract of DOI:10.1002/ejoc.201901363

A new, simple, and rapid photochemical reaction of o-phthalaldehyde with C₆₀ has been disclosed. This reaction afforded exclusively the unanticipated C₆₀-phthalide derivative 11 in good yield. The formation of this product is consistent with the intervention of a key free-radical intermediate –a benzylic radical– in the course of the reaction. The structure of 11 has been determined by the combined use of experimental and theoretical NMR studies. In the course of this study we further disclosed a new mode of radical reactivity of C₆₀ with a series of substituted phthalides (2, 15–17) catalyzed by tetrabutylammonium decatungstate.

Full text of DOI:10.1002/ejoc.201901363

DOI: 10.1002/ejoc.201901363
Communication
Photochemistry of Fullerenes
Photochemical Reaction of o-Phthalaldehyde with Fullerene C60:
The Stimulus for the Phthalide Additions to C60
Manolis D. Tzirakis,*[a] Nikitas G. Malliaros,[a] and Michael Orfanopoulos*[a]
Abstract: A new, simple, and rapid photochemical reaction of been determined by the combined use of experimental and
o-phthalaldehyde with C60 has been disclosed. This reaction af- theoretical NMR studies. In the course of this study we further
forded exclusively the unanticipated C60-phthalide derivative 11 disclosed a new mode of radical reactivity of C60 with a series
in good yield. The formation of this product is consistent with of substituted phthalides (2, 15–17) catalyzed by tetrabutyl-
the intervention of a key free-radical intermediate –a benzylic ammonium decatungstate.
radical– in the course of the reaction. The structure of 11 has
Introduction
to give phthalide 2.[2] The involvement of 7 has been confirmed
by laser flash photolysis studies performed by Scaiano et al.[4]
It was shown that the photochemistry of o-phthalaldehyde in-
volves an intramolecular hydrogen transfer of two isomeric trip-
let states of 1 to afford a 1,4-biradical 9 (Figure 1), which then
decays to an isomeric mixture of enols (E- and Z-7).[4] These
enols revert back to 1 quite efficiently, and it was proposed
that the formation of 2 and 3 should arise from an independent
reaction pathway that involves the intermediacy of cyclic enol
10 (Figure 1). The suggested ketene-enol (7) involved in the
decay of biradical 9 (from o-phthalaldehyde) has been further
characterized by IR matrix-isolation spectroscopy.[5,6]
The photochemical isomerization of o-phthalaldehyde (1) to
phthalide (2) and an isomeric mixture of the corresponding di-
mers (3), has been the subject of several studies (Scheme 1).[1–7]
Scheme 1. Photochemical transformation of o-phthalaldehyde (1) to phthal-
ide (2) and its dimer (3).
In general, these studies have given prominence to the im-
portant role of radical intermediates involved in this transforma-
tion (Figure 1). Nonetheless, there is a considerable controversy
concerning the exact operating mechanism. Kagan has sug-
gested the intermediacy of biradical intermediate 4 (Figure 1).[1]
It was further shown that the relative ratio of 2 to 3 depends
largely on the solvent used. For example, chlorinated solvents,
such as CCl4 and CHCl3, favor the formation of 2, while dimer
3 is favored in heptane or benzene.[1] Harrison et al. have
shown that the irradiation of 1 at 77 K leads to the generation
of radical intermediates 5 and 6 (Figure 1), which are stable at
this low temperature.[3] They further proposed that the ob-
served photochemistry of 1 that leads to 2 and 3 proceeds
from the short-lived singlet excited state of 1.[3] In contrast,
Pappas and Blackwell have suggested that the phototransfor-
mations of 1 may proceed through the ketene-enol intermedi-
ate 7 (Figure 1) which can undergo an intramolecular cycliza-
tion to afford 8 (Figure 1), followed by ionic hydrogen transfer
Figure 1. Suggested intermediate species 4–10 involved in the photochemi-
cal transformation of o-phthalaldehyde (1) to phthalide (2).
The aim of the present study was to investigate the photo-
chemical reaction of 1 in the presence of [60]fullerene. We envi-
sioned that C60 –a well-known radical scavenger–[8] would react
promptly with the free-radical intermediates formed upon irra-
diation of o-phthalaldehyde, especially with acyl radical 5 as
shown in a previous study that regards the prompt reactivity of
C60 towards acyl radicals (Figure 1).[9] This reaction not only
provides new evidence on the intermediate species involved in
[a] Department of Chemistry, University of Crete
Voutes Campus, 71003, Heraklion, Crete, Greece
E-mail: tzirakis@chemistry.uoc.gr
morfanop@uoc.gr
Supporting information and ORCID(s) from the author(s) for this article are
available on the WWW under https://doi.org/10.1002/ejoc.201901363.
Eur. J. Org. Chem. 0000, 0–0
1
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
the photochemical transformation of 1[1–7] but, at the same solvent increased considerably the rate of the reaction (Table 1,
time, also provide a novel synthetic route for the functionaliza- entry 5), thus signifying the intermediacy of a polar transient
tion of [60]fullerene and access to new nanostructures with po- species in the reaction mechanism. We further confirmed that
tentially interesting properties.
the reaction proceeds efficiently also in ambient diffuse sun-
light to afford 11 in 43 % isolated yield (Table 1, entries 7–8).
Control experiments showed that no reaction occurred in the
absence of light for several hours. In addition, irradiation of 1
in the absence of C60 in chlorobenzene, benzene or a mixture of
chlorobenzene/CH3CN (85:15) solution afforded quantitatively
phthalide 2 (see Experimental Section).
Results and Discussion
At the onset of our studies we investigated the reaction of C60
(20 mg, 0.028 mmol) with 1 by simply irradiating a solution of
C60 with 15 equiv. of 1 in chlorobenzene (65 mL), using a 300 W
Xenon lamp (Table 1, entry 1) at 5 °C. The reaction was followed
The structural assignment of 11 was deduced from the com-
by HPLC analysis of aliquots taken at varying time intervals. bined use of 1H/13C NMR, UV/Vis and FTIR spectroscopy, as well
Under these conditions, the unanticipated monofunctionalized as from mass spectrometry [see the Experimental Section and
fullerene derivative 11 was detected by HPLC in 12 % yield after Supporting Information (SI)]. Additional evidence for the pro-
6 min of irradiation. The reaction was greatly accelerated by posed structure of 11 was obtained from the correlation be-
using additional equivalents of 1 (Table 1, entries 2–3). Also, in tween the DFT-calculated (δcalcd; ppm) and experimental (δexpt
;
1
contrast to C6H6 which did not alter the outcome of the reac- ppm) H and 13C NMR chemical shifts (Table 2). All calculations
tion (Table 1, entry 4), the addition of CH3CN as a polar co- were performed using the GAUSSIAN 09 package.[10] The ener-
Table 1. Reaction of C60 with 1.[a]
Entry
Equiv. of 1
Solvent
Irradiation time
Yield[%][b,c]
1
2
3
4
15
50
100
50
50
C6H5Cl
C6H5Cl
C6H5Cl
C6H6
C6H5Cl/CH3CN (85:15)
C6H5Cl/CH3CN (85:15)
C6H5Cl/CH3CN (85:15)
6 min
6 min
3 min
6 min
3 min
3 min
21 h
12
30 (19)
30 (21)
30
33
56
5
6
100
50
7[d]
38
72 h
21 h
55 (43)
47 (33)
8[d]
100
C6H5Cl/CH3CN (85:15)
[a] For experimental details see the Experimental Section. [b] Yield based on integration of HPLC peaks. [c] Numbers in parentheses correspond to isolated
yield. [d] The reaction was carried out with exposure to ambient light in 30 mL of solvent.
Table 2. Calculated [δcalcd; ppm; B3LYP/6-311+G(2d,p)] and experimental (δexpt, in ppm) 1H and 13C chemical shift values of 11 in chloroform.
Atom[a]
δexpt
δcalcd
|δcalcd – δexpt|
Atom[a]
δexpt
δcalcd
|δcalcd – δexpt|
[b]
[c]
[b]
[d]
C(1)
68.33
55.75
87.18
169.80
124.70
135.02
69.01
57.02
84.62
172.21
124.00
133.32
0.68
1.27
2.56
2.41
0.70
H(61)
H(64)
H(70)
H(73)
H(75)
H(76)
6.31
7.16
8.27
7.95
8.26
7.83
6.45
6.76
8.42
7.87
8.55
7.72
0.14
0.40
0.15
0.08
0.29
0.11
C(11)
C(62)
C(66)
C(67)
C(71)
1.70
C(72)
C(74)
126.85
131.02
125.12
MAE[e] = 0.195
1.73
2.20
128.82
MAE[e] = 1.6563
[a] All atoms are numbered according to the numbering system introduced in Figure 2. [b] NMR spectra were recorded on a 500 MHz spectrometer, in CDCl3
at 25 °C. [c] Chemical shifts were derived from application of scaling factors (slope = –0.9854, intercept = 173.92) to the 13C NMR shielding tensors computed
at the B3LYP/6-311+G(2d,p) level of theory, followed by Boltzmann averaging over all three conformers at 298 K (further details are provided in the Experimen-
tal Section). [d] Chemical shifts were derived from application of scaling factors (slope = –0.5899, intercept = 28.358) to the 1H NMR shielding tensors
computed at the B3LYP/6-311+G(2d,p) level of theory, followed by Boltzmann averaging over all three conformers at 298 K (further details are provided in
the Experimental Section). [e] MAE is defined as the sum of the absolute errors (Σ|δcalcd – δexpt|) divided by the total number (n) of chemical shifts values
included in the comparison; MAE=Σ|δcalcd – δexpt|/n.
Eur. J. Org. Chem. 0000, 0–0
www.eurjoc.org
2
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
gies and optimal geometries of the three low-energy conforma- involves the intermediacy of a carbanionic species which can
tional isomers 11a–c (rotation around the C1–C62 bond) were be effectively trapped by D2O. To further probe this mechanistic
first determined at a higher level of theory, using the two-layer hypothesis, we carried out the photochemical transformation
ONIOM [M06-2X/6-31+G(d,p):HF/3-21G] approach and implicit of 1 in the presence of 0.08 % D2O without adding C60. Not
solvation in CHCl3 (IEFPCM) (Figure 2; see also Section 7/Figure surprisingly, deuterium incorporation was again observed at the
S25 in SI). At this level of theory, 11b was found to be the same C-atom affording a mixture of 2-d0/2-d1 in a ratio of
lowest energy hence most highly populated conformer (Fig- ca. 70:30 (Figures S23–S24 in SI).
ure 2).
Scheme 2 depicts a plausible mechanistic scenario for the
photochemical reaction of 1 with C60, consistent with the ex-
perimental evidences described above and the preceding litera-
ture. The first step involves an intramolecular hydrogen atom
transfer (HAT) by the triplet excited state of 1. The resulting 1,4-
biradical 9 can either (i) decay to intermediate ketene-enol 7
and revert back to 1,[4] or (ii) react intermolecularly with 1 via
a HAT process to afford radical intermediates 5 and 12.[3] Acyl
radical 5 rearranges to radical intermediate 6, which is subse-
quently trapped by C60 to afford the final product 11 upon
coupling with a H-atom derived from 12. In another possible
route, 7 undergoes intramolecular ionic closure to afford 14
which, in turn, rearranges to 8 and subsequently to 2 via an
ionic hydrogen transfer.[2] In consistence with the mechanistic
experiments described above, the presence of deuterated water
in this latter step can result in the formation of deuterated
phthalide, 2-d1. Phthalide 2 can subsequently enter the path-
way of radical addition to C60 via a HAT process catalyzed by 5.
The significant increase of the reaction rate induced by the
addition of CH3CN as a polar co-solvent (Table 1, entries 5–6)
further supports the intermediacy of polar transient species, i.e.
8 and 14, in the reaction mechanism.
Figure 2. Optimized geometries, relative Gibbs free energies (ΔG in kcal mol–1
)
and calculated Boltzmann populations (%) calculated thereof, for the three
low-energy conformers of 11; geometry optimizations and energy calculations
were carried out at the ONIOM2[M06-2X/6-31+G(d,p):HF/3-21G] [IEFPCM-
chloroform] level of theory (298.15 K). The atoms treated at a high level in the
ONIOM approach are shown in ball and stick representation, while the lower
level atoms are represented in tube model. This figure also shows the number-
ing system adopted in the present work for identifying the relevant C- and
H-atoms. Color code: white = H; gray = C; red = O. Cartesian coordinates and
energies of all three conformers considered in our GIAO NMR calculations are
provided in SI (Tables S4–S6).
Then, the NMR shielding tensors (σ; ppm) for each conformer
were computed at the B3LYP/6-311+G(2d,p) level of theory us-
ing the gauge-including atomic orbital (GIAO) method in
chloroform (IEFPCM solvation model). The computed set of
isotropic magnetic shielding tensor values (σiso; ppm) for each
nucleus in all conformers were scaled, referenced, and averaged
(using the mole fraction of each conformation), to generate a
set of Boltzmann-weighted average chemical shifts (δcalcd; ppm)
(Table 2). The mean absolute errors (MAEs) between the experi-
mental and computed chemical shifts associated with each nu-
1
cleus type (i.e. H and 13C) of 11 are given in the last row of
Table 2. A perusal of this table indicates a good agreement
between the experimental and the computed 1H and 13C chem-
ical shifts (δexpt vs. δcalcd), which further supports the assigned
structure of 11. Full details on the computational methods, as
well as the optimized geometries and energies of all three con-
formers of 11 are provided in the Experimental Section and SI
(Tables S4–S6), respectively.
The reaction mechanism was next investigated by perform-
ing deuterium-labeling experiments. Initially, we performed the
photocatalyzed reaction of 1 with C60 in a 85:15 mixture of
C6H5Cl/CH3CN containing 0.08 % D2O. A significant percentage
of deuterium incorporation (ca. 30 %) at the C(62) carbon atom
of 11 was observed by 1H NMR spectroscopy (Figure S22 in SI).
Scheme 2. Proposed mechanism for the photochemical reaction of 1 with
fullerene C60
.
Next, in an attempt to provide solid experimental evidence
Apparently, this result suggests that the reaction mechanism for the suggested structure of 11, we sought to access it from
Eur. J. Org. Chem. 0000, 0–0
www.eurjoc.org
3
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
different synthetic pathways. We envisioned that structure 11
can result from the reaction of C60 with phthalide (2) in the
presence of tetrabutylammonium decatungstate [TBADT,
(nBu4N)4W10O32] through a Hydrogen Atom Transfer (HAT)
mechanism (Table 3, entry 1). TBADT is a well-established pho-
tocatalyst for C–H bond activation in several classes of organic
compounds.[11] For example, we have shown that TBADT can
activate the benzylic C–H bond in toluenes[12] and also the α-
C–H bond in ethers[13] to afford the corresponding C-centered
radicals that can be effectively trapped by C60 and give the
respective mono-functionalized hydrofullerenes in moderate to
good yields. We therefore envisaged that TBADT can activate
the benzylic C–H bond in 2 to afford the anticipated radical
addition adduct of C60, viz. 11. To our delight this was indeed
the case and product 11 was isolated in 20 % yield (see Experi-
mental Section). This result not only provides further evidence
in support of the structure suggested for the product obtained
from the reaction of C60 with 1 (Table 1), but also constitutes a
new synthetic example for the employment of TBADT catalysis
in fullerene chemistry.[9,12–14] The scope of this new TBADT-cata-
lyzed reaction was further explored by studying the reaction
of C60 with three commercially available phthalides, namely 5-
bromophthalide (15), 5-cyanophthalide (16), and 6-nitrophthal-
ide (17) (Table 3, entries 2–4). Indeed, following the previous
photocatalytic procedure, the corresponding radical intermedi-
ates were efficiently trapped by C60 to yield selectively the
monofunctionalized fullerene derivatives 18–20, respectively, in
decent isolated yields (and almost quantitative recovery of un-
reacted C60), thus validating the versatility and efficiency of this
method.
Scheme 3. Intermolecular primary KIE in the TBADT-mediated reaction of 2-
d0/2-d2 with C60
.
methylene –CH2– and –CD2– groups in 2-d0/2-d2, is propor-
tional to the primary isotope effect of kH/kD. 1H NMR of the
methine CH signal of 11-d0 as well as the phenyl hydrogens of
both 11-d0 and 11-d1/11-d2, determined the primary isotope
effect kH/kD = 1.7 0.1 (Scheme 3; see also Figure S21 in SI).
This substantial primary KIE indicates an extensive C–H(D) bond
breaking in the transition state of the first slow radical forming
step. Another important finding from this experiment concerns
the source of the fullerenyl hydrogen. The adduct 11-d1 is par-
tially deuterated (24 % D incorporation on the fullerene cage),
which indicates, in consistence to the mechanism proposed in
our previous studies,[9,12–14] that the intermediate C60 anion is
mainly trapped by a proton originated from the residual mois-
ture in the reaction mixture.
Conclusions
We disclosed a new mode of radical reactivity of C60, which
involves the photochemical addition of o-phthalaldehyde (1)
to C60. This simple and rapid reaction affords exclusively the
unanticipated C60-phthalide derivative 11 in good yield. The
formation of this product is consistent with the intervention of
a key free-radical intermediate in the course of the reaction,
which is further supported by our mechanistic studies. The ob-
served single addition of radicals to C60 is of particular interest,
given the well-established high affinity of C60 (a “radical
sponge”) with free radicals,[8] which generally results in multiple
radical additions, hence a difficult to separate mixture of multi-
adducts. Thus, the present reaction is an excellent example con-
cerning the isolation and characterization of C60 monoadduct
via a single radical addition, especially in good yield and high
purity. The structure of 11 has been determined by the com-
bined use of experimental and theoretical NMR studies.
In the course of this study, we further disclosed a new mode
of radical reactivity of C60 with a series of substituted phthalides
catalyzed by tetrabutylammonium decatungstate (TBADT). The
operating mechanism for this latter reaction was investigated
by determining the intermolecular kinetic isotopic effect of the
reaction of C60 with 2-d0/2-d2. The presence of a normal isotope
effect (kH/kD = 1.7 0.1) shows that the rate-determining step
of the reaction is the cleavage of the C–H or C–D bond as
similarly found in our previous reactions on the TBADT-cata-
lyzed reaction of C60 with different substrates.[9,11–14] This latter
reaction expands further the synthetic utility of TBADT catalyst
Table 3. TBADT-catalyzed reaction of C60 with phthalide 2 and 15–17.[a]
Entry
Substrate
Product
Irradiation time
Isolated yield [%]
1
2
3
4
2
11
18
19
20
1.5 h
3.5 h
4 h
20
15
17
14
15
16
17
1.5 h
[a] For experimental details see the Experimental Section.
To probe the mechanism of this reaction and obtain informa-
tion on the extent of bond breaking we measured the primary
intermolecular kinetic isotope effect (KIE) in the reaction of C60
with an equimolar mixture of phthalide-d0 (2-d0) and phthalide-
d2 (2-d2) (Scheme 3; see also Figures S20–S21 in SI). This reac-
tion was performed similarly with those described previously in
this work (Table 3), using a 40-fold excess of 2-d0/2-d2.
The ratio of products 11-d0 and 11-d1/11-d2, which is the in fullerene functionalization chemistry by providing a new
result of an intermolecular isotopic competition between the route to previously unexplored fullerene-based structures.
Eur. J. Org. Chem. 0000, 0–0
www.eurjoc.org
4
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
Photochemical Reaction of C60 with 1 at Ambient Light: In a
typical experiment, a solution of C60 (20 mg, 0.028 mmol) in 30 mL
of chlorobenzene/acetonitrile (85:15) was added in a 50-mL glass
flask followed by addition of 50 equiv. o-phthalaldehyde (190 mg,
1.4 mmol). The resulting mixture was septum capped and depos-
ited on the lab bench next to the window exposed to ambient,
diffused sunlight. The progress of the reaction was monitored by
HPLC. After 72 hours the solvent was distilled from the reaction
mixture under reduced pressure, and the remaining crude product
was centrifuged three times with acetonitrile HPLC grade to remove
excess o-phthalaldehyde. The isolated product was further purified
by flash column chromatography (eluent: hexane/toluene, 4:1 v/v)
to afford fullerene adduct 11 in 43 % isolated yield (10.2 mg;
0.012 mmol), C60 (2 mg; 0.0028 mmol), and a small amount of a
complex mixture of several unidentified by-products.
Experimental Section
General: Reagents and solvents were purchased in reagent grade
quality from Fluka, Merck, and Sigma-Aldrich and used without fur-
ther purification. Tetrabutylammonium decatungstate (TBADT) was
prepared according to procedure reported in the literature.[15] 1
H
NMR and 13C NMR spectra were recorded on a Bruker AMX-500 MHz
(125 MHz for 13C) spectrometer in CDCl3 solutions.
Chemical shifts (δ) are reported in ppm downfield from tetramethyl-
silane using the residual deuterated solvent signals as an internal
reference (CDCl3: δH = 7.26 ppm, δC = 77.16 ppm). For 1H NMR,
coupling constants J are given in Hz and the resonance multiplicity
is described as s (singlet), d (doublet), t (triplet), m (multiplet), and
br (broad). All spectra were recorded at 25 °C. UV/Vis spectra were
performed on a Shimadzu MultiSpec-1501 UV/Visible spectrometer.
HPLC analysis was conducted on a Marathon III instrument
equipped with a 5C18-MS (4.6 × 250 mm, Nacalai Tesque) reversed
phase column with detection at 310 nm. A mixture of toluene/
acetonitrile (1:1 v/v) was used as eluent at 1 mL/min flow rate. Flash
column chromatography was carried out on SiO2 (silica gel 60, SDS,
230–400 mesh ASTM) by applying an overpressure of 0.1–0.6 bar.
Evaporation of the solvents was accomplished with a rotary evapo-
rator or by high vacuum distillation. Photochemical reactions were
carried out by using a 300-W xenon lamp as the light source. Infra-
red (IR) spectra were recorded on a Perkin-Elmer 1600 FT-IR spec-
trometer (ATR, Golden Gate) and values are reported as wavenum-
bers (cm–1) with band intensities indicated as: s (strong), m (me-
dium), and w (weak).
Photochemical Isomerization of 1 to 2: A solution of o-phthal-
aldehyde 1 (200 mg, 1.49 mmol) in 70 mL of solvent (C6H5Cl, or
CHCl3, or C6H6, or chlorobenzene/acetonitrile, 85:15) was added in
a 100-mL glass flask. The resulting solution was septum capped,
purged for 20 min with argon, and was subsequently irradiated
with a 300 W Xenon lamp at 5–10 °C using an ice water bath. A
constant, slow argon stream was maintained over the solution. The
progress of all reactions was monitored by NMR. The irradiation
time was varied from 25 min (CHCl3) to 40 min (C6H5Cl/CH3CN,
C6H5Cl, C6H6). Then, the solvent was distilled from the reaction mix-
ture under reduced pressure to afford substantially pure 2 as a
white solid. The identity of 2 was confirmed by comparison with a
commercial sample.
Phthalide (2): Rf = 0.3 (SiO2; hexane/EtOAc, 2:1); 1H NMR (CDCl3,
500 MHz, 25 °C): δH = 5.33 (2H, s), 7.49–7.56 (2H, m), 7.67–7.69 (1H,
Experimental Procedures and Spectroscopic Data
3
3
Photochemical Reaction of C60 with 1: In a typical experiment, a
solution of C60 (20 mg, 0.028 mmol) in 65 mL of solvent (C6H5Cl)
was added in a 100-mL glass flask followed by addition of 50 equiv.
o-phthalaldehyde (188 mg, 1.4 mmol). The resulting solution was
septum capped and deoxygenated in the absence of light by purg-
ing with argon for 20 min. This solution was subsequently irradiated
for 6 min with a 300 W Xenon lamp at 5–10 °C using an ice water
bath. A constant, slow argon stream was maintained over the solu-
tion. The progress of all reactions was monitored by HPLC. The sol-
vent was distilled from the reaction mixture under reduced pres-
sure, and the remaining crude product was centrifuged three times
with acetonitrile HPLC grade to remove excess o-phthalaldehyde.
The isolated product was further purified by flash column chroma-
tography (eluent: hexane/toluene, 4:1 v/v) to afford fullerene adduct
11 in 19 % yield (4.5 mg, 0.0053 mmol).
m), 7.94 (1H, d, JHH = 8 Hz), 4.88 (1H, m), 5.43 (1H, dd, JHH 8.4,
3
15.0 Hz), 5.79 (1H, d, 3JHH 11.0 Hz), 6.45 (1H, dd, JHH 11.0, 15.0 Hz),
7.46–7.24 (10H, m) ppm; 13C NMR (CDCl3, 125 MHz, 25 °C): δC
=
69.79, 122.22, 125.93, 125.97, 129.20, 134.15, 146.66, 171.24 ppm.
General Photocatalytic Procedure for the TBADT-Mediated Pho-
tochemical Reaction of C60 with Phthalides (2, 15, 16, or 17): A
solution of C60 (20 mg, 0.028 mmol) in 65 mL of chlorobenzene/
acetonitrile (85:15) was added in a 100-mL glass flask followed by
addition of 100 equiv. phthalide (2, 15, 16, or 17) (2.8 mmol) and
0.2 equiv. TBADT (19 mg, 0.0055 mmol). The resulting mixture was
septum capped and degassed in the absence of light by purging
with argon for 20 min. This solution was subsequently irradiated
with a 300 W Xenon lamp for 1.5–4 h at 5–10 °C in an ice water
bath, while a constant, slow argon stream was maintained over the
solution. The progress of the reaction was monitored by HPLC.
Then, the solvent was distilled from the reaction mixture under
reduced pressure, and the remaining crude product was centri-
fuged three times with acetonitrile HPLC grade to remove excess
phthalide (2, 15, 16, or 17) and TBADT. The crude product was
further purified by flash column chromatography (eluent: hexane/
toluene, 4:1 v/v).
11: IR (KBr): ν = 1775 (s), 1720 (w), 1702 (w), 1688 (w), 1655 (m),
˜
1638 (w), 1618 (w), 1561 (m), 1543.3 (w), 1499 (w), 1509 (w), 1459
(m), 1439 (w), 1261 (s), 1015 (s), 800 (s), 527 (s) cm–1; UV/Vis (CHCl3):
λmax (nm) = 326, 405, 433; 1H NMR (CDCl3, 500 MHz, 25 °C): δH
=
8.26 (2H, m), 7.95 (1H, m), 7.83 (1H, m), 7.16 (1H, s), 6.31 (1H, s)
ppm; 13C NMR (CDCl3, 125 MHz, 25 °C): δC = 169.80, 152.97, 152.57,
151.05, 149.18, 147.80, 147.52, 146.95, 146.92, 146.83, 146.79,
146.77, 146.65, 146.61, 146.52, 146.50, 146.48, 146.44, 146.42,
146.11, 146.06, 146.01, 145.83, 145.77, 145.66, 145.59, 145.58,
145.00, 144.97, 144.66, 144.47, 143.47, 143.40, 143.25, 142.89,
142.85, 142.80, 142.53, 142.46, 142.35, 142.31, 142.22, 142.16,
141.93, 141.90, 141.81, 141.76, 141.74, 141.55, 140.63, 140.62,
140.16, 140.11, 138.51, 137.85, 136.41, 136.06, 135.02, 131.02,
128.36, 126.85, 124.70, 87.18, 68.33, 55.75 ppm; HR-MALDI-MS
(DCTB): m/z (%): 856.0424 (18), 855.0393 (48), 854.0361 (67, M+,
calcd. for C68H6O2+: 854.0362), 721.0073 (100, C60H+, calcd. for
C60H+: 721.0073).
TBADT-Mediated Photochemical Reaction of C60 with Phthalide
2: The reaction was carried out following the general photocatalytic
procedure described above using phthalide 2 (375 mg, 2.8 mmol).
The reaction mixture was irradiated for 90 min. The crude product
was purified by flash column chromatography (eluent: hexane/tolu-
ene, 4:1 v/v) to afford fullerene adduct 11 in 20 % isolated yield
(4.7 mg; 0.0055 mmol), along with a small amount of a complex
mixture of several unidentified by-products.
TBADT-Mediated Photochemical Reaction of C60 with 5-Bromo-
phthalide (15): The reaction was carried out following the general
Eur. J. Org. Chem. 0000, 0–0
www.eurjoc.org
5
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
photocatalytic procedure described above using 5-bromophthalide
(15) (600 mg, 2.8 mmol). The reaction mixture was irradiated for 3.5
h. The crude product was purified by flash column chromatography
(eluent: hexane/toluene, 4:1 v/v) to afford fullerene adduct 18 in
15 % isolated yield (4.0 mg; 0.0043 mmol).
raphy (eluent: hexane/toluene, 4:1 v/v) to afford fullerene adduct
20 in 14 % isolated yield (3.5 mg; 0.0039 mmol).
20: IR (KBr): ν = 2375.1 (w), 2346.1 (m), 1832.0 (w), 1782.5 (s), 1735.7
˜
(w), 1719.8 (m), 1702.5 (w), 1686 (m), 1676.5 (w), 1664.2 (w), 1655.2
(m), 1648.0 (w), 1637.9 (m), 1617.9 (m), 1578.0 (w), 1570.9 (w),
1560.8 (m), 1532.9 (s), 1508.8 (w), 1499.5 (w), 1491.4 (w), 1475.3 (w),
1459.3 (m), 1438.7 (w), 1427.7 (w), 1342.5 (s), 1298.8 (w), 1252.1 (w),
1209.6 (w), 1183.3 (w), 1092.5 (s), 1042.0 (w), 998.1 (m), 837.9 (w),
811.3 (w), 794.8 (w), 752.5 (s), 705.0 (m), 669.5 (w), 577.3 (m), 554.0
18: IR (KBr): ν = 2372.9 (m), 2346 (s), 1871 (w), 1846.2 (w), 1832 (w),
˜
1774.6 (s), 1735.7 (m), 1719.6 (m), 1702.4 (m), 1686 (w), 1676.5 (m),
1671.1 (w), 1664.1 (w), 1655.2 (s), 1648.1 (m), 1637.9 (m), 1628.9 (w),
1618.3 (m), 1603.7 (w), 1578.1 (m), 1570.9 (m), 1560.7 (s), 1543.1 (s),
1523.5 (m), 1508.7 (s), 1499.3 (m), 1491.4 (w), 1475.5 (m), 1465.8
(w), 1459.2 (s), 1449.9 (w), 1438.8 (m), 1421.2 (m), 1388.9 (w), 1325
(w) 1294.1 (w), 1261.7 (w), 1211.1 (m), 1182.7 (w), 1058.3 (w), 999.3
(m), 845.9 (w), 753.4 (s), 708.6 (w), 670 (w), 661.2 (w), 573.9 (w),
526.9 (s), 420 (w) cm–1; UV/Vis (CHCl3): λmax (nm) = 329, 405, 431;
1H NMR (CDCl3, 500 MHz, 25 °C): δH = 8.37 (1H, br. s), 8.10 (1H, d,
3JHH = 8.0 Hz), 7.95 (1H, dd; 3JHH(1)= 8.0 Hz, 3JHH(2)= 2.0 Hz), 7.11
(w). 527.0 (s), 475.7 (w) cm–1; UV/Vis (CHCl3): λmax (nm) = 329, 405,
3
431; 1H NMR (CDCl3, 500 MHz, 25 °C): δH = 9.05 (1H, d, JHH
=
3
3
2.0 Hz), 8.77 (1H, dd, JHH(1) = 8.5 Hz; JHH(2) = 2.0 Hz), 8.41 (1H, d,
3JHH = 8.5 Hz), 7.28 (1H, s), 6.46 (1H, s) ppm; 13C NMR (CDCl3,
125 MHz, 25 °C): δC = 167.31, 159.80, 152.43, 151.87, 150.94, 150.18,
149.28, 147.89, 147.87, 147.52, 146.89, 146.85, 146.73, 146.68,
146.66, 146.64, 146.57, 146.56, 146.52, 146.50, 146.24, 146.20,
146.19, 146.09, 145.89, 145.87, 145.86, 145.74, 145.70, 145.66,
145.61, 145.04, 145.01, 144.55, 144.39, 143.50, 143.44, 143.24,
142.98, 142.98, 142.90, 142.85, 142.42, 142.38, 142.17, 142.13,
141.97, 141.92, 141.89, 141.80, 141.57, 141.47, 140.82, 140.77,
140.15, 140.14, 138.73, 138.07, 136.13, 135.89, 130.16, 129.53,
(1H, s), 6.34 (1H, s) ppm; 13C NMR (CDCl3, 125 MHz, 25 °C): δC
=
168.82, 159.90, 152.70, 152.24, 151.16, 148.72, 148.62, 148.56,
148.15, 147.88, 147.81, 147.48, 146.86, 146.81, 146.65, 146.62,
146.54, 146.49, 146.46, 146.20, 146.15, 146.13, 145.95, 145.83,
145.81, 145.65, 145.59, 145.57, 145.44, 145.40, 144.98, 144.95,
144.74, 144.62, 144.44, 143.47, 143.24, 142.87, 142.50, 142.34,
142.21, 142.16, 141.94, 141.93, 141.56, 140.68, 140.14, 138.54,
137.42, 135.94, 134.67, 130.38, 128.01, 127.24, 86.52, 68.11, 55.90.
ppm; ESI-MS (negative ion mode): m/z 930.926 ([M – H]–, calcd. for
126.17, 122.26, 111.38, 101.48, 87.28, 68.00, 56.40 ppm; ESI-MS (neg-
ative ion mode): m/z 898.016 ([M – H]–, calcd. for C68H4NO4
:
–
898.013).
TBADT-Mediated Photochemical Reaction of C60 with Equimolar
Quantities of Phthalide-d0 (2-d0) and Phthalide-d2 (2-d2): The
reaction was carried out following the general photocatalytic proce-
dure described above using 2-d0/2-d2 (150 mg, 1.1 mmol). The reac-
tion mixture was irradiated for 13 h. The crude product was purified
by flash column chromatography (eluent: hexane/toluene, 4:1 v/v)
to afford fullerene adduct 11-d0 and 11-d1/d2 in 18 % isolated yield
(4.2 mg; 0.0050 mmol).
C
68H4BrO2–: 930.939).
TBADT-Mediated Photochemical Reaction of C60 with 5-Cyano-
phthalide (16): The reaction was carried out following the general
photocatalytic procedure described above using 5-cyanophthalide
16 (445 mg, 2.8 mmol). The reaction mixture was irradiated for 4 h.
The crude product was purified by flash column chromatography
(eluent: hexane/toluene, 4:1 v/v) to afford fullerene adduct 19 in
17 % isolated yield (4.3 mg; 0.0049 mmol), along with a small
amount of a complex mixture of several unidentified by-products.
Computational Methods: All calculations were performed using
the GAUSSIAN 09 package (Revision A.02).[10] The energies and opti-
mal geometries of the three conformers 11a–c were determined at
the B3LYP/6-31G(d) level of theory, in the gas phase (Tables S1–S3
in SI). All optimized structures were verified as ground-state minima
by performing frequency calculations at the same level of theory
(no imaginary frequencies were found). For NMR calculations, the
energies and optimal geometries of 11a–c were first determined at
a higher level of theory, using the two-layer ONIOM [M06-2X/6-
31+G(d,p):HF/3-21G] approach and implicit solvation in CHCl3
(IEFPCM; integral equation formalism variant of the polarizable con-
tinuum model), as implemented in Gaussian 09 software package
(Revision A.02)[10] (Tables S4–S6 in SI). Solute cavities were con-
structed using default united-atom radii (UA0). A pruned (99,590)
integration grid (99 radial shells and 590 angular points per shell;
“ultrafine” grid in Gaussian 09) was employed. All optimized struc-
tures were verified as ground-state minima by performing fre-
quency calculations at the same level of theory (no imaginary fre-
quencies were found), the data from which were also used to com-
pute their Boltzmann distribution at 298 K. All optimized structures
were then used for NMR calculations. NMR shielding tensors (σ, in
ppm) were computed with the gauge-independent atomic orbitals
(GIAO) method at the B3LYP/6-311+G(2d,p) level of theory, includ-
ing chloroform solvation effects (using the IEFPCM solvation
model), while employing solute cavities built from Bondi radii as
implemented in Gaussian 09. A pruned (99,590) integration grid
was employed. The computed isotropic magnetic shielding tensors
(σiso, in ppm) for each nucleus in all conformers was converted to
referenced and empirically scaled chemical shifts (δscaled, in ppm)
by applying scaling and referencing factors (slope and intercept,
19: IR (KBr): ν = 2370.8 (m), 2345.8 (s), 1871 (w), 1846.2 (w), 1832
˜
(w), 1774.7 (s), 1735.6 (m), 1719.5 (s), 1702.4 (m), 1686 (s), 1676.5
(m), 1671 (w), 1664.2 (w), 1655.2 (s), 1648.1 (m), 1637.9 (s), 1628.8
(w), 1618.2 (s), 1577.9 (m), 1570.9 (w), 1560.7 (s), 1543.1 (s). 1523.5
(m), 1508.7 (s), 1499.2 (m), 1491.3 (w), 1475.5 (m) 1466 (w), 1459
(m), 1438.7 (m), 1421.3 (m), 1406.7 (w), 1389 (w), 1301.5 (w), 1272.2
(w), 1200.9 (w), 1043 (w), 1002.2 (w), 849.3 (w), 752.5 (s), 708.8 (w),
670.3 (w), 658.1 (w), 543.2 (w), 526.8 (s), 453 (w) cm–1; UV/Vis
(CHCl3): λmax (nm) = 328, 405, 431; 1H NMR (CDCl3, 500 MHz, 25 °C):
3
3
δH = 8.48 (1H, s), 8.36 (1H, d, JHH = 8 Hz), 8.09 (1H, d, JHH = 8 Hz)
7.21 (1H, s), 6.41 (1H, s) ppm; 13C NMR (CDCl3, 125 MHz, 25 °C): δC =
167.74, 152.39, 151.88, 147.89, 147.87, 147.56, 146.90, 146.87,
146.74, 146.68, 146.66, 146.59, 146.52, 146.51, 146.26, 146.18,
146.18, 146.16, 145.96, 145.90, 145.88, 145.78, 145.70, 145.67,
145.65, 145.03, 145.02, 144.59, 144.42, 143.51, 143.45, 143.26,
142.99, 142.93, 142.91, 142.86, 142.50, 142.46, 142.39, 142.21,
142.20, 142.16, 142.01, 141.98, 141.97, 141.92, 141.82, 141.59,
141.56, 140.80, 140.77, 140.29, 140.17, 138.69, 138.03, 136.17,
135.89, 134.70, 131.92, 128.54, 127.71, 118.41, 117.52, 87.05, 67.98,
56.27 ppm; HR-MALDI-MS (DCTB): m/z (%): 879.089 (100, M+, calcd.
for C69H6NO2+: 879.031), 721.116 (80, C60H+, calcd. for C60H+:
721.007).
TBADT-Mediated Photochemical Reaction of C60 with 6-Nitro-
phthalide (17): The reaction was carried out following the general
photocatalytic procedure described above using 6-nitrophthalide
(17) (500 mg, 2.8 mmol). The reaction mixture was irradiated for
90 min. The crude product was purified by flash column chromatog-
Eur. J. Org. Chem. 0000, 0–0
www.eurjoc.org
6
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
respectively) according to the equation: δscaled = (σiso – intercept)/
slope, where δscaled is the referenced/scaled chemical shift (in ppm),
and σiso is the isotropic computed NMR shielding tensor (in ppm).
The slope and intercept values are typically obtained by linear re-
gression analysis of a plot of the calculated isotropic magnetic
shielding tensors (σiso) against the corresponding experimental
chemical shifts (δexpt). In the present study, we used generic-scaling
factors obtained from large datasets [i.e. obtained from linear re-
gression analysis of a plot of calculated isotropic magnetic shielding
tensors (σiso) against experimental chemical shifts (δexpt) of a large
series of compounds]; these scaling/referencing factors are specific
of the level of theory used. For the level of theory used herein
[B3LYP/6-311+G(2d,p)//M06-2X/6-31+G(d,p)] the following values
have been calculated by Willoughby et al.:[16] for 1H: intercept =
31.9477, slope = –1.0767; for 13C: intercept = 181.2412, slope =
–1.0522. The final calculated chemical shift values (δcalcd, in ppm;
Table 2) were determined by Boltzmann averaging the scaled chem-
ical shift values (δscaled, in ppm), based on the calculated Gibbs
free energy values of each conformer. Specifically, by using the free
energy data (kcal mol–1) obtained from the frequency calculations,
a Boltzmann weighting factor [e(–E/RT), where E is the relative energy
(kcal mol–1) with respect to the most stable conformer, T is the tem-
perature (in K), and R is the gas constant (0.001986 kcal mol–1 K–1)]
was determined for each conformer at 298 K, which was, in turn,
converted into the relative mole fraction by dividing the calculated
Boltzmann factor of each conformer by the sum of all the Boltz-
mann factors of all contributing conformers. The resulting weight-
ing factors (mole fraction contributions) were applied to the scaled
chemical shift values (δscaled, in ppm) for each nucleus of each indi-
vidual conformer. Summation of the weighted chemical shifts
across all conformers generates the final Boltzmann-weighted aver-
age chemical shifts (δcalcd, in ppm) used to compare against experi-
mental data (δexpt, in ppm; Table 2). It should be noted that only
one enantiomer [i.e. the one having (S)-configuration at C(62)] was
considered in our computational NMR studies.
lowship grant (GA. no. 4813). We are also grateful to Prof. T.
Drewello and Dr. Ina D. Kellner (University of Erlangen-Nürn-
berg) as well as Prof S. Pergantis (University of Crete) for per-
forming the MS analyses.
Keywords: Fullerene C60 · Photochemistry · Radical
reactions · o–Phthalaldehyde · Phthalides
[1] J. Kagan, Tetrahedron Lett. 1966, 7, 6097–6102.
[2] S. P. Pappas, J. E. Blackwell Jr., Tetrahedron Lett. 1968, 9, 3337–3340.
[3] D. A. Harrison, R. N. Schwartz, J. Kagan, J. Am. Chem. Soc. 1970, 92, 5793–
5795.
[4] J. C. Scaiano, M. V. Encinas, M. V. George, J. Chem. Soc., Perkin Trans. 2
1980, 724–730.
[5] J. Gebicki, S. Kuberski, J. Chem. Soc., Chem. Commun. 1988, 1364–1365.
[6] J. Gebicki, S. Kuberski, R. Kaminski, J. Chem. Soc., Perkin Trans. 2 1990,
765–769.
[7] J. C. Netto-Ferreira, J. C. Scaiano, Can. J. Chem. 1993, 71, 1209–1215.
[8] M. D. Tzirakis, M. Orfanopoulos, Chem. Rev. 2013, 113, 5262–5321.
[9] M. D. Tzirakis, M. Orfanopoulos, J. Am. Chem. Soc. 2009, 131, 4063–4069.
[10] M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R.
Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G. A. Petersson, H. Na-
katsuji, M. Caricato, X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G.
Zheng, J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hase-
gawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, J. A.
Montgomery Jr., J. E. Peralta, F. Ogliaro, M. Bearpark, J. J. Heyd, E. Broth-
ers, K. N. Kudin, V. N. Staroverov, R. Kobayashi, J. Normand, K. Raghava-
chari, A. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi, N. Rega,
J. M. Millam, M. Klene, J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, J.
Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi,
C. Pomelli, J. W. Ochterski, R. L. Martin, K. Morokuma, V. G. Zakrzewski,
G. A. Voth, P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels, Ö.
Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski, D. J. Fox, Gaussian 09,
Revision A.1, Gaussian, Inc., Wallingford CT, 2009.
[11] M. D. Tzirakis, I. N. Lykakis, M. Orfanopoulos, Chem. Soc. Rev. 2009, 38,
2609–2621.
[12] M. D. Tzirakis, M. Orfanopoulos, Org. Lett. 2008, 10, 873–876.
[13] M. D. Tzirakis, M. Orfanopoulos, Angew. Chem. Int. Ed. 2010, 49, 5891–
5893; Angew. Chem. 2010, 122, 6027.
Supporting Information (see footnote on the first page of this
article): Copies of 1H NMR, 13C NMR, FT-IR, and UV/Vis spectra of 11
and 18–20; optimized geometries (in Cartesian coordinates), total
energies, relative free energies, and Boltzmann populations of the
three conformers of 11. Experimental Details.
[14] M. D. Tzirakis, M. N. Alberti, M. Orfanopoulos, Chem. Commun. 2010, 46,
8228–8230.
[15] C. Cui, Y. Li, Y. Li, Adv. Energy Mater. 2016, 1601251; D. C. Duncan, T. L.
Netzel, C. L. Hill, Inorg. Chem. 1995, 34, 4640–4646.
[16] P. H. Willoughby, M. J. Jansma, T. R. Hoye, Nat. Protocols 2014, 9, 643–
660.
Acknowledgments
The research work was supported by the Hellenic Foundation
for Research and Innovation (HFRI) and the General Secretariat
for Research and Technology (GSRT), under the HFRI PhD Fel-
Received: September 14, 2019
Eur. J. Org. Chem. 0000, 0–0
www.eurjoc.org
7
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
Photochemistry of Fullerenes
M. D. Tzirakis,* N. G. Malliaros,
M. Orfanopoulos* .............................. 1–8
Photochemical Reaction of o-
Phthalaldehyde with Fullerene C60:
The Stimulus for the Phthalide Addi-
tions to C60
C60 reacts under ambient light with fullerene can be accessed via a novel
o-phthalaldehyde to afford an un- tetrabutylammonium decatungstate
anticipated C60-phthalide derivative in photocatalyzed reaction of C60 with
43 % yield. Alternatively, this hydro- phthalides (4 examples).
DOI: 10.1002/ejoc.201901363
Eur. J. Org. Chem. 0000, 0–0
www.eurjoc.org
8
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim