Crystal structures and photoreactivity of transand cis-valerophenone diperoxides

Article abstract of DOI:10.1007/s11224-011-9887-8

The crystal structures of valerophenone diperoxides trans-1 and cis-1 were elucidated by X-ray crystallographic analysis. The 1,2,4,5-tetraoxane rings of both compounds adopt chair conformations. Intermolecular CH...O hydrogen bond, π-π, and CH/π interactions exist in cis-1, whereas only CH/π interactions exist in trans-1. In the asymmetric unit of the crystal, a half molecule exists for trans-1, while one molecule for cis-1 which shows wholemolecule disorder. Solid-state photolysis (at 254 nm) or solution-state thermolysis (at 150 °C) of trans-1 and cis^-1 produced valerophenone (2) and butyl benzoate (3). Rationalization of the solid-state photoreactivity of the diperoxides by their crystal structures was attempted. Springer Science+Business Media, LLC 2011.

Full text of DOI:10.1007/s11224-011-9887-8

Struct Chem (2012) 23:441–449
DOI 10.1007/s11224-011-9887-8
ORIGINAL RESEARCH
Crystal structures and photoreactivity of trans-
and cis-valerophenone diperoxides
•
Hiroki Takahashi Yoshikatsu Ito
Received: 23 May 2011 / Accepted: 20 September 2011 / Published online: 9 October 2011
Ó Springer Science+Business Media, LLC 2011
Abstract The crystal structures of valerophenone diper-
oxides trans-1 and cis-1 were elucidated by X-ray crystal-
lographic analysis. The 1,2,4,5-tetraoxane rings of both
compounds adopt chair conformations. Intermolecular
CHÁÁÁO hydrogen bond, p–p, and CH/p interactions exist in
cis-1, whereas only CH/p interactions exist in trans-1. In the
asymmetric unit of the crystal, a half molecule exists for
trans-1, while one molecule for cis-1 which shows whole-
molecule disorder. Solid-state photolysis (at 254 nm) or
solution-state thermolysis (at 150 °C) of trans-1 and cis-1
produced valerophenone (2) and butyl benzoate (3).
Rationalization of the solid-state photoreactivity of the
diperoxides by their crystal structures was attempted.
reactions, avariety of investigationsto manipulatethe crystal
structuresthroughsupramoleculeformationwereconducted:
for instance, organic salts by Scheffer [1–3], host–guest
crystals by Toda [4, 5], photodimerization in templates by
MacGillivray [6–8], and designed topochemical polymeri-
zation by Lauher [9]. One of the present authors also con-
tributed several papers for this purpose (vide infra).
Recently, some 1,2,4,5-tetroxane derivatives containing
a diperoxide structure were found to exhibit remarkable
antimalarial activities in vitro and in vivo [10–12]. This
activity is probably associated with the reactivity of the
peroxide group and the steric and electronic effects of the
substituents surrounding the diperoxide group. For this
reason, new synthetic methods, structure and reactivity of
cyclic diperoxides have been investigated [13, 14].
Keywords 1,2,4,5-Tetraoxane Á X-ray analysis Á
Intermolecular interaction Á Photolysis Á Thermolysis
Preparation of trans-valerophenone diperoxide (trans-1)
and cis-valerophenone diperoxide (cis-1) (Scheme 1) and
their thermal and photosensitized decomposition in solution
were previously reported [15–18]. In continuation of these
studies, it has been found that the thermal decomposition of
cis-1 is faster than that of trans-1 in solution, whereas their
relative rates are reversed under direct photolysis in the
crystalline state. In order to interpret this contrary relative
stability of trans-1 and cis-1, we now elucidated their
crystal structures. The results are reported below.
Introduction
Studies on the crystal structure–reactivity correlation are
fundamentally important, because the reactivity and the
product stereochemistry are predictable from the crystal
structure [1–3]. Aiming at achieving desired solid-state
H. Takahashi (&)
Department of Interdisciplinary Environment, Graduate School
of Human and Environmental Studies, Kyoto University,
Kyoto 606-8501, Japan
Experimental
Synthesis
e-mail: takahashi.hiroki.2x@kyoto-u.ac.jp
Y. Ito
Trans- and cis-valerophenone diperoxides (trans-1 and cis-
1) were previously prepared [15–17]. These samples were
purified by preparative TLC (silica gel, CCl4), followed by
recrystallization from n-pentane.
Department of Synthetic Chemistry and Biological Chemistry,
Graduate School of Engineering, Kyoto University, Katsura,
Kyoto 615-8510, Japan
e-mail: yito@sbchem.kyoto-u.ac.jp
123

442
Struct Chem (2012) 23:441–449
v goniometer (graphite-monochromated Mo Ka radiation
˚
(k = 0.71073 A); x scans) with a Rigaku low-temperature
equipment. The data were collected at -173 °C. The crystal
structures were solved by direct methods (SIR97 [19]) and
refined by full-matrix least-squares methods (SHELXL-97
[20]). All the non-hydrogen atoms were refined anisotropi-
cally, whereas the hydrogen atoms were refined using a
riding model. With respect to crystal of cis-1, whole mole-
cule was disordered, but constraints or restraints were not
applied for any atom in the structure. All the calculations
were performed using crystallographic software packages
CrystalStructure [21] and Yadokari-XG 2009 [22]. Crystal
data and details of the structure determination and refinement
can be found in Table 1. The programs Mercury [23] and
PLATON [24] were used for analysis of intermolecular
interactions and for making graphics.
Scheme 1 Chemical structure of valerophenone diperoxides trans-1,
cis-1, and the products 2 and 3
X-ray crystallography of trans-1 and cis-1
Suitable crystals of trans-1 and cis-1 to be used for X-ray
analysis were obtained by recrystallization from n-pentane.
The X-ray experiments were performed on a Rigaku Sat-
urn724? CCD area detector diffractometer mounted on a 1/4
Solid-state photolysis of trans-1 and cis-1
Irradiation was carried out with a 10-W low-pressure
mercury lamp at 254 nm. HPLC analyses were performed
Table 1 Crystal and experimental data
Compound
Trans-1
Cis-1
CCDC deposit no.
Color/shape
CCDC-787284
CCDC-787283
Colorless/prism
C22H28O4
Colorless/needle
Empirical formula
Formula weight
C22H28O4
356.44
356.44
Temperature, °C
Crystal system, space group
-173(1)
-173(1)
ꢀ
Trigonal, R3
Monoclinic, P21/n
Unit cell dimensions
˚
a (A)
30.426(3)
30.426(3)
5.685(3)
90
5.7850(19)
21.258(7)
15.973(5)
90
˚
b (A)
˚
c (A)
a (°)
b (°)
c (°)
90
94.802(6)
90
120
3
˚
V (A )
4557(6)
9
1957.5(11)
4
Z
Density (calculated), g/cm3
Absorption coefficient, mm-1
F(000)
1.169
0.079
1728
1.209
0.079
768
h range for collection (°)
Reflections measured
Independent reflections
3.54–27.55
12,597
2326
3.15–27.46
31,767
4466
Index ranges h, k,
l
-39 B h B 38, -39 B k B 39, -7 B l B 7
-7 B h B 7, -27 B k B 27, -20 B l B 20
Data/restraints/parameters
Goodness of fit on F2
2326/0/174
1.110
4466/0/473
1.120
Final R indicates [I [ 2r(I)]
Final Rw indicates [I [ 2r(I)]
0.0389
0.0457
0.1132
0.0925
123

Struct Chem (2012) 23:441–449
443
Fig. 1 ORTEP drawings of
trans-1: a crystal and
(a)
(b)
b molecular structure with
labeled atoms. Atoms with the
suffix prime are generated by
the symmetry operation (1 - x,
-y, 1 - z). Thermal
a
C(9)
C(8)
C(7)
C(10)
C(11)
C(6)
C(2’)
O(1’)
C(4’)
O(2’)
displacement ellipsoids are
drawn at 50% probability level.
The H atoms are represented by
spheres, and the intermolecular
interactions (CHÁÁÁp) are
C(3)
C(4)
C(5)
C(1)
C(1’)
C(5’)
C(3’)
O(2)
O(1)
C(2)
C(6’)
C(11’)
C(10’)
C(7’)
C(8’)
indicated by green broken lines
C(9’)
b
o
with a JASCO PU-980 pump and a SHIMADZU SPD-6AV
detector by using a Cosmosil 5C18-AR column (4.6 mm
i.d. 9 150 mm) and were eluted (1 mL/min) with a mix-
ture of MeOH–H2O (90:10 or 65:35 v/v).
Results and discussion
ꢀ
Trans-1 crystallizes in the trigonal space group R3: One-
half molecule exists in the asymmetric unit of trans-1.
The 1,2,4,5-tetraoxane ring adopts a chair conformation
Crystalline samples of trans-1 and cis-1 were crushed in an
agate mortar and pestle with a similar strength, respectively.
Then ca. 10 mg of each sample was placed in a special quartz
vessel [25, 26] and was irradiated for 0.5 h under the stream of
Ar gas at 0 °C. After the irradiation, the sample was dissolved
in MeCN, followed by the addition of benzophenone as the
standard to determine the amounts of 2 and 3, and then the
solution was analyzed with HPLC. Next, dibenzyl was added
to the solution as the standard to determine the amount of
trans-1 or cis-1, and the solution was analyzed with HPLC.
In the case of the time-course experiments, ca. 10 mg of
each sample was similarly irradiated. A tiny part of the
sample was taken out after an appropriate irradiation time,
anditwasdissolvedinMeCN. Then, thesolutionwasdirectly
analyzed with HPLC. The material balance of 1, 2, and 3 was
assumed to be quantitative (total 100%). The decomposition
of the diperoxides 1 stopped at low conversions and the
HPLC retention times of 1differed very much from those of 2
and 3 (retention times with 90/10 MeOH–H2O: trans-1 9.4,
cis-18.2, 22.6, 33.0 min;retention timeswith65/35MeOH–
H2O: trans-1 [ 60, cis-1 [ 60, 2 10.8, 3 18.9 min). Fur-
thermore, irradiation of all parts of the solid sample equally
was difficult with our set-up [25, 26], although the powder
was mixed with a spatula from time to time. Therefore,
determinations of the diperoxide recovery were subjected to
large experimental errors as seen from Fig. 4.
˚
(puckering parameter: Q = 0.634(1) A, h = 179.17(1)°
and / = 0.00°) [27], and the Ph and Bu groups are axial
and equatorial, respectively (Fig. 1b). CH/p interactions
are observed between the Bu and the center of gravity of Ph
i
i
˚
groups (Cg) [C2–H5ÁÁÁCg 3.19(2) A and C4–H9ÁÁÁCg
˚
3.45(2) A, (symmetry code: (i) 1/3 ? y, 2/3 - x ? y,
ii
˚
2/3 - z) and C9–H12ÁÁÁCg 2.82(2)A, (symmetry code: (ii)
4/3 - x ? y, 2/3 - x, z - 1/3)] (Fig. 1a).
Cis-1 crystallizes in the monoclinic space group
P21/n. One molecule exists in the asymmetric unit of the
crystal. The molecule shows whole-molecule disorder and
has two conformational isomers (A and B) with occupancy
factors of 0.55(2) and 0.45(2), respectively, as shown in
Fig. 2b–e. The conformers A and B are virtually the same
(Bu and Ph are axial and equatorial, respectively, in one
half of the molecule, whereas they are equatorial and axial,
respectively, in the other half), except that they have
slightly different bond lengths and angles (Fig. 2e) so that
˚
the puckering parameters are Q = 0.646(2) A, h = 1.3(2)°,
and / = 83(6)° for conformer A; Q = 0.649(2) A,
h = 0.0(3)°, and / = 65(7)° for conformer B. Selected
bond lengths and angles for cis-1 are listed in Table 3. The
average lengths of O–O and C–O in reported cyclic di-
˚
1
˚
peroxide derivatives are 1.472 and 1.436 A, respectively.
1
Referring to the Cambridge Crystallographic Database (CSD Ver.
Thermolysis of trans-1 and cis-1
5.32), 21 crystals containing 1,2,4,5-tetraoxane have been reported.
Their refcodes are as follows: AROZEM, BMTTRX, CEDDOE,
DPTOCH, GAGTEO, GAGTIS, GAHXIW, IBEJAA, LEVSOV,
LEVSUB, MUPTEW, QINDAS, QINDEW, QINDIA, SCHOCH10,
SCHOHX10, SIFGAQ, SPOCTC10, SPOCTC10, VIBHIX, and
YIZFET. We calculated the average bond lengths by using these data.
Asolution oftrans-orcis-1(0.05 M)in1 mLofdiphenylether
was heated in an oil bath at 150 °C for 5 h. Then, the reaction
mixture was analyzed with HPLC, as described above.
123

444
Struct Chem (2012) 23:441–449
atom of phenyl group and tetraoxane oxygen atom
iii
˚
[CHÁÁÁO: 3.187 A C20–H26ÁÁÁO1 (symmetry code: (iii)
(a)
1/2 ? x, 1/2 - y, -1/2 ? z)], between the Bu and the
˚
center of gravity of Ph groups [CH/p: 2.986 A C13B–
H19BÁÁÁCg(Be)iv [For the detail of the two characters
in parentheses, see note of Table 2.] (symmetry code:
˚
C20B–
(iv) x - 1/2, 1/2 - y, 1/2 ? z), 2.780
A
H26BÁÁÁCg(Ae)v (symmetry code: (v) 2 - x, -y, -1 - z)
vi
˚
and 2.83 A C20B–H26BÁÁÁCg(Be) (symmetry code: (vi)
2 - x, -y, -1 - z)], and between Ph groups [p–p : Cg–Cg
Cg(Aa)–Cg(Aa)vii, 3.977(6)
˚
A
˚
˚
distance: 4.250(5)
A
vii
vii
Cg(Aa)–Cg(Be) , and 3.722(6) A Cg(Be)–Cg(Be) ;
(symmetry code: (vii) 2 - x, -y, -1 - z)], respectively.
The detailed parameters and schematic representation of
the p–p interactions are shown in Table 2 and Fig. 3,
respectively.
c
b
o
Equivalent isotropic displacement parameters (Ueq) [29]
for cis-1 and trans-1 were listed in Table 4. The averaged
ratios of Ueq(cis)/Ueq(trans), Ph, Bu and 1,2,4,5-tetraoxane
group(s), which are calculated from the averaged Ueq val-
ues for each group of cis-1 and trans-1, are 1.6, 1.2, and
1.3, respectively. The Ph groups for cis-1 have large
displacement parameters in the molecule. This result
[Ueq(cis)/Ueq(trans) ratios [ 1] indicates that the molecules
in the cis-1 crystal may have more librational motion than
those in the trans-1 crystal.
(b)
(e)
Reactivity of trans-1 and cis-1
Irradiation of the diperoxide crystals of trans-1 and cis-1
with 254 nm light at 0 °C for a short time (0.5 h) led to
their decomposition to a small extent and valerophenone
(2) along with butyl benzoate (3) were produced as almost
exclusive products with a molar ratio 2/3 ꢀ 1 (Eqs. 1 and
3 in Scheme 2). In a separate experiment, the time course
of photolysis was examined for both diperoxides (Figs. 4,
5). It appears that at first trans-1 decomposed more rapidly
than cis-1, but the decompositions stopped after prolonged
irradiation in both cases (maximum conversion *6%,
Fig. 4). This may be attributed to the inner filter effect by
the products (2 and 3), which were formed on the crystal
surface and absorb efficiently the incident light (254 nm)
(molar absorptivity e in ethanol at 254 nm: trans-1 520,
cis-1 340, 2 5200, 3 1180 M-1cm-1). Figure 5 displays
that the 2/3 ratio gradually decreased with the irradiation
time. This decrease must have been caused by further
photolysis of the product 2, i.e., the product 2 was con-
sumed by the concurrent Norrish type II photoreaction
[30]. Because of various experimental difficulties (see
‘‘Experimental’’ section), the time-course experiments for
the diperoxide recovery were subjected to large experi-
mental errors, as seen from Fig. 4. Experiments presented
in Eqs. 1 and 3 (Scheme 2) were carried out to confirm the
(c)
(d)
C(20)
C(21)
C(22)
C(20B)
C(21B)
C(22B)
C(9)
C(10)
C(9B)
C(8B)
C(10B)
C(11B)
C(19)
C(19B)
C(8)
C(7)
C(11)
C(18)
C(17)
C(17B)
O(2B)
C(18B)
O(4)
C(6)
O(3)
C(7B)
O(4B) C(6B)
C(16B)
C(16)
C(15)
C(5)
C(15B)
O(2)
O(1)
C(5B)
O(3B)
C(3)
O(1B)
C(4B)
C(14B)
C(4)
C(14)
C(3B)
C(13)
C(13B)
C(2)
C(12B)
C(2B)
C(1)
C(1B)
C(12)
Fig. 2 ORTEP drawings of cis-1: a crystal packing, b disordered
form, c conformer A, d conformer B, and e overlay of the molecular
structures of two disordered form by using AutoMolFit [28]
(PLATON); the red and blue molecules correspond to conformers
A and B, respectively. Thermal displacement ellipsoids are drawn at
50% probability level. The H atoms are represented by spheres, and
the intermolecular interactions are indicated by black (CHÁÁÁO), green
(CHÁÁÁp), and red (pÁÁÁp) broken lines
The O–O and C–O bond lengths for both the conformers
˚
˚
range from 1.470 to 1.486 A and 1.421 to 1.444 A,
respectively, whereas the O–O and C–O bond lengths for
trans-1 are slightly shorter or equal to the average value, as
given in Table 3.
Intermolecular CHÁÁÁO hydrogen bond, CH/p, and p–p
interactions are observed for cis-1 between the hydrogen
123

Struct Chem (2012) 23:441–449
445
Table 2 Geometrical parameters of p–p interactions for cis-1
e
f
g
Cg(I)–Cg(J)a
a (°)b
b (°)c
c (°)d
˚
Distance (A)
˚
CgI_PP (A)
˚
CgJ_PP (A)
˚
Slippage (A)
Cg(Aa)–[Cg(Aa)vii
Cg(Be)–[Cg(Be)vii
Cg(Aa)–[Cg(Be)vii
a
4.250(5)
3.722(6)
3.977(6)
0.0
35.06
25.38
31.91
35.06
25.38
28.53
3.479(4)
3.362(4)
3.494(4)
3.479(4)
3.362(4)
3.376(4)
2.441
1.595
–
0.0
5.80
Geometrical center of phenyl rings. The two characters in parentheses indicate the conformer (A or B) and the position of the phenyl ring
b
c
(e: equatorial or a: axial), respectively. Dihedral angle between phenyl planes I and J. Angle between Cg(I) ? Cg(J) vector and normal to
plane I. Angle between the Cg(I) ? Cg(J) vector and the vector normal to plane I. d Angle between Cg(I) ? Cg(J) vector and normal to plane J.
Perpendicular distance of Cg(I) on ring J. Perpendicular distance of Cg(J) on ring I. This distance corresponding to the interplaner distance.
e
f
g
Distance between Cg(I) and perpendicular projection of Cg(J) on ring I. Symmetry code (vii) (2 - x, -y, -1 - z)
contrary to the above-mentioned solid-state photolysis, the
decomposition of cis-1 (conversion 82%) was faster than
that of trans-1 (conversion 38%). The 2/3 ratio from trans-
1 (2.8) was nearly twice larger than that from cis-1 (1.5),
but these ratios were considerably smaller when compared
with the case of solid-state photolysis (4–10 from Eqs. 1, 3
and Fig. 5).
Thus, it was observed that (i) the 2/3 selectivity was
always more than one (1.5–10), (ii) the 2/3 selectivity was
considerably higher by the solid-state photolysis (4–10)
c
b
a
than by the solution thermolysis (1.5–2.8), and (iii) the 2/3
selectivity from trans-1 was nearly twice higher than that
from cis-1. Photochemical or thermal decomposition of
peroxidic compounds usually begins with cleavage of the
weakest bond, i.e., a peroxide bond. Therefore, diperoxides
trans-1 and cis-1 should decompose mainly from the O–O
a
bond cleavage of path a in Scheme 3 [31–34]. It seems
from the observation (i) that this mode of decomposition
ends up with formation of 2 more than 3. The observation
(ii) is as expected, since the reaction is generally more
selective in crystals than in a solution. Scheme 3 describes
another possible decomposition pathway (path b), where
the cleavage of two C–O bonds occurs first and the ensuing
carbonyl oxide intermediate, as recently described in a
review [35], leads mainly to the formation of ester 3. It may
be assumed that path b is occurring to some extent in
the solution, since the 2/3 selectivity is low. The observed
lower 2/3 selectivity from cis-1 than from trans-1 (the
observation (iii)) may infer that cis-1 has a somewhat
stronger tendency to decompose through path b. It seems
difficult to prove this inference on the basis of the present
X-ray structures, because all the O–O and C–O bond
lengths in the 1,2,4,5-tetraoxane ring of cis- and trans-1 are
normal and there are no clear features in the thermal
parameters of the ring atoms (Tables 3, 4). However, a
preliminary analysis of the X-ray data of cis-1 at 0 °C,
which will be published elsewhere, suggests that its
1,2,4,5-tetraoxane ring atoms are undergoing large libra-
tional motions at 0 °C. Large thermal vibrations at the ring
carbon atoms of cis-1 may be taken to induce the C–O
b
c
Fig. 3 The schematic representation of p–p interactions between the
neighboring phenyl planes for cis-1: symmetry operation (2 - x, -y,
-1 - z). The red and blue dots are the center of gravity for
conformer A and B, respectively. The Cg–Cg distances are shown by
˚
red [Cg(Aa)–Cg(Aa), 4.250 A], blue [Cg(Be)–Cg(Be), 3.722 A], and
˚
black [Cg(Aa)–Cg(Be), 3.977 A] broken lines, respectively. The
interplanar distance 3.479 A (conformer A, red) and 3.362 A
(conformer B, blue). Projections in two directions are shown. There is
no such p–p interaction for trans-1
˚
˚
˚
difference of initial photoreactivity between trans-1 and
cis-1. Now, from inspection of Eqs. 1 and 3 (Scheme 2)
and Figs. 4, 5, it seems apparent that (a) the irradiation of
trans-1 gave nearly twice the 2/3 ratio than the irradiation
of cis-1 and (b) the initial stage of photochemical decom-
position of trans-1 was about three times faster than that of
cis-1.
Thermolysis of trans-1 and cis-1 in diphenyl ether at
150 °C for 5 h was performed. Valerophenone (2) and
butyl benzoate (3) were produced as major products
(Eqs. 2 and 4 in Scheme 2). Formation of the previously
reported byproducts such as phenyl valerate and 2-pheno-
xybiphenyl [15–18] was neglected. It can be noticed that,
123

446
Struct Chem (2012) 23:441–449
Scheme 2 Reactivity of trans-1
and cis-1
products (%)
Ph-OO-Bu Ph-CO-OBu
2/3
conversion
(%)
(molar ratio)
(2)
(3)
Ph
Bu
h (254 nm), 0.5 h
crysral, 0 °C
4.5
38
91
9
10
(1)
(2)
O
O
O
O
150 °C, 5 h
in Ph-O-Ph
a
Bu
Ph
74
26
2.8
trans-1
Ph
Bu
h (254 nm), 0.5 h
crysral, 0 °C
1.6
82
85
60
15
40
(3)
(4)
5.7
1.5
O
O
O
O
150 °C, 5 h
in Ph-O-Ph
a
Ph
Bu
cis-1
a
The formation of by-products [15–18] was neglected.
100
98
bond cleavage (the occurrence of path b). Further X-ray
crystallographic and other work is desired along this line.
Previously one of the present authors has pointed out
that a lower-melting member among the homologous
crystalline compounds may have more free space for
molecular motions and consequently exhibits higher pho-
toreactivity [36–38]. Also, the crystal density was used as a
measure of the free space in crystal packing [36–39].
Besides, he studied coerced control over solid-state [2 ? 2]
and [4 ? 4] photocyclodimerization by varying the crystal
structures through formation of two-component crystals
[25, 26]. As a modifier component, hydrogen-bonding
intermolecular linkers (e.g., gauche 1,2-diamine, phthalic
acid, and oxalic acid) [39–44], small molecules (e.g., NH3
and divalent metal) [26, 45], and relatively small and
planar molecules (e.g., imidazole, trans,trans-muconic
acid, and acetylenedicarboxylic acid) [45, 46] were
employed. In the course of these studies, two-component
crystals were frequently found to be more loosely packed
than one-component crystals and hence two-component
crystals appeared to undergo decreased topochemical
restriction to molecular motions. For example, in addition
to static disorder, dynamic disorder caused by a pedal-like
dynamic conformational change and volume-demanding
reactions such as trans–cis photoisomerization of olefin
occurred relatively easily in cocrystals [25, 26, 43, 44].
Trans-1 melts at 103–104 °C and cis-1 melts at 54–59 °C
[15]. The lower and un-sharp melting point for cis-1 may
partly reflect its completely disordered crystal structure
described above (Fig. 2). From the topochemical point of
view on the melting point [36–38], it is expected that upon
solid-state photolysis cis-1 would decompose faster than
trans-1. The larger librational motions (Ueq) for all atoms of
cis-1 than those of trans-1 (Table 4) might signify that cis-1
would be more reactive than trans-1. However, as Eqs. 1 and
96
94
92
90
0
1
2
3
4
5
irradiation time, h
Fig. 4 Time-course of the decomposition for trans- or cis-valero-
phenone diperoxide (trans-1 or cis-1) upon solid-state photolysis at
254 nm and 0 °C
10
9
trans
8
7
6
5
cis
4
3
0
1
2
3
4
5
irradiation time, h
Fig. 5 Change of the valerophenone/butyl benzoate (2/3) ratio with
time upon solid-state photolysis of trans-1 or cis-1 at 254 nm and
0 °C
123

Struct Chem (2012) 23:441–449
447
Scheme 3 Probable
decomposition pathways for
trans-1 and cis-1
2 Ph-CO-Bu + O2
or
less likely 2 Ph-CO-OBu
O
O
O
O
Bu
Ph
Bu
Ph
(a) O–O cleavage
O
O
O
O
Bu
Ph
Bu
Ph
Bu
Ph
Bu
Ph
Bu
(b) C–O cleavage
O
O
2
O
O
O O
2
2 Ph-CO-OBu
Ph
Table 3 Selected bond lengths
and angles for cis-1 and trans-1
˚
Length (A)
Atom
Atom
Angle (°)
Conformer A of cis-1
C(5)–O(1)
1.421(2)
1.436(2)
1.438(3)
1.428(2)
1.486(2)
1.470(2)
C(5)–O(1)–O(2)
C(5)–O(3)–O(4)
O(1)–O(2)–C(16)
O(3)–O(4)–C(16)
O(1)–C(5)–O(3)
O(2)–C(16)–O(4)
106.75(16)
108.04(16)
106.89(13)
106.84(13)
108.46(16)
107.75(17)
C(5)–O(3)
C(16)–O(2)
C(16)–O(4)
O(1)–O(2)
O(3)–O(4)
Conformer B of cis-1
C(5B)–O(1B)
C(5B)–O(3B)
C(16B)–O(2B)
C(16B)–O(4B)
O(1B)–O(2B)
O(3B)–O(4B)
Trans-1
1.434(3)
1.444(3)
1.431(3)
1.423(3)
1.470(2)
1.476(2)
C(5B)–O(1B)–O(2B)
C(5B)–O(3B)–O(4B)
O(1B)–O(2B)–C(16B)
O(7)–O(8)–C(16B)
107.41(17)
107.31(17)
107.9(2)
106.4(2)
106.8(2)
108.3(2)
O(1B)–C(5B)–O(3B)
O(2B)–C(16B)–O(4B)
C(5)–O(1)
C(5)–O(2viii
1.423(2)
1.429(2)
1.473(1)
C(5)–O(1)–O(2)
O(1)–C(5)–O(2viii
O(1)–O(2)–C(5viii
107.51(6)
108.50(8)
107.50(6)
)
)
)
Symmetry code (viii) (1 - x,
-y, 1 - z)
O(1)–O(2)
Table 4 Ueq values for cis-1
and trans-1
Cis-1
Atom
Trans-1
Ueq
Atom
Ueq
Atom
Ueq
1,2,4,5-Tetraoxane ring
O(1)
O(3)
0.0251(4)
0.0264(4)
O(2)
O(4)
C(16)
0.0248(4)
0.0250(4)
0.0232(6)
O(1)
O(2)
C(5)
0.0187(2)
0.0190(2)
0.0175(2)
C(5)
0.0226(6)
Bu groups
C(1)
0.0556(12)
0.0417(10)
0.0316(9)
0.0283(7)
C(12)
C(13)
C(14)
C(15)
0.0478(13)
0.0361(8)
0.0312(8)
0.0264(7)
C(1)
C(2)
C(3)
C(4)
0.0554(6)
0.0306(3)
0.0239(3)
0.0208(3)
C(2)
C(3)
C(4)
Ph groups
C(6)
0.0253(10)
0.0303(11)
0.0402(19)
0.0411(16)
0.038(2)
C(17)
C(18)
C(19)
C(20)
C(21)
C(22)
0.0249(9)
0.0283(12)
0.0325(13)
0.0272(9)
0.0288(14)
0.0269(15)
C(6)
C(7)
C(8)
C(9)
C(10)
C(11)
0.0175(3)
0.0205(3)
0.0231(3)
0.0237(3)
0.0234(3)
0.0201(3)
C(7)
C(8)
C(9)
C(10)
C(11)
0.0301(13)
123

448
Struct Chem (2012) 23:441–449
3 in Scheme 2 and Fig. 4 demonstrate, trans-1 decomposed
initially faster than cis-1 in the crystal. Recently, we have
reported another topochemical factor that a locally narrow
and rigid reaction cavity due to p–p interactions between the
neighboring aryl planes plays a crucial role in suppressing
the solid-state photoreactivity of as many as seventeen 2,4,6-
trialkylbenzophenones, where alkyl is i-Pr, Et, or Me and the
corresponding benzocyclobutenols are formed [47]. Now
from inspection of the crystal packing of trans-1 and cis-1, it
has become noticed that a similar Ph–Ph p–p overlap exists
for the cis-1 crystal as shown in Fig. 3, but does not for the
trans-1 crystal. Furthermore, from the list of Table 1, cis-1
(Dc 1.209 g/cm3) is more densely packed than trans-1 (Dc
1.169 g/cm3) (Dc 1.172 and 1.132 g/cm3, respectively, at
0 °C). Hence, we propose that the topochemical restriction
of molecular motions by the p–p interaction between the
neighboring Ph planes [47] and the higher crystal packing
[36–39] may be responsible for the observed lower photo-
reactivity of the cis-1 crystal as compared with the trans-1
crystal. However, it must be warned that many other factors
such as defects, surface effects, different extent of light
scattering, particle sizes, formation of different solid–solid
mixtures/eutectic solutions are known to influence the solid-
state reactivity [48–51], but the influences by these factors
were not experimentally investigated in this brief report.
Data Centre (CCDC), 12 Union Road, Cambridge CB2 1EZ,
UK; fax: 144(0) 1223-336033; e-mail: deposit@ccdc.cam.
ac.uk).
Acknowledgments The authors are indebted to Professor Yoshiki
Matsuura (Osaka University), who first solved the crystal structure of
trans-1. We also thank Dr. Sadayuki Asaoka for his assistance in
doing photolysis experiments.
References
1. Ihmels H, Scheffer JR (1999) Tetrahedron 55:885–907
2. Braga D, Chen S, Filson H, Maini L, Netherton MR, Patrick BO,
Scheffer JR, Scott C, Xia W (2004) J Am Chem Soc 126:3511–
3520
3. Scheffer JR, Xia WJ (2005) Top Curr Chem 254:233–262
4. Tanaka K, Toda F (2000) Chem Rev 100:1025–1074
5. Toda F, Tanaka K, Miyamoto H (2001) In: Schanze KS (ed)
Understanding and manipulating excited-state processes, vol 8.
Marcel Dekker, New York, 2001, p. 385
6. MacGillivray LR (2008) J Org Chem 73:3311–3317
7. MacGillivray LR, Papaefstathiou GS, Friscic T, Hamilton TD,
Bucar DK, Chu Q, Varshney DB, Georgiev IG (2008) Acc Chem
Res 41:280–291
8. Georgiev IG, MacGillivray LR (2007) Chem Soc Rev 36:
1239–1248
9. Lauher JW, Fowler FW, Goroff NS (2008) Acc Chem Res 41:
1215–1229
10. Kim HS, Shibata Y, Wataya Y, Tsuchiya K, Masuyama A,
Nojima MJ (1999) J Med Chem 42:2604–2609
11. Vennerstrom JL, Fu HN, Ellis WY, Ager AL Jr, Wood JK,
Andersen SL, Gerena L, Milhous WK (1992) J Med Chem
35:3023–3027
Conclusion
12. O’Neill PM, Amewu RK, Nixon GL, ElGarah FB, Mungthin M,
Chadwick J, Shone AE, Vivas L, Lander H, Barton V, Muang-
noicharoen S, Bray PG, Davies J, Park BK, Wittlin S, Brun R,
Preschel M, Zhang K, Ward SA (2010) Angew Chem Int Ed
49:5693–5697
13. Kim HS, Tsuchiya K, Shibata Y, Wataya Y, Ushigoe Y, Mas-
uyama A, Nojima M, McCullough KJ (1999) J Chem Soc Perkin
Trans 1(13):1867–1870
The crystal structures of valerophenone diperoxides trans-1
andcis-1 weresuccessfullysolvedat-173 °C. The molecular
structures of both the compounds adopt chair conformations.
In the crystal of cis-1, the molecular structure showed whole-
molecule disorder. The thermal parameters for all atoms of
cis-1 were slightly larger than those of trans-1.
Solid-state photolysis (with 254 nm light at 0 °C) or
solution-state thermolysis (at 150 °C) of trans-1 and cis-1
produced valerophenone (2) and butyl benzoate (3) with
the molar ratio 2/3 of 1.5–10. The observed lower 2/3
selectivity for cis-1 than that for trans-1 in the solid state
was not explicable by the present X-ray structures. The
lower solid-state photoreactivity of cis-1 as compared with
trans-1 at the initial stage of photolysis was ascribed to the
former’s locally narrow and rigid reaction cavity due to the
p–p overlap between the neighboring Ph planes.
14. Badawi HM, Al-Saadi AA, Ali AS (2010) J Mol Struct
969:197–203
15. Ito Y, Konishi M, Matsuura T (1979) Photochem Photobiol
30:53–57
16. Ito Y, Matsuura T, Yokoya H (1979) J Am Chem Soc
101:4010–4011
17. Ito Y, Yokoya H, Umehara Y, Matsuura T (1980) Bull Chem Soc
Jpn 53:2407–2408
18. Ito Y, Tone M, Yokoya H, Matsuura T, Schuster GB (1986) J Org
Chem 51:2240–2245
19. Altomare A, Burla M, Camalli M, Cascarano G, Giacovazzo C,
Guagliardi A, Moliterni A, Polidori G, Spagna R (1999) J Appl
Crystallogr 32:115–119
20. Sheldrick GM (2008) Acta Crystallogr A 64:112–122
21. CrystalStructure 3.8, Crystal Structure Analysis Package; Rigaku
and Rigaku Americas (2000–2007): 9009 New Trails Dr., The
Woodlands, TX 77381, USA
Supplementary data
22. Kabuto C, Akine S, Nemoto T, Kwon E (2009) J Cryst Soc Jpn
51:218–224
23. Macrae CF, Edgington PR, McCabe P, Pidcock E, Shields GP,
Taylor R, Towler M, van de Streek J (2006) J Appl Crystallogr
39:453–457
CCDC-787284 (trans-1) and CCDC-787283 (cis-1) contain
the supplementary crystallographic data for this paper. These
data can be obtained free of charge at www.ccdc.cam.ac.uk/
data_request/cif (or from the Cambridge Crystallographic
123

Struct Chem (2012) 23:441–449
449
24. Spek AL (2009) Acta Crystallogr D 65:148–155
25. Ito Y (1998) Synthesis 1:1–32
26. Ito Y (1999) In: Ramamurthy V, Schanze KS (eds) Molecular and
supramolecular photochemistry, vol 3. Marcel Dekker, New
York, pp 1–70
27. Cremer D, Pople JA (1975) J Am Chem Soc 97:1354–1358
28. Mackay AL (1984) Acta Crystallogr A 40:165–166
29. Fischer RX, Tillmanns E (1988) Acta Crystallogr C 44:775–776
30. Wagner PJ (1971) Acc Chem Res 4:168–177
31. McCullough KJ, Morgan AR, Nonhebel DC, Pauson PL (1980) J
Chem Res (S) 35–37
38. Ito Y (1989) In: Anpo M, Matsuura T (eds) Photochemistry on
solid surfaces. Elsevier, Amsterdam, pp 469–480
39. Ito Y, Borecka B, Scheffer JR, Trotter J (1995) Tetrahedron Lett
36:6083–6086
40. Ito Y, Borecka B, Olovsson G, Trotter J, Scheffer JR (1995)
Tetrahedron Lett 36:6087–6090
41. Ito Y (1996) Mol Cryst Liq Cryst 277:247–253
42. Ito Y, Olovsson
1(11):127–133
G (1997) J Chem Soc Perkin Trans
43. Ito Y, Hosomi H, Ohba S (2000) Tetrahedron 56:6833–6844
44. Ohba S, Hosomi H, Ito Y (2001) J Am Chem Soc 123:6349–6352
45. Ito Y, Kitada T, Horiguchi M (2003) Tetrahedron 59:7323–7329
46. Horiguchi M, Ito Y (2006) J Org Chem 71:3608–3611
47. Ito Y, Takahashi H, Hasegawa J, Turro NJ (2009) Tetrahedron
65:677–689
32. Miura M, Nojima M, Kusabayashi
1:1950–1954
S (1980) JCS Perkin
33. Cafferata LFR, Eyler GN, Mirifico MV (1984) J Org Chem
49:2107–2111
34. Cafferata LFR, Eyler GN, Svartman EL, Canizo AI, Borkowski
EJ (1990) J Org Chem 55:1058–1061
35. Ishiguro K, Nojima T, Sawaki Y (1997) J Phys Org Chem
10:787–796
36. Ito Y, Ito H, Ino M, Matsuura T (1988) Tetrahedron Lett
29:3091–3094
48. Garcia-Garibay MA (2003) Acc Chem Res 36:491–498
49. Garcia-Garibay MA (2007) Angew Chem Int Ed 46:8945–8947
50. Kuzmanich G, Natarajan A, Chin K, Veerman M, Mortko CJ,
Garcia-Garibay MA (2008) J Am Chem Soc 130:1140–1141
51. Lebedeva NV, Tarasov VF, Resendiz MJE, Garcia-Garibay MA,
White RC, Forbes MDE (2010) J Am Chem Soc 132:82–84
37. Matsuura T, Meng J, Ito Y, Irie M, Fukuyama K (1987) Tetra-
hedron 43:2451–2456
123