Engineering reactions in crystals: Gem-dialkoxy substitution enables the photodecarbonylation of crystalline 2-indanone

Article abstract of DOI:10.1016/S0040-4039(02)01549-6

The photochemical reactivity of three 1,1,3,3-tetrasubstituted 2-indanones was investigated in solution and in crystals. While solution irradiation of tetramethyl-2-indanone 2 leads to 1-isopropenyl-2-isopropyl benzene 5 in high yield, crystals of 2 were

Full text of DOI:10.1016/S0040-4039(02)01549-6

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 43 (2002) 7063–7066
Engineering reactions in crystals: gem-dialkoxy substitution
enables the photodecarbonylation of crystalline 2-indanone
Danny Ng, Zhe Yang and Miguel A. Garcia-Garibay*
Department of Chemistry, University of California, Los Angeles, CA 90095-1569, USA
Received 24 May 2002; revised 21 July 2002; accepted 26 July 2002
Abstract—The photochemical reactivity of three 1,1,3,3-tetrasubstituted 2-indanones was investigated in solution and in crystals.
While solution irradiation of tetramethyl-2-indanone 2 leads to 1-isopropenyl-2-isopropyl benzene 5 in high yield, crystals of 2
were completely inert. In contrast, samples of 1,3-bis(ethylenedioxy)-2-indanone 3 reacted efficiently in both media. While solution
irradiation yielded benzocyclobutane 7 and o-xylylene 8, only the former was formed in the solid state. Finally, the 1,3-bis-thioke-
tal analog 4 reacted sluggishly in solution but not at all in the solid state. Differences in reactivity between the three indanones
were analyzed in terms of radical stabilization effects and possible intramolecular quenching mechanisms. © 2002 Elsevier Science
Ltd. All rights reserved.
Although many reactions in crystals proceed with
remarkably high selectivity and specificity,1 their limited
predictability has prevented chemists from exploiting
their potential in synthetic applications. With the pur-
pose of identifying reliable solid state reactions, we
have explored the use of compounds that form high-
energy species such as carbenes,2 biradicals, and radical
pairs.3 We anticipate that highly reactive intermediates
will break weak bonds and make new stronger ones
despite having severely limited motion in the solid state.
The more exothermic these reactions, the earlier and
lower their transition states, and the higher their rate
constants and quantum efficiencies.4a,6,7
In order to document the scope of the reaction, we have
analyzed several a-substituents3,5 and potential quench-
ing mechanisms that may prevent ketones from reacting
in the solid state.8 With regard to the effect of a-phenyl
groups, it is well known that conformational factors
have a strong effect on the stability of benzylic radi-
cals.9 Experimental10 and computational9 studies sug-
gest that conformers with ideal overlap between the
radical p-orbital and the aromatic p-system are 12.5
1.5 kcal/mol lower in energy than those where the two
groups are orthogonal. In this context, the solid state
reactivity of 2-indanones is of significant interest.
Although the carbonyl group of 2-indanones is for-
mally a,a%-diaryl substituted, the perpendicular orienta-
tion between the s-bonds undergoing cleavage (C1ꢀC2
and C2ꢀC3) and the aromatic p-system may deprive
them of the benefits of benzylic stabilization (Scheme
2). However, the photochemical a-cleavage and decar-
bonylation reactions of several 2-indanones have been
well documented in solution, suggesting that benzylic
stabilization may not be essential in the first step.11,12 If
To generate radical pairs and biradicals in crystals, we
have analyzed the stepwise a-cleavage (Norrish type-I
reaction)4 and decarbonylation of triplet dialkyl
ketones.3,5 We have shown that cyclic and open chain
ketones with a,a%-diphenyl and a,a%-diester substituents
can react smoothly in the solid state (Scheme 1). As the
bond dissociation energies of the a-bonds are reduced
by the a-substituents, the bond cleavage reactions
become more exothermic from the triplet excited state.6
3
formation of the acyl–alkyl triplet biradical BR-1 by
Scheme 1.
a-cleavage also occurs in the solid state, differences in
reactivity may arise from different radical stabilization
during the decarbonylation step. If rotation about the
Keywords: solid state reactivity; a-cleavage; decarbonylation; radical
stability; 2-indanones; benzyl radicals.
* Corresponding author. Tel.: 1-310-825-3158; fax: 1-310-825-0767;
e-mail: mgg@chem.ucla.edu
3
CꢀCꢁO bond in BR-1 is hindered in the solid state,
diminished benzylic stabilization may slow down the
rate of decarbonylation to the extent that formation of
0040-4039/02/$ - see front matter © 2002 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(02)01549-6

7064
D. Ng et al. / Tetrahedron Letters 43 (2002) 7063–7066
Scheme 2.
3
the triplet biradical BR-2 may not compete with inter-
system crossing to the singlet acyl–alkyl biradical 1BR-1
which would undergo rapid bond-formation the ground
state (Scheme 2).
Solution- and solid-state photochemistry of 1,1,3,3-tetra-
methyl-2-indanone (2). As reported by Starr and East-
man,12 the photoreaction of 2 proceeds in solution with
high chemical efficiency to give 1-iso-propenyl-2-iso-
propylbenzene 5 in 95% yield. However, it was shown
that 5 originates (at least partially) from tetramethyl-
In the absence of benzylic stabilization, the reactivity of
2-indanones in crystals should depend on the enabling
effect of substituents R1–R4. Interestingly, cis- and
trans-1,3-diphenyl-2-indanones 1a and 1b (Scheme 2)
are known to give cis- and trans-diphenylbenzocy-
clobutanes stereoselectively upon irradiation in the
solid state.13 While their crystals react efficiently, it is
not known whether the indanone structure can facili-
tate the solid state reaction on its own, or whether the
a-phenyl groups are required. In order to answer this
question and to expand on our previous observations,
we have analyzed the solution and solid state reactivity
of 2-indanones 2, 3 and 4 (Scheme 3).
ortho-xylylene
6 by a photochemical 1,5-H shift
(Scheme 4).19 When crystals of 2 were irradiated at
ambient temperature for 12 h, no trace of a reaction
was observed.
The lack of reactivity in the solid state may come from
inefficient a-cleavage and/or decarbonylation steps.
Although the absolute efficiency of solid state reactions
is difficult to quantify, we have noticed that a-cleavage
reactions that are efficient in solution can also be
observed in the solid state.3,5 Based on that, we
attribute the difference in photoreactivity between solu-
tion and the solid state to the restriction imposed by the
rigid medium on rotation about the a,b-bond and its
effects on decarbonylation (Scheme 2). Flash photolysis
studies in solution by Scaiano and co-workers indicate
Differences in reactivity between 2-indanones 2, 3 and 4
should reflect the effects of gem-dialkyl, ketal, and
thioketal substituents on the a-carbons. Since we have
shown that alkyl groups alone do not enable decar-
bonylation in the solid state,3 results with the tetra-
methyl derivative 2 should reveal whether or not the
indanone ring satisfies the required stabilization in the
solid state. With compounds 3 and 4 we can examine
whether the added effect of dialkoxy and dithioxy
substituents can assist (or hinder) the reaction. Tetra-
methyl indanone 2 (mp=75–76°C)12 was prepared from
2-indanone by standard procedures. Compounds 3 and
4 were selected for their known crystallinity (mp=
115°C14 and mp=197–198°C,15 respectively), simple
preparation, and because of the potential radical stabi-
lizing abilities of the gem-dialkoxy and gem-dithioxy
groups.16–18
3
that decarbonylation of BR-1 (R1–R4=Me) occurs in
3
less than 10 ns. The dialkyl biradical BR-2 formed in
this step goes on to products (5 and 6) with a time
constant of 580 ns,11b,c which is limited by intersystem
crossing to the singlet state. In the absence of bond
rotation, poor alignment between the aromatic p-sys-
tem and the breaking CꢀCO bond should slow decar-
bonylation with respect to intersystem crossing,
resulting in bond formation to regenerate the ground
state starting material.
Solution- and solid-state photochemistry of 1,3-bis
(ethylenedioxy)-2-indanone (3). The photochemical
decarbonylation of 1,3-bis(ethylenedioxy)-2-indanone 3
was recently reported as a convenient way to prepare
benzocyclobutanedione 9 (Scheme 5).14 Irradiation of 3
in THF was reported to yield 48% of bisketal 7.14 We
noticed that irradiation of 3 in benzene through a Pyrex
filter produced a new UV–vis absorption band with a
umax=420 nm, which is a typical value for o-xylylenes11
(Fig. 1, dashed line). Continued photolysis of 3 led to
disappearance of this band, as expected from the
known photolabile nature of these species.11
Although pure samples of 8 cannot be isolated due to
their low steady state concentrations, NMR analysis of
partially irradiated samples of 3 in d6-benzene gave
clear evidence of benzocyclobutane 7 and xylylene 8
Scheme 3.

D. Ng et al. / Tetrahedron Letters 43 (2002) 7063–7066
7065
Scheme 4.
Irradiation of pulverized crystals of bisketal 3 at room
temperature with a Pyrex filter for 3 h afforded benzo-
cyclobutane 7 as the only product after 25% conversion
(Scheme 5). Longer irradiation under these conditions
led to partial melting and formation of additional prod-
ucts. Photolysis at 0°C was sluggish with less than 5%
formation of cyclobutane 7 after 10 h. The absence of
o-xylylene 8 in the solid state is to be expected and is in
agreement with results previously reported with 1a and
1b.13 While conversion of bisketal 3 to benzocyclobu-
tane 7 involves only a small change in shape and
volume, the formation of o-xylylene requires rotation
of the two ketal units from nearly orthogonal to nearly
co-planar with respect to the aromatic group. The
crystalline environment should restrict such drastic
motion.
Scheme 5.
Solution- and solid-state photochemistry of 1,3-bis
(ethylenedithioxy)-2-indanone (4). To our knowledge,
the photochemistry of bisthioketal 4 has not been docu-
mented. The photochemical reactivity of 4 in solution
was extremely sluggish and only complex mixtures were
observed upon extended irradiation times. Crystals of 4
were completely stable at room temperature despite the
potential radical stabilization endowed to the a-carbons
by the thioketal substituents. We attribute the high
photostability of 4 to intramolecular quenching. There
are several reported examples of triplet ketone 3n,p*
quenching by photoinduced electron transfer from
thioethers.20 In support of this hypothesis, we were able
to quench the solution reactivity of the bisketal 3 with
added 1,3-dithiane. A Stern–Volmer analysis that
assumes diffusion-controlled quenching suggests an
excited state lifetime of <4 ns, as expected for a highly
reactive excited state.
Figure 1. UV–vis spectra of 2-indanone 3 (—); spectra after
15 min (- - - -), and 45 min (· · · ·) of irradiation, respectively.
The transient absorption between 400 and 475 nm is assigned
to ortho-xylylene 8.
(Scheme 5). In addition to the signals of 7, a doublet of
triplets and a broad doublet assigned to o-xylylene 8
were observed at l =6.0 and l=5.7 ppm, respectively.
Other signals present were attributed to secondary pho-
toproducts of 8. No olefinic protons were observed in
the NMR spectrum when 3 was irradiated in the pres-
ence of maleic anhydride suggesting effective trapping
of 8. However, we established that reaction of 8 with
maleic anhydride is photochemical as no reaction was
observed after 5 days in the absence of light. The
formation of the photolabile o-xylylene 8 and its sec-
ondary photoproducts may account for low yield of 7
in solution.14
Concluding remarks. The lack of solid state reactivity in
tetramethyl indanone 2 suggests that the photodecar-
bonylation of 1a and 1b must be facilitated by the
radical stabilizing abilities of the a-phenyl groups. The
solid-state reactivity of bisketal 3 appears to be due to
the gem-dialkoxy substituents rather than to benzylic
stabilization by the indanone aryl group. These results
indicate that gem-dialkoxy groups may enable decar-
bonylation of a wide variety of crystalline ketones,
potentially allowing synthetic applications with suitable
substrates. In contrast, the lack of reactivity of the
bisthioketal 4 is probably due to fast intramolecular

7066
D. Ng et al. / Tetrahedron Letters 43 (2002) 7063–7066
quenching. Additional examples and other functional
groups are currently under investigation.
11. (a) Paul, T.; Hassan, M. A.; Korth, H.-G.; Sustmann, R.;
Avila, D. V. J. Org. Chem. 1996, 61, 6835–6848; (b)
Netto-Ferreira, J. C.; Wintgens, V.; Scaiano, J. C. Tetra-
hedron Lett. 1989, 30, 6851–6854; (c) Scaiano, J. C.;
Wintgens, V.; Netto-Ferreira, J. C. Pure Appl. Chem.
1990, 62, 1557–1564; (d) McCullough, J. J. Acc. Chem.
Res. 1980, 13, 270–276.
Acknowledgements
This work was supported by the NSF (grants
DMR9988439 and CHE0073431). Support by Merck
Pharmaceutical Co. is gratefully acknowledged.
12. Starr, C. R.; Eastman, R. H. J. Org. Chem. 1966, 31,
1393–1402.
13. (a) Quinkert, G.; Tabata, T.; Hickman, E. A. J.; Dobrat,
W. Angew. Chem., Int. Ed. Engl. 1971, 10, 199; (b)
Quinkert, G.; Palmovski, J.; Lorenz, H.-P.; Wiersdorff,
W.-W.; Finke, M. Angew. Chem., Int. Ed. Engl. 1971, 10,
198.
14. Leinweber, D.; Butenscho¨n, H. Tetrahedron Lett. 1997,
38, 6385–6386.
15. Scho¨nberg, A.; Singer, E.; Eschenhof, B.; Hoyer, G.-A.
Chem. Ber. 1978, 111, 3058–3067.
16. The radical stabilizing energy of substituent X [RSE(X)]
refers to the decrease in the bond dissociation energy of
the HꢀCH2X bond as compared to HꢀCH3. See Ref. 6
and: Parkinson, C. J.; Mayer, P. M.; Radom, L. J. Chem.
Soc., Perkin Trans. 2 1999, 2305–2313.
References
1. (a) Scheffer, J. R.; Garcia-Garibay, M. A.; Nalamasu, O.
Org. Photochem. 1987, 8, 249–347; (b) Toda, F. Acc.
Chem. Res. 1995, 28, 480–486; (c) Gamlin, J. N.; Jones,
R.; Leibovitch, M.; Patrick, B.; Scheffer, J. R.; Trotter, J.
Acc. Chem. Res. 1996, 29, 203–209; (d) Ramamurthy, V.;
Venkatesan, K. Chem. Rev. 1987, 87, 433–481.
2. Garcia-Garibay, M. A.; Shin, S. H.; Sanrame, C. Tetra-
hedron 2000, 56, 6729–6737.
3. (a) Peterfy, K.; Garcia-Garibay, M. A. J. Am. Chem. Soc.
1998, 120, 4540–4541; (b) Choi, T.; Peterfy, K.; Khan, S.
I.; Garcia-Garibay, M. A. J. Am. Chem. Soc. 1996, 118,
12477–12478.
4. (a) Weiss, D. S. Org. Photochem. 1981, 5, 347–420; (b)
Bohne, C. In CRC Handbook of Organic Photochemistry
and Photobiology; Horspool, W. M.; Ong, P.-S., Eds.;
CRC Press: Boca Raton, 1995.
5. (a) Yang, Z.; Garcia-Garibay, M. A. Org. Lett. 2000, 2,
1963–1965; (b) Yang, Z.; Ng, D.; Garcia-Garibay, M. A.
J. Org. Chem. 2001, 66, 4468–4475.
6. Campos, L. M.; Dang, H.; Ng, D.; Yang, Z.; Martinez,
H.; Garcia-Garibay, M. A. J. Org. Chem. 2002, 67,
3749–3754.
7. Fisher, H.; Paul, H. Acc. Chem. Res. 1987, 20, 200–206.
8. Garcia-Garibay, M. A.; Ng, D.; Yang, Z. Tetrahedron
Lett. 2001, 42, 9113–9116.
9. (a) Hrovat, D. A.; Borden, W. T. J. Phys. Chem. 1994,
98, 10460–10464; (b) Dorigo, A. E.; Li, Y.; Houk, K. N.
J. Am. Chem. Soc. 1989, 111, 6942–6948.
17. Davidson, E. R.; Chakravorty, S.; Gajewski, J. J. New J.
Chem. 1997, 21, 533–537.
18. Based on thermochemical data, we estimate the RSE of
simple alkoxy and thioxy substituents to be ca. 9 and 12
kcal/mol, respectively. (a) CRC Handbook of Chemistry
and Physics; CRC Press: Boca Raton, FL, 1984; Chapter
9, pp. 64–74; (b) Good, D. A.; Francisco, J. S. Chem.
Phys. Lett. 1997, 266, 512–514; (c) Jefferson, A.;
Nicovich, J. M.; Wine, P. H. J. Phys. Chem. 1994, 98,
7128–7135.
19. de Fonseka, K. K.; McCullough, J. J.; Yarwood, A. J. J.
Am. Chem. Soc. 1979, 101, 3277–3282.
20. (a) Guttenplan, J. B.; Cohen, S. G. J. Org. Chem. 1973,
38, 2001–2007; (b) Inbar, S.; Linschitz, H.; Cohen, S. G.
J. Am. Chem. Soc. 1982, 104, 1679–1682; (c) Marciniak,
B.; Bobrowski, K.; Hug, G. L. J. Phys. Chem. 1993,
97, 11937–11943; (d) Marciniak, B.; Bobrowski, K.; Hug,
G. L. J. Photochem. Photobiol. A: Chem. 1994, 81, 159–
168.
10. Conradi, M. S.; Zeldes, H.; Livingsrton, R. J. Phys.
Chem. 1979, 83, 2160–2161.