Two-photon laser-induced reaction of 1,8-bis(halomethyl)naphthalenes from different excited states and transient targeting of its intermediate by time-delayed, two-color photolysis and argon ion laser-jet photolysis techniques

Article abstract of DOI:10.1021/ja962884l

Two-photon chemistry of 1,8-bis(bromomethyl)naphthalene (1a) and 1,8-bis(chloromethyl)naphthalene (1b) was studied by (a) laser photolysis with use of XeCl (308 nm), KrF (248 nm), and ArF (193 nm) excimer lasers, (b) time-delayed, two-color photolysis with use of XeCl and XeF (351 nm) lasers, and (c) argon ion laser-jet (333, 351, and 364 nm) photolysis by both direct and benzophenone sensitization. The reaction proceeds through an intermediate monoradical 2, which is generated by a one-photon process, followed by additional photolysis to the two-photon product acenaphthene (4). On excitation of substrate 1 to its S₁ state, monoradical 2 is formed directly from the S₁ state and subsequently from the T₁ state through intersystem crossing, alternatively on benzophenone sensitization. Upper excited states, namely S₂ and S₃, of 1 are proposed in the KrF and AsF excimer laser irradiation, generating intermediate 2 through fast fragmentation of the T(σ*) state. Transient targeting of intermediate 2 by time-delayed, two-color photolysis and argon ion laser-jet photolysis increases considerably the yield of the two-photon product 4.

Full text of DOI:10.1021/ja962884l

592
J. Am. Chem. Soc. 1997, 119, 592-599
Two-Photon Laser-Induced Reaction of
1,8-Bis(halomethyl)naphthalenes from Different Excited States
and Transient Targeting of Its Intermediate by Time-Delayed,
Two-Color Photolysis and Argon Ion Laser-Jet Photolysis
Techniques
Akihiko Ouchi,*,† Yoshinori Koga,† and Waldemar Adam‡
Contribution from the National Institute of Materials and Chemical Research, AIST, MITI,
Tsukuba, Ibaraki 305, Japan, and Institute of Organic Chemistry, UniVersity of Wu¨rzburg,
Am Hubland, D-97074 Wu¨rzburg, Germany
ReceiVed August 16, 1996X
Abstract: Two-photon chemistry of 1,8-bis(bromomethyl)naphthalene (1a) and 1,8-bis(chloromethyl)naphthalene
(1b) was studied by (a) laser photolysis with use of XeCl (308 nm), KrF (248 nm), and ArF (193 nm) excimer
lasers, (b) time-delayed, two-color photolysis with use of XeCl and XeF (351 nm) lasers, and (c) argon ion laser-jet
(333, 351, and 364 nm) photolysis by both direct and benzophenone sensitization. The reaction proceeds through
an intermediate monoradical 2, which is generated by a one-photon process, followed by additional photolysis to the
two-photon product acenaphthene (4). On excitation of substrate 1 to its S1 state, monoradical 2 is formed directly
from the S1 state and subsequently from the T1 state through intersystem crossing, alternatively on benzophenone
sensitization. Upper excited states, namely S2 and S3, of 1 are proposed in the KrF and ArF excimer laser irradiation,
generating intermediate 2 through fast fragmentation of the T(σ*) state. Transient targeting of intermediate 2 by
time-delayed, two-color photolysis and argon ion laser-jet photolysis increases considerably the yield of the two-
photon product 4.
Introduction
Scheme 1
In spite of the wide application of lasers in spectroscopy and
analytical chemistry, comparatively little synthetic work has
been done in the field of organic chemistry.1,2 The most
prominent features of a laser are its high intensity, monochro-
macity, and intermittence in the case of pulsed lasers. Photo-
chemical activation of multifunctional molecules may be carried
out by lasers by utilizing the high intensity of the laser. Most
of such studies have been carried out for the spectral detection
and identification of the short-lived transients involved and the
measurement of their kinetics. However, while such spectro-
scopic studies are, of course, useful to understand multiphoton
reactions, they usually provide no information on the products
of the photochemical process because the sub-milligram quanti-
ties of products that are formed do not allow characterization
and quantitative analysis. Nevertheless, to assess complete
mechanistic details of a chemical transformation, the kinetics
of the process and the detection and identification of the
intermediates as well as product studies are indispensable.
For the present mechanistic study, we have chosen 1,8-bis-
(halomethyl)naphthalenes 1 as the substrate and conducted the
multiphoton reaction shown in Scheme 1.3 This class of
compounds was chosen because extensive spectroscopic studies
have been conducted on the photolysis of monosubstituted
analogs, namely the 1-(halomethyl)naphthalenes 5 (Scheme 2),4
which should provide the necessary background for the mecha-
nistic interpretations of the reactions of the disubstituted
derivatives 1. Analogous to the formation of the 1-naphthyl-
methyl radical 6 in the laser photolysis of 5, it was expected
that the corresponding naphthylmethyl radicals 2 should be
formed in the photolysis of 1. Subsequent photochemical
† National Institute of Materials and Chemical Research.
‡ Institute of Organic Chemistry.
X Abstract published in AdVance ACS Abstracts, December 15, 1996.
(1) (a) Kleinermanns, K.; Wolfrum, J. Angew. Chem., Int. Ed. Engl. 1987,
26, 38-58. (b) Wilson, R. M.; Schnapp, K. A. Chem. ReV. 1993, 93, 223-
249. (c) Johnston, L. J. Chem. ReV. 1993, 93, 251-266. (d) Wilson, R.
M.; Schnapp, K. A.; Hannemann, K.; Ho, D. M.; Memarian, H. R.; Azadnia,
A.; Pinhas, A. R.; Figley, T. M. Spectrochim. Acta, Part A 1990, 46, 551-
558. (e) Wilson, R. M.; Adam, W.; Schulte Oestrich, R. Spectrum. 1991,
4, 8-17.
(2) (a) Wilson, R. M.; Hannemann, K.; Peters, K.; Peters, E.-M. J. Am.
Chem. Soc. 1987, 109, 4741-4743. (b) Netto-Ferreira, J. C.; Wintgens,
V.; Scaiano, J. C. Tetrahedron Lett. 1989, 30, 6851-6854. (c) Bendig, J.;
Mitzner, R. Ber. Bunsen-Ges. Phys. Chem. 1994, 98, 1004-1008. (d)
Mathews, M. G.; Wang, Z.; Koplitz, B. Chem. Phys. Lett. 1994, 227, 633-
638.
(3) Preliminary results: (a) Ouchi, A.; Yabe, A. Tetrahedron Lett. 1992,
33, 5359-5362. (b) Adam, W.; Ouchi, A. Tetrahedron Lett. 1992, 33,
1875-1878. (c) Ouchi, A.; Yabe, A. Chem. Lett. 1995, 945-946.
S0002-7863(96)02884-3 CCC: $14.00 © 1997 American Chemical Society

Reaction of 1,8-Bis(halomethyl)naphthalenes from Excited States
J. Am. Chem. Soc., Vol. 119, No. 3, 1997 593
Scheme 2
activation of the monoradical intermediate 2 should afford the
1,8-naphthalenediylbismethyl 3.5 The chemistry of this non-
Kekule´ species is well-established, which cyclizes to acenaph-
thene (4).5d,f In Scheme 1, product 4 derives from a two-photon
process and by monitoring its yield, the efficiency of the double
activation may be assessed.
Figure 1. Absorption spectra of 1,8-bis(bromomethyl)naphthalene (1a)
[s], 1,8-bis(chloromethyl)naphthalene (1b) [- - -], and acenaphthene
(4) [‚‚‚]. Concentration 10-5 M in cyclohexane, optical path 10 mm.
The wavelengths of the laser emissions are marked in the figure.
The efficiency of the photochemical formation of the mono-
radical 6 from 5 has been reported to depend on the type of
excited state that is involved. Thus, it is anticipated that the
efficiency of the two-photon process 1 f 4 will be increased
not only by applying high-intensity light for the reaction but
also by generating the appropriate excited state, which is
accomplished by selecting the proper wavelength in the laser
irradiation. In addition, the timing of a second photon absorp-
tion by the monoradical intermediate 2 is also expected to be
crucial for the effective conversion of 2 f 4 since the generation
of 6 from 5 has been reported to proceed from its T1 state whose
lifetime is far longer than the pulse width of an excimer laser.
These features are demonstrated in the present quantitative
product studies on the multiphoton photolysis of substrates 1a,b
by using (a) XeCl (308 nm), KrF (248 nm), and ArF (193 nm)
excimer lasers, (b) time-delayed, two-color pulsed photolysis
with XeCl and XeF (351 nm) excimer lasers, and (c) argon ion
laser-jet photolysis (333, 351, and 364 nm) both by direct and
benzophenone sensitization.
maximum at 294 nm and the second at 229 nm. The
monosubstituted analogs 1-(bromomethyl)naphthalene (5a) and
1-(chloromethyl)naphthalene (5b) also possess similar absorp-
tion spectra; the absorption maxima of 5a are located at 291
and 226 nm and those of 5b at 284 and 224 nm.
Excimer laser photolyses were conducted in order to study
the wavelength and the fluence dependence of 1a,b. Figure
2a,b shows the yield of 4 and the consumption of 1a,b at high
(3 × 1022 photon‚m-2‚pulse-1) excimer laser fluence photolyses
as a function of the number of laser shots. As seen in the traces,
the conversion of 1a,b depended markedly on the laser that was
used. The conversion increased with increasing laser emission
wavelength, namely XeCl > KrF > ArF. These results did
not simply reflect the absorbance of the substrates because the
magnitude of the absorption coefficient at each laser wavelength
was ArF > KrF ≈ XeCl for 1a and ArF > XeCl > KrF for 1b
(cf. Figure 1). Although the absorbance of 1b was smaller than
that of 1a, the conversion rate for 1b was greater than for 1a in
the XeCl and KrF laser photolyses. In the case of ArF laser
photolysis, however, the rate of the consumption of 1a and 1b
was almost the same.
Results
As shown in Figure 2a,b, the increase in the yield of 4
depended somewhat irregularly on the wavelength of the lasers
used, i.e., KrF > XeCl > ArF. This is the result of fluctuations
in both the conversion and the efficiencyswhich is defined as
100 × (yield of 4)/(conversion of 1)sat each laser irradiation.
The efficiency (Figure 3) of the formation of 4 for the various
excimer laser photolyses decreased with increasing laser wave-
length, namely ArF > KrF > XeCl. The substrate dependence
on the efficiency was 1a > 1b in the XeCl and KrF laser
photolyses, but it did not change much in the case of the ArF
excimer laser photolysis.
Despite the high conversions of 1 for the XeCl and KrF lasers
the efficiency and the yield of 4 were low. This may be
explained by the coupling of the monoradicals 2 (Scheme 1),
which form dimers4e,g and possibly oligomeric and polymeric
products, as well as by unfavorable secondary reactions of the
substrate 1 and the product 4 with the released halogen atoms.
Indeed, higher-molecular-weight compounds were observed by
TLC analysis of the crude photolysate, as evidenced by
considerable amounts of non-eluting products at the origin. As
a control experiment, the decomposition of 4 was tested in the
high-intensity KrF laser irradiation in the presence and absence
of bromine molecules. The results (Table 1) show considerable
consumption of 4 in this photolysis, especially in the presence
of bromine.
Absorption spectra of 1,8-bis(bromomethyl)naphthalene
(1a), 1,8-bis(chloromethyl)-naphthalene (1b), and acenaphthene
(4) in cyclohexane are shown in Figure 1. The spectra in
acetonitrile were almost the same as those in cyclohexane; the
shift of the absorption maxima was less than 5 nm. As seen in
the spectra, 1a has its first absorption maximum at 302 nm and
the second at 230 nm; similarly, 1b shows the first absorption
(4) (a) Slocum, G. H.; Kaufman, K.; Schuster, G. B. J. Am. Chem. Soc.
1981, 103, 4625-4627. (b) Kelley, D. F.; Milton, S. V.; Huppert, D.;
Rentzepis, P. M. J. Phys. Chem. 1983, 87, 1842-1843. (c) Gould, I. R.;
Tung, C.; Turro, N. J.; Givens, R. S.; Matuszewski, B. J. Am. Chem. Soc.
1984, 106, 1789-1793. (d) Hilinski, E. F.; Huppert, D.; Kelley, D. F.;
Milton, S. V.; Rentzepis, P. M. J. Am. Chem. Soc. 1984, 106, 1951-1957.
(e) Slocum, G. H.; Schuster, G. B. J. Org. Chem. 1984, 49, 2177-2185.
(f) Johnston, L. J.; Scaiano, J. C. J. Am. Chem. Soc. 1985, 107, 6368-
6372. (g) Tolumura, K.; Udagawa, M.; Itoh, M. J. Phys. Chem. 1985, 89,
5147-5149. (h) Arnold, B.; Donald, L.; Jurgens, A.; Pincock, J. A. Can.
J. Chem. 1985, 63, 3140-3146. (i) Scaiano, J. C.; Johnston, L. J. Pure
Appl. Chem. 1986, 58, 1273-1278. (j) van Haelst, L.; Haselbach, E.;
Suppan, P. Chimia 1988, 42, 231-234. (k) Redmond, R. W.; Wayner, D.
D. M.; Kanabus-Kaminska, J. M.; Scaianao, J. C. J. Phys. Chem. 1989, 93,
6397-6401. (l) Kawai, A.; Okutsu, T.; Obi, K. Chem. Phys. Lett. 1990,
174, 213-218. (m) Kawai, A.; Okutsu, Y.; Obi, K. Chem. Phys. Lett. 1995,
235, 450-455.
(5) (a) Gisin, M.; Wirz, J. HelV. Chim. Acta 1976, 59, 2273-2277. (b)
Pagni, R. M.; Burnett, M. N.; Dodd, J. R. J. Am. Chem. Soc. 1977, 99,
1972-1973. (c) Platz, M. S. J. Am. Chem. Soc. 1979, 101, 3398-3399.
(d) Platz, M. S. J. Am. Chem. Soc. 1980, 102, 1192-1194. (e) Platz, M.
S.; Carrol, G.; Pierrat, F.; Zayas, J.; Auster, S. Tetrahedron 1982, 38, 777-
785. (f) Haider, K.; Platz, M. S.; Despres, A.; Lejenne, V.; Migirdicyan,
E.; Bally, T.; Haselbach, E. J. Am. Chem. Soc. 1988, 110, 2318-2320. (g)
Biewer, M. C.; Platz, M. S.; Roth, M.; Wirz, J. J. Am. Chem. Soc. 1991,
113, 8069-8073.
When the irradiations were conducted at medium fluence of
the excimer lasers, i.e., 6.2 × 1020 photon‚m-2‚pulse-1 (1/50 of
the high-fluence photolyses), the rate of the consumption

594 J. Am. Chem. Soc., Vol. 119, No. 3, 1997
Ouchi et al.
Figure 2. The consumption of substrate 1 (open symbols) and yield of acenaphthene (4) (closed symbols) as a function of the number of excimer
laser shots. Substrate and fluence of excimer lasers: (a) 1a, high; (b) 1b, high; (c) 1a, medium; (d) 1b, medium. Eximer lasers: 0, 9 (ArF); O,
b (KrF); 4, 2 (XeCl). Concentration 10-5 M in cyclohexane. Fluence of excimer lasers 3.1 × 1022 (high) and 6.2 × 1020 (medium)
photon‚m-2‚pulse-1. The results are the average of two independent runs; the experimental errors are given as Supporting Information.
On further decrease of the light intensity, the photolysis
became expectedly still less efficient. The low-fluence KrF
1
excimer laser (6.2 × 1018 photon‚m-2‚pulse-1
,
/
of the
100
medium-fluence photolysis) and the low-pressure mercury lamp
(254 nm, 3.4 × 1018 photon‚m-2‚s-1) photolyses of 1a,b only
showed slow consumption of the starting materials and no
formation of acenaphthene (4).6
Time-Delayed, Two-Color Photolyses. This technique
involves irradiation with a single XeCl laser pulse (6.2 × 1020
photon‚m-2‚pulse-1), followed by a time-delayed XeF laser
pulse (9.4 × 1020 photon‚m-2‚pulse-1). The photolysis with
the XeF laser alone showed no consumption of 1a,b, which
was in accord with the lack of absorption of 1a,b at this
wavelength (cf. Figure 1). The consumption of 1a,b in the two-
color photolysis was independent of the delay times; the average
consumption was 29% for 1a and 91% for 1b after three sets
of two-color irradiationssa set of irradiation comprises a pulse
of XeCl and XeF lasersswhich were the same as those of the
one-color laser photolysis by the XeCl laser. These results
indicate that the consumption of 1a,b depends only on the XeCl
laser and not on the subsequent XeF laser irradiation.
Figure 3. Efficiency of acenaphthene (4) formation on irradiation with
various high-fluence excimer lasers. Concentration 10-5 M in cyclo-
hexane. Fluence of excimer lasers 3.1 × 1022 photons‚m-2‚pulse-1
.
Utilized excimer lasers and substrates: O (ArF, 1a); b (ArF, 1b); 0
(KrF, 1a); 9 (KrF, 1b); 4 (XeCl, 1a); 2 (XeCl, 1b).
Table 1. Decomposition of Acenaphthene (4) by KrF Laser
Irradiationsa in the Presence and Absence of Bromine
decomposition of 4 (%)
no. of laser pulses
no additive
with Br2 (1 equiv)
Figure 4 shows the effect of delay time on the formation of
acenaphthene (4) in terms of the increment factor, which is
defined by (A - B)/B, where A is the yield of 4 in the two-
color photolysis and B is the yield of 4 in the one-color XeCl
laser irradiation. In these figures, an increment factor of 0
means that 4 was formed only by the first XeCl laser and no
additional amounts of 4 are produced by the subsequent XeF
laser irradiation. In other words, an increment factor >0
indicates additional formation of 4 in the photolysis of an
intermediate by the second XeF laser pulse. The expected
intermediate is the monoradical 2 with strong absorption at 330-
360 nm,7 similar to that of the parent 1-naphthylmethylradical
6 (ca. 340 and 365 nm).4e,g
1
2
3
5
7
4.7 ( 1.0
11.3 ( 5.1
14.7 ( 3.4
20.0 ( 6.7
23.9 ( 3.6
32.1 ( 5.2
15.1
29.7
39.1
45.8
61.3
63.5
10
a Fluence: 3.1 × 1022 photon‚m-2‚pulse-1. Pulse width (fwhm):
30 ns. Concentration: 10-5 M in cyclohexane.
decreased for both 1a and 1b (Figure 2c,d). The yield of 4
also diminished in the photolyses; for the KrF laser photolysis
it dropped to ca. 1/4 for 1a and 1/6 for 1b compared with those
of the high-fluence photolyses. In the case of the XeCl
photolyses, practically no acenaphthene was formed for both
1a and 1b.
(6) The experimental results are given as Supporting Information.

Reaction of 1,8-Bis(halomethyl)naphthalenes from Excited States
J. Am. Chem. Soc., Vol. 119, No. 3, 1997 595
Figure 4. Increment factor for the yield of acenaphthene (4) as a
function of delay time: (a) from 1a, (b) from 1b. Concentration of 1
10-5 M in cyclohexane; laser power of XeCl 6.2 × 1020 photons‚
m-2‚pulse-1 and of XeF 9.4 × 1020 photons‚m-2‚pulse-1. Three sets
of two-color irradiations were conducted at each run. The jitter between
the two laser pulses is (3 ns. The results are the average of four
independent runs. The insert shows the increment factor in the range
of 0.5-10 (a) and 0.7-10 µs (b) with second-order kinetic fitting.
Figure 5. The consumption of 1 (open symbols) and yield of 4 (solid
symbols) as a function of the number of laser-jet cycles. Substrate
and solvent: (a) 1a, cyclohexane; (b) 1a, acetonitrile; (c) 1b, acetonitrile.
Laser power 3.5 W; concentration of 1a 10-4 M; flow rate 2.0
mL‚min-1; (i) standard reaction (O, b), (ii) with 0.1 M NB (4, 2),
(iii) with 10-3 M BP (0, 9), (iv) with 10-3 M BP and 0.1 M NB (],
[). The results are the average of two independent runs; the
experimental errors are given as Supporting Information.
As seen in the traces of Figure 4, the increment factor strongly
depends on the delay time. It is significant to note that the
increment factor has two maxima for both 1a and 1b, the first
maximum at 0-30 ns and the second at 0.2-0.5 µs, although
the separation of the first and the second maxima of substrate
1a is not clearly observed. The existence of two maxima
suggests the presence of at least two photochemical paths for
the formation of 2.
The yield of acenaphthene (4) between 0.5-0.7 and 10 µs
delay time followed second-order decay (Figure 4, insert). The
result implies that the decrease in the concentration of 2 is due
to dimerization, analogous to the related naphthylmethyl radical
6, for which a second-order decay (dimerization) was reported.4g
After a delay time of ca. 10 µs, the increment factor decreased
to 0, which indicates complete absence of intermediate 2 in this
time domain.
In Figure 5a are displayed the results of the laser-jet
photolyses in cyclohexane solutions. Four types of laser-jet
photolyses were conducted by using all three emissions of the
argon ion laser (total output energy 3.5 W): (i) standard
irradiation without any additive, (ii) irradiation in the presence
of the halogen atom scavenger8 norbornene (NB), (iii) ben-
zophenone (BP)-sensitized irradiation, and (iv) BP-sensitized
irradiation in the presence of NB. As shown in Figure 5a, the
addition of the halogen atom scavenger NB does not affect the
consumption of 1a but increases the yield of acenaphthene (4)
both in the direct [type (i) Versus (ii)] and BP-sensitized [type
(iii) Versus (iv)] photolyses. This provides evidence that
undesirable side reactions due to the released halogen atoms
occur, which decrease the yield of 4. This is consistent with
the results obtained in the excimer laser photolyses (Figure 2
and Table 1). The results for the BP sensitization revealed that
both the consumption of 1a and the yield of 4 increase [type
(i) Versus (iii) and (ii) Versus (iv)].
Argon ion laser-jet photolyses were conducted for the direct
and the triplet-sensitized irradiation under continuous high-
intensity conditions. The argon ion laser emits continuously at
the three near-UV wavelengths 333, 351, and 364 nm (output
energy distribution 1:2:2).
(7) Ito, O.; Alam, M. M.; Koga, Y.; Ouchi, A. J. Photochem. Photobiol.,
A 1996, 97, 19-23.
(8) The effect of other halogen atom scavengers was also studied and
the results are given as Supporting Information.

596 J. Am. Chem. Soc., Vol. 119, No. 3, 1997
Ouchi et al.
Figure 6. The dependence of the laser intensity on the consumption
of 1a (open symbols) and yield of 4 (closed symbols). Concentrations
in acetonitrile: 1a 10-4 M, BP 10-4 M, NB 0.1 M. Flow rate 2.0
mL‚min-1. Laser-jet cycles: O, b (one); 4, 2 (three); 0, 9 (five).
Figure 8. The dependence of the flow rate on (a) the consumption of
1a and yield of 4 and (b) the consumption of 1a and yield of 4 per
photon: 1a (O), 4 (b) for one laser-jet cycle. Concentrations in
acetonitrile: 1a 10-4 M, BP 10-4 M, NB 0.1 M.
Figure 7. The dependence of the wavelength on the consumption of
1a (dashed line) and yield of 4 (solid lines). Laser power 3.5 W (total
output); NB: 0.1 M. Irradiations at (i) 333 nm (O), (ii) 333 nm + 0.1
M BP (b); (iii) 351 nm (4), (iv) 351 nm + 0.1M BP (2), (v) 364 nm
(0), (vi) 364 nm + 0.1M BP (9). The results are the average of two
independent runs; the experimental errors are given as Supporting
Information.
of 61% when all three lines are employed [Figure 5b (iv)]. The
slight decrease in the former case is due to the decrease of the
laser intensity by using a quartz prism for the separation of the
three laser lines. In contrast, the yield of 4 for BP sensitization
(at the fifth cycle) was 1.2% for 333 nm, 1.9% for 351 nm, and
0.9% for 364 nm, or a total of 4.0%, which is about half of that
[8.1%, Figure 5b (iv)] with all three lines.
Laser-jet photolyses were also conducted in acetonitrile under
the type (i)-(iv) modes (Figure 5b). A similar trend of reactions
was observed compared with those in the cyclohexane solution
but with a remarkable increase in both the consumption of 1a
and the yield of 4; however, a prominent difference is the lower
consumption of 1a in the BP-sensitized reaction in the presence
of NB, although the yield of 4 was increased [type (ii) Versus
(iv)].
For the explicit assessment for the participation of a two-
photon process, the intensity effect was studied for the laser-
jet photolysis (Figure 6). The consumption of 1a displayed unit
slope for a double-logarithmic plot Versus light intensity, which
indicates that the reaction proceeded by a one-photon process.
In contrast, the slope of the yield of 4 of such a plot was 2,
which confirms the expected two-photon process.
The effect of each wavelength in the laser-jet photolysis was
examined in Figure 7. In the direct photolysis, the consumption
of 1a was only achieved by the 333-nm emission of the laser
but not by 351- or 364-nm emissions since substrate 1a does
not absorb at 351 and 364 nm; however, the yield of 4 was
low. In contrast, in the BP-sensitized photolyses, consumption
of 1a and the formation of 4 were observed at all three laser
lines. The consumption of 1a in the BP sensitization (at the
fifth cycle) was 26% for 333 nm, 18% for 351 nm, and 8% for
364 nm, or a total of 52%, which is similar to the consumption
The dependence on the flow rate of the liquid jet, which
corresponds to the contact time of the substrate molecules in
the laser beam, is shown in Figure 8, in which the horizontal
axis is shown in both flow rate (lower scale) and contact time
(upper scale). As seen in Figure 8a, the consumption of 1a
increases with decreasing flow rate, while the yield of 4 levels
off at flow rates below 1.2 mL‚min-1. In Figure 8b the contact
time dependence on the consumption of 1a and the yield of 4
is displayed but normalized for the number of irradiated photons.
As seen in the figure, the consumption per photon is essentially
independent of the contact time, which means that the consump-
tion increases linearly with the number of photons; however,
the yield of 4 per photon decreased linearly with the contact
time, which suggests decomposition of 4 probably by the one-
photon process. As a control experiment, the decomposition
of substrate 4 by BP-sensitized laser-jet photolysis was tested.
Indeed, with increasing laser-jet cycles and BP concentration,
product 4 was more effectively decomposed.6
For comparison purpose, the laser-jet photolysis of derivative
1b was also conducted in acetonitrile for the four reaction types,
(i)-(iv), as shown in Figure 5c. Without BP no consumption
of 1b was observed in type (i) and (ii) modes, which is in accord
with the lack of absorption at the argon ion laser emissions (cf.
Figure 1). In the case of BP-sensitized reactions, types (iii)

Reaction of 1,8-Bis(halomethyl)naphthalenes from Excited States
J. Am. Chem. Soc., Vol. 119, No. 3, 1997 597
Scheme 3
and (iv), consumption of 1b was observed which was somewhat
slower than that of 1a; however, practically no formation of
product 4 was observed in both cases.
tives 1 to generate intermediate 2 (Scheme 1 and 3), probably
with a shorter S1 lifetime due to the larger heavy atom effect in
the facile ISC to the T1 state; i.e., 1(S0) f 1*(S1) f 1*(T1) f
2. Indeed, the existence of the monoradicals 2 is established
by the detection of a transient absorption at 330-360 nm in
the laser flash photolysis of 1a,b.7
Discussion
The results of the direct photolysis can be interpreted in terms
of a mechanism established for the monosubstituted substrates
5 (Scheme 2). The first absorption maximum of the monosub-
stituted analogs, i.e., λmax 291 nm of 5a and λmax 284 nm of
5b, by irradiation with a XeCl excimer laser or a nitrogen laser
(337 nm), has been reported to involve the excitation to the
S1(π,π*) state of the naphthalene ring.4f Therefore, the pho-
tolysis of 1a,b by a XeCl excimer laser or by the 333-nm
emission of the argon ion laser should also proceed through
the S1(π,π*) state (Scheme 3). The irradiation to the second
maximum of 5a,b, i.e., λmax 226 nm of 5a and λmax 224 nm of
5b, by irradiation with the fourth harmonic of a Nd:YAG laser
(266 nm) represents the excitation to the S2(π,π*) state.4b,d Thus,
the photolysis of 1a,b by a KrF excimer laser should also
proceed through the S2(π,π*) state of the naphthalene moiety.
The type of excitation in the ArF excimer laser irradiation is
still not well understood because there are no previous reports
on 5a,b which characterize the higher excited states that are
produced. However, the excited state is different from the S1
or S2 state because the results are different from those of XeCl
and KrF laser photolyses (cf. Figures 2 and 3); presumably they
are either the S3(n,σ*) state of the C-X bond9 or the S3(π,π*)
state of the naphthalene ring.
An additional photon is necessary for the formation of
acenaphthene (4), i.e., 2 f 2* f 4. This is supported by the
fact that with the decrease of the XeCl excimer laser fluence,
the decrease on the yield of acenaphthene (4) is larger than the
decrease in the consumption of 1. This is also confirmed clearly
by the argon ion laser-jet experiments in that the consumption
of the substrate followed a one-photon process whereas the
formation of acenaphthene required two photons (Figure 6).
In spite of the large consumption of substrate 1, only a small
yield of the two-photon product 4 was obtained in the XeCl
excimer laser photolysis. This result can be explained in terms
of the time lag between the laser pulse and the formation of
intermediate 2. In view of the fact that the pulse width of the
XeCl laser is only 14 ns, which is one order in magnitude shorter
than the lifetime of the T1 state of 1a,b, the monoradicals 2
formed from the T1 state of substrate 1 in the one-color
photolysis have little probability of absorbing a second photon
within the same pulse for the formation of the two-photon
product 4. Thus, the monoradical 2 experiences extensive side
reactions to afford dimers and higher-molecular-weight products.
In the case of the KrF laser photolysis of 1 (Figure 2), upper
excited states are expected in analogy to substrate 5. When 5
is irradiated by the fourth harmonic of a Nd:YAG laser (266
nm), it is excited to the S2 state which, in addition to the internal
conversion to the S1 state, leads extremely quickly (within 25
ps, i.e., within the pulse width of the Nd:YAG laser) to excited
naphthylmethyl radical 6. The fast formation of the excited
naphthylmethyl radical 6 derives from ISC of the S2 to the T(σ*)
state, subsequent homolytic cleavage of the C-X bond, and
absorption of a second photon by radical 6. It is likely that
such a mechanism also operates for the disubstituted naphtha-
lenes 1 in the KrF laser irradiation (Scheme 3), i.e., 1(S0) f
1*(S2) f 1*[T(σ*)] f 2 f 2* f 4. The pulse width of 30 ns
Acenaphthene (4) Formation from the Different Excited
States. In the case of the XeCl laser photolysis, substrate 5 is
excited to its S1(π,π*) state,4f,g whose lifetime is <5 ps for 5a
and 490 ps for 5b,10 followed by facile intersystem crossing
(ISC) to the T1(π,π*) state.4f Substantial homolytic cleavage
of the C-X bond in the T1 state, which has the lifetime of 0.5-
0.7 µs,4f,g was reported to generate radical 6.4l A similar
mechanism is proposed to operate for the disubstituted deriva-
(9) (a) Calvert, J. G.; Pitts, J. G., Jr. Photochemistry; John Wiley &
Sons: New York, 1966; Chapter 5, Section 5-8. (b) Horspool, W.;
Armesto, D. Organic Photochemistry; Ellis Horwood PTR Prentice Hall:
Chichester, 1992; Chapter 6, Section 6.1.

598 J. Am. Chem. Soc., Vol. 119, No. 3, 1997
Ouchi et al.
for the KrF excimer laser is long enough for such a two-photon
process to occur and a higher yield of product 4 is obtained
accordingly.
In the case of ArF laser photolysis, no studies have been
conducted with 5 previously. The fact that the results (yield,
efficiency, and consumption) are different from those obtained
by XeCl and XeF excimer laser photolyses (Figure 2) suggests
the participation of an additional reaction path, most likely from
higher excited states; however, intermediary states are not clear
at present.
observed spectroscopically for substrates 1 and 5 because of
the overlap of the absorption spectra derived from the multiple
transient species at the initial stage of the photolysis; however,
such a fast process has been proposed through product analysis
in the photolysis of 5,4e from which formation of radical 6
directly from the S1 state was proposed. Therefore, in analogy
we suggest that additionally the monoradical 2 is generated from
the substrate 1 through the sequence 1(S0) f 1*(S1) f 2 in
Scheme 3.
Transient Targeting of Intermediate 2 by Argon Ion
Laser-Jet Photolyses. Another way of overcoming the disad-
vantage of the XeCl excimer laser photolysis is the utilization
of high-intensity, continuous light sources so that intermediate
2, efficiently generated from the T1(π,π*) state, can have
sufficient time for the absorption of a second photon. The laser-
jet method fulfills this requirement and also has the advantage
that the argon ion laser emits continuously at the three
wavelengths 333, 351, and 364 nm. The 333-nm emission is
suitable for the first and the 351- and 364-nm lines for the
second step photolysis (Scheme 1). An additional advantage
is that undesirable side reactions of the intermediate monoradical
2 are minimized because of the favorable 1:2:2 intensity
distribution of the three wavelengths. Thus, the sum of the 351-
and 364-nm emissions is better than 4-fold more intense than
the 333-nm emission and the monoradical 2 is photochemically
more quickly consumed at the 351- and 364-nm wavelength
than generated at the 333-nm wavelength and thereby prevented
from accumulating sufficiently for bimolecular side reactions
(dimerization).
The existence of the postulated non-Kekule´ type diradical
1,8-naphthalenediylbis(methyl) (3)5 as an intermediate in step
2* f 4 (in Scheme 3) is still not certain. The absorption
spectrum of 3 has been measured in an EPA glass and by flash
photolysis of 1-(diazomethyl)-8-methylnaphthalene, which showed
several strong absorption maxima in the 300-340-nm region
and some weak maxima in the 400-500-nm region.5g Our
attempt to detect the intermediate 3 in the nanosecond flash
photolysis of 1a,b in cyclohexane showed no clear absorption
maxima in these wavelength regions.7 Unfortunately, the
transient absorption spectrum of 3 has not been measured in
the direct photolysis of 1 in the liquid phase, but its ESR
spectrum has been reported during the photolysis of 1,1,2,2-
tetrakis(dimethylamino)ethene matrix, probably formed by
electron transfer.5f An intramolecular radical SH2-type reaction
2 f 4 is also likely to occur, without the intervention of
intermediary diradical 3.
Transient Targeting of Intermediate 2 by Time-Delayed,
Two-Color Photolysis. Although the consumption of 4 was
highest for the three excimer lasers used, the disadvantage of
the XeCl excimer laser for the formation of acenaphthene (4)
lies in the time lag between the laser pulse and the formation
of intermediate 2. This disadvantage can be circumvented by
applying a second laser pulse when the concentration of
intermediate 2 is maximal. Indeed, a substantial increase of
the yield of 4 was accomplished in this way (Figure 4).
A significant feature of this method is that the increment
factor, i.e., the yield of 4, reflects the concentration of
intermediate 2 at each delay time. Only the intermediate can
absorb XeF excimer laser emission and the additional yield of
4 should depend linearly on the concentration of the transient.
Figure 4 shows two maxima for the increment factor, the first
at 0-30 ns and the second at 0.2-0.5 µs. The second maximum
pertains to the monoradical 2 derived from the long-lived T1
state of 1 through the sequence 1 f 1*(S1) f 1*(T1) f 2 f
2* f 4 (Scheme 3). The delay time of 0.2-0.5 µs at the
maximum is in good accord with the expected rise time of the
monoradical 2 through the T1 state, since the lifetimes of the
S1 and T1 states of the monosubstituted naphthalene 5 are <1
ns10 and 0.5-0.7 µs.4f,g Additional alkyl substitution on radical
6 would be in line with a slight decrease in the T1 lifetime11 of
1a,b compared with 5a,b. Additional evidence for the proposed
reaction path in Scheme 3 is given by the fact that the second
maximum in Figure 4 is higher than the first for 1a, but Vice
Versa for 1b. This is rationalized in terms of the more efficient
population of the T1 state in the case of 1a because of more
effective ISC 1*(S1) f 1*(T1), which is promoted by the heavy
atom effect of bromine.
Indeed, although the fluence (7.9 × 1026 for the laser-jet
Versus 4.5 × 1028 photon‚m2‚s-1 for the XeCl excimer laser)
and the number of photons per contact (1.5 × 1014 photon‚cycle-1
for the laser-jet Versus 6.3 × 1016 photon‚pulse-1 for the XeCl
1
1
excimer laser) of the argon ion laser-jet were /57 and /420 of
that of the medium fluence XeCl excimer laser, an ca. 4.5-fold
increase in the yield of 4 (at 36-38% conversion of 1a) during
the photolysis in cyclohexane was observed in the argon ion
laser-jet [Figure 5a, type (i)] compared with the medium fluence
XeCl excimer laser (Figure 2). Furthermore, the actual increase
of the yield of 4 is expected to be higher than 4.5-fold because
the laser-jet experiments were conducted at 10-fold higher
concentration of 1a than those of the XeCl excimer laser
photolyses. In this context, it has been reported that the yield
of 4 decreased with the increase of the substrate concentration.12
Further increase in the consumption of 1a and the yield of
acenaphthene (4) was accomplished by the BP sensitization
[Figure 5a, type (i) Versus (iii)]. This is due to the more efficient
formation of intermediate 2 through the 1*(T1) state. In the
direct photolysis, as shown in Figure 7, the excitation is only
accomplished by the 333-nm emission whose energy is merely
20% of the total output. In contrast, for the BP sensitization
the 351- and 364-nm emissions of the argon-ion laser generate
also the intermediate 2 through the 1*(T1) state in addition to
the 333-nm emission. The change of solvent from cyclohexane
to acetonitrile showed significant increase of the yield and
consumption in the reaction (Figure 5a Versus Figure 5b). The
results can be rationalized by the facility of the halogen atom
escape from the solvent cage13 or by the participation of
The first maximum observed at very short delay time (0-30
ns) in the time-delayed, two-color photolyses (Figure 4) provides
evidence for another path in the formation of the monoradical
2. The existence of this fast process has not been unequivocally
(12) In the case of the medium- and high-fluence KrF excimer laser
photolysis of 1a, the yield of 4 decreased to 1/75 when the concentration
of 1a was changed from 10-5 to 10-3 M.6
(13) The fraction of escaped radical pairs from the cage is a linear
function of (1/η)n, where η is the viscosity of solvent and n is a value
between 0.5 and 1.16 The viscosity of cyclohexane is 0.608 cP (30 °C)
and that of acetonitrile is 0.325 cP (30 °C).17 Therefore, the escape of
halogen atoms will be larger in acetonitrile than in cyclohexane so that the
higher consumption of 1 and the yield of 4 should be observed.
(10) Huppert, D.; Rand, S. D.; Reynolds, A. H.; Rentzepis, P. M. J. Chem.
Phys. 1982, 77, 1214-1224.
(11) Carmichael, I.; Hug, G. L. Handbook of Organic Photochemistry;
Scaiano, J. C., Ed.; CRC Press: Boca Raton, 1989; Vol. I, Chapter 16.

Reaction of 1,8-Bis(halomethyl)naphthalenes from Excited States
J. Am. Chem. Soc., Vol. 119, No. 3, 1997 599
[XeF, 9.4 × 1020 photon‚m-2‚pulse-1, pulse width (fwhm) 26 ns]
excimer laser. The pulse energy was measured by means of a Gentec
joulemeter ED-500 and a Tektronic T912 10MHz storage oscilloscope.
The delay times between the two laser pulses were controlled by a
four-channel digital delay/pulse generator (Stanford Research Systems
Inc., Model DG535). The delay times of the two laser pulses were
measured with a biplaner photoelectric tube (Hamamatsu Photonics
K. K., model R1193U-05) and a storage scope (Iwatsu Electric Co.
Ltd., model TS-8123).
additional ionic reaction paths;14 however, the explicit reason
for this solvent effect is not clear at present.
Conclusions
The laser chemistry of 1,8-bis(bromomethyl)naphthalene (1a)
and 1,8-bis(chloromethyl)naphthalene (1b) proceeds through an
intermediate monoradical 2, which is generated by a one-photon
process (Scheme 1), followed by successive photolysis of 2 to
afford the two-photon product acenaphthene (4). The yield of
two-photon product 4 was optimized by means of time-delayed,
two-color photolysis and argon ion laser-jet photolysis, the latter
by direct and benzophenone sensitization; significant increase
in the yield of product 4 was observed in both photolysis
techniques. On excitation to the S1 state by using the XeCl
excimer laser, the monoradical 2 was formed directly from the
S1 state and also from the T1 state, which was generated by
ISC from the S1 state. Participation of higher excited states is
also proposed, e.g., population of the S2 state by irradiation with
the KrF excimer laser, in which presumably fast C-X homolysis
though the T(σ*) state generates the monoradical intermediate
2.
Argon Ion, Laser-Jet Photolyses. These photolyses were conducted
under an argon gas atmosphere by using the reported apparatus.1d,e The
argon ion laser was an Innova 100-20 system from Coherent [output
energy distribution 333 (20%), 351 (40%), and 364 nm (40%)] and
the laser was focused by a quartz convex lens (f ) 80 mm). The laser
output was 3.5 W, whose fluences at each emission (photons‚m-2‚s-1
)
were 1.5 × 1026 for 333 nm, 3.1 × 1026 for 351 nm, and 3.3 × 1026 for
364 nm. In some experiments, the three emissions of the laser were
separated by means of a quartz prism. The solutions were prepared
by dissolving the substrate in an appropriate solvent and norbornene
was added in some experiments. The sensitized photolyses were
conducted with benzophenone as the triplet sensitizer. The solution
was cycled through the apparatus for 30 to 60 min under an argon gas
atmosphere before photolysis in order to eliminate the dissolved oxygen
in the solution. The diameter of the capillary in the laser-jet apparatus
was 100 µm and the specified flow rate was controlled by means of a
Bischoff HPLC 2200 pump.
Experimental Section15
Analysis of the Photolysis Products. The yield of the two-photon
product, namely acenaphthene (4), and the consumption of the starting
material were determined by HPLC analysis on a Merck Supersphere
60 RP-8 column (244 mm × 4 mm i.d.) and their retention times
compared with the authentic samples. The HPLC system consisted of
a JASCO TRI Rotar-V HPLC pump and a JASCO 870-UV UV/VIS
detector, or a Shimadzu LC-6A HPLC pump and a Shimadzu SPD-
6AV UV-Vis spectrometric detector.
Excimer Laser Photolyses. These reactions were conducted in
cyclohexane solution under a nitrogen gas atmosphere. High-intensity
photolyses (3.1 × 1022 photon‚m-2‚pulse-1) were conducted by using
an ca. 1.5-cm capillary of 1-mm diameter, one of whose ends was
closed. A 10-µL aliquot of the solution was placed into a horizontally
placed capillary and the laser light was focused at the opening of the
capillary, at which a stream of nitrogen gas was directed. Medium-
and low-intensity photolyses (6.2 × 1020 and 6.2 × 1018 photon‚
m-2‚pulse-1) were conducted on 0.5-1.0-mL solution under a nitrogen
gas atmosphere by using a synthetic quartz cuvette of 10-mm width
and 10-mm optical path. For the irradiations, the excimer lasers used
were Lambda Physik EMG 102 [pulse width (fwhm) 14 ns (XeCl)]
and Lambda Physik EMG 201 [pulsewidth (fwhm) 23 ns (ArF, typical),
30 ns (KrF), and 26 ns (XeF)]. The pulse energy was measured by
means of a Gentec joulemeter ED-500 and a Tektronix T912 10MHz
storage oscilloscope.
Acknowledgment. We thank the Deutsche Forschungsge-
meinschaft, the Fonds der Chemischen Industrie, and the Agency
of Industrial Science and Technology (MITI, Japan) for financial
support. A.O. expresses his gratitude to the Alexander von
Humboldt Foundation for a research fellowship (1991).
Time-Delayed, Two-Color Photolyses. These reactions were
conducted on 0.5 mL of 10-5 M cyclohexane solutions under a nitrogen
gas atmosphere by using a synthetic quartz cuvette of 10-mm width
and 10-mm optical path. The first pulse was generated by a Lambda
Physik EMG 102 [XeCl, 6.2 × 1020 photon‚m-2‚pulse-1, pulse width
(fwhm) 14 ns] and the second pulse by a Lambda Physik EMG 201
Supporting Information Available: Preparation of com-
pounds 1a,b. Experimental errors in Figures 2, 5, and 7;
photolysis of 1a,b by low-fluence KrF and KrCl excimer lasers,
and low-pressure mercury lamp; substrate concentration effect
on the high- and medium-fluence KrF excimer laser photolyses
of 1a,b; dependence on the benzophenone concentration and
halogen atom scavengers in the argon-ion laser-jet photolysis
of 1a; decomposition of 4 by benzophenone-sensitized argon-
ion laser-jet photolysis (12 pages). See current masthead page
for ordering and Internet access instructions.
(14) In addition to a radical reaction path, the presence of an ionic path
has been reported in the photolysis of 5 in methanol: ref 4e.
(15) Preparation of compounds 1a,b is given as Supporting Information.
(16) Leffler, J. E. An introduction to free radicals; John Wiley & Sons:
New York, 1993; Chapter 4, Section 4.2.2.
(17) Riddick, J. A.; Bunger, W. B. Techniques of Chemistry, Organic
SolVents, 3rd ed.; Wiley-Interscience Publication, John Wiley & Sons: New
York, 1970; pp 139 and 399.
JA962884L