4-Substituted 1-acyloxypyridine-2(1H)-thiones: Experimental and computational studies of the substituent effect on electronic absorption spectra

Article abstract of DOI:10.1021/jo901522r

(Chemical Equation Presented) A series of eight 4-substituted 1-(adamantane-1-carbonyloxy)pyridine-2(1H)-thiones (1, X=H, OC₇H ₁₅, Me, CF₃, SC₃H₇, CN, COOMe, and Cl) was prepared and characterized by UV-vis spectroscopy in MeCN and cyclohexane. The observed lowest energy transition, designated as π_CS→ π*_ring, exhibits a substantial substituent effect and λ_max ranges from 333 (X=OC ₇H₁₅) to 415 nm (X=CN). Experimental λ _max values for all esters except for 1b (X=OC₇H ₁₅) correlate with the σ_p - parameter (ρ = 0.41 ± 0.03, r²= 0.95). In contrast, the energy of the absorption band at about 295 nm, designated as π_CS → π*_CS, is practically substituent independent. Both absorption bands exhibit a modest negative solvatochromic effect. The experimental absorption energies correlate better with excitation energies calculated for N-acetyloxy analogues 2 with the ZINDO//DFT than with the TD-DFT//DFT method. Calculations for a series of 12 N-acetates 2 predict the most blueshifted π → π* transition for the alkoxy substituent and most red-shifted for the NO₂ group relative to the parent 2a (X = H). 2009 American Chemical Society.

Full text of DOI:10.1021/jo901522r

pubs.acs.org/joc
4-Substituted 1-Acyloxypyridine-2(1H)-thiones: Experimental and
Computational Studies of the Substituent Effect on Electronic
Absorption Spectra†
Aleksandra Jankowiak and Piotr Kaszynski*
Organic Materials Research Group, Department of Chemistry, Vanderbilt University, Nashville,
Tennessee 37235
piotr.kaszynski@vanderbilt.edu
Received July 15, 2009
A series of eight 4-substituted 1-(adamantane-1-carbonyloxy)pyridine-2(1H)-thiones (1, X = H,
OC7H15, Me, CF3, SC3H7, CN, COOMe, and Cl) was prepared and characterized by UV-vis
spectroscopy in MeCN and cyclohexane. The observed lowest energy transition, designated as πCS f
π*ring, exhibits a substantial substituent effect and λmax ranges from 333 (X = OC7H15) to 415 nm
-
(X=CN). Experimental λmax values for all esters except for 1b (X=OC7H15) correlate with the σp
parameter (F = 0.41 ( 0.03, r2 = 0.95). In contrast, the energy of the absorption band at about
295 nm, designated as πCS f π*CS, is practically substituent independent. Both absorption bands
exhibit a modest negative solvatochromic effect. The experimental absorption energies correlate
better with excitation energies calculated for N-acetyloxy analogues 2 with the ZINDO//DFT than
with the TD-DFT//DFT method. Calculations for a series of 12 N-acetates 2 predict the most blue-
shifted π f π* transition for the alkoxy substituent and most red-shifted for the NO2 group relative to
the parent 2a (X = H).
Introduction
of carboxylic acids.1 One such an example is adamantane-1-
carbonyloxy derivative 1a.1 To date, ring-substituted Barton
esters are practically unknown2,3 despite the fact that a
handful of substituted 1-hydroxypyridine-2(1H)-thiones,
including 4-propyl,4 4-fluoroalkyl,5 mercapto,6 4-methoxy-
carbonyl,7 3-ethoxy,8 and several methyl8-10 and dimethyl9
derivatives, have been described in the literature. The substituent
O-Acyl derivatives of 1-hydroxypyridine-2(1H)-thione,
so-called Barton esters, are popular precursors of free radicals
and are often used as reagents in free radical transformations
† Presented in part at 9th Annual Florida Heterocyclic and Synthetic Con-
ference (Flohet-9), Gainesville, FL, March 9-12, 2008, and the 237th ACS
National Meeting, March 22-26, 2009, Salt Lake City, Utah.
(1) Knight, D. W. In Encyclopedia of Reagents for Organic Synthesis;
Paquette, L. A., Ed.; Wiley&Sons: New York, 1995; Vol. 4, pp 2775-2778.
Crich, D. In Comprehensive Organic Synthesis; Trost, B. M., Fleming, I., Ley,
S. V., Eds.; Pergamon: New York, 1991; Vol. 7, pp 717-734 and references
cited therein.
(2) Edrissi, M.; Jadbabaee, M. J.; Dalziel, J. A. W. Microchem. J. 1971,
16, 536–537.
(3) Pyridine-2(1H)-thiones substituted with methyl (positions 4 and 5)
and NO2 (position 5) were implied in a synthetic scheme, but no details or
characterization data were reported: Rao, U. N.; Biehl, E. J. Org. Chem.
2002, 67, 3409–3411.
(4) Rees, C. W. J. Chem. Soc. 1956, 3684–3685.
(5) Sun, J.-Y.; Qiu, X.-L.; Meng, W.-D.; Qing, F.-L. Tetrahedron 2006,
62, 8702–8706.
(6) Puszko, A.; Talik, Z. Pol. J. Chem. 1991, 65, 377–380.
(7) Yale, H. L.; Losee, K.; Martins, J.; Holsing, M.; Perry, F. M.;
Bernstein, J. J. Am. Chem. Soc. 1953, 75, 1933–1942.
(8) Shaw, E.; Bernstein, J.; Losee, K.; Lott, W. A. J. Am. Chem. Soc. 1950,
72, 4362–4364.
(9) Abramovitch, R. A.; Knaus, E. E. J. Heterocycl. Chem. 1975, 12, 683–
690.
(10) Puszko, A. Pol. J. Chem. 1994, 68, 657–664.
DOI: 10.1021/jo901522r
r
Published on Web 09/09/2009
J. Org. Chem. 2009, 74, 7441–7448 7441
2009 American Chemical Society

Jankowiak and Kaszynski
JOCArticle
SCHEME 1
SCHEME 2
could presumably allow for controlling the sensitivity of the
Barton esters to visible light and tune the wavelength of
radical chain initiation.11,12 In addition, appropriately func-
tionalized Barton esters could be incorporated into more
complex molecular architectures of radiation/light sensitive
functional materials.
In this context, we recently described a method for the
preparation of 4-alkoxy-1-hydroxypyridine-2(1H)-thiones
including 1b.13 In contrast to the yellowish parent 1a, the
4-heptyloxy derivative 1b is a colorless solid.13 This finding
prompted us to investigate the substituent effect on photo-
physical properties of Barton esters, and to develop synthetic
access to several diverse 4-substituted 1-hydroxypyridine-
2(1H)-thiones. Here, we describe the preparation and spec-
troscopic studies of eight esters 1a-h which contain the
adamantyl group chosen for the convenience of isolation
and purification of the products. For a better understanding
of the substituent-photophysical properties relationship in
esters 1, we conducted TD-DFT and ZINDO calculations
for a dozen of their acetyl analogues, esters 2.
the corresponding thiones 3. Both thioacetates 5 and thiones
3 were sensitive to silica gel and light. Therefore, without
rigorous purification the crude products were reacted with
MeONa to generate the sodium salts 3[Na] used for the
formation of esters 1. The overall yield for the three-step
procedure was 40-66%.
The 4-cyano, 4-methoxycarbonyl, and 4-chloro deriva-
tives exhibited sensitivity to the basic MeONa. Therefore,
the 4-cyanopyridine-2(1H)-thione (3f) was liberated from
the thioacetate 5f by using aqueous NH4Cl giving an ap-
proximately 75% pure product used for the esterification
step (Scheme 2). The purity of the thioacetate 5g obtained
either from 2-chloro or 2-bromo N-oxide 4g and AcSNa was
low. Therefore, thione 3g was obtained by using thiourea
followed by hydrolysis with aqueous Na2CO3, according to a
modified patent procedure.14 Both halides, the 2-chloro and
2-bromo esters 4g[Cl] and 4g[Br], respectively, gave compar-
able results. The isolated thione 3g was then converted to the
sodium salt 3g[Na] with NaH (Scheme 3) and subsequently
to 1g. Our brief investigation of the formation of the
isopropyl analogue of 3g from 6 demonstrated that the
reaction was much less clean and the product was only
∼40% pure, and consequently was not pursued further.
The thioacetate route for the preparation of 4-chloro ester
1h was unsuccessful. Thioacetate 5h was easily obtained from
the known 2,4-dichloropyridine N-oxide15 (4h) as a typical
mixture with the partially deprotected thione 3h. However,
the reaction of 5h with MeONa gave multiple products.
Attempts to liberate the thione form the thioacetate 5h by
using aqueous NH4Cl, in a manner analogous to the pre-
paration of 3f, were plagued with low mass recovery and the
formation of multiple products. Therefore, we focused on
the reaction of dichloride 4h with thioacetamide, which was
previously used to prepare 2,4-dimercaptopyridine N-oxide.6
Thus, a reaction of 4h with limited amounts of thioacetamide
(1.3 equiv) in an inert atmosphere with use of literature
conditions6 gave, in one instance, a monomercapto deriva-
tive in 63% yield. The regioselectivity of the thiolation reaction
was established by alkylation of a sample of the isolated
Results
Synthesis. Adamantane-1-carboxylic acid O-esters 1 were
prepared by acylation of the corresponding 1-hydroxypyr-
idine-2(1H)-thione sodium salts 3[Na] with adamantane-1-
carbonyl chloride (Scheme 1). The 4-cyano and 4-chloro
derivatives 1f and 1h were obtained in a reaction of the acid
chloride with appropriate thione 3f or 3h in the presence of
pyridine. Attempts to generate the sodium salt 3f[Na] from 3f
with use of NaH led to significant decomposition of the
substrate. Salts 3b[Na]-3e[Na] were obtained from 2-chlor-
opyridine N-oxides 4b-e according to our recently reported
general procedure (Scheme 1).13 Thus, treatment of 4 with
sodium thioacetate gave crude thioacethoxy derivatives 5
which, isolated after acidic workup, were contaminated with
(11) Hartung, J.; Spehar, K.; Svoboda, I.; Fuess, H.; Arnone, M.; Engels,
B. Eur. J. Org. Chem. 2005, 869–881.
(12) Arnone, M.; Engels, B. J. Phys. Chem. A 2007, 111, 3161–3165.
(13) Jankowiak, A.; Jakowiecki, J.; Kaszynski, P. Pol. J. Chem. 2007, 81,
1869–1877.
(14) Bernstein, J.; Losee, K. A. U.S. patent 2713049, 1955.
(15) Abramovitch, R. A.; Deeb, A.; Kishore, D.; Mpango, G. B. W.;
Shinkai, I. Gaz. Chim. Ital. 1988, 118, 167–171.
7442 J. Org. Chem. Vol. 74, No. 19, 2009

Jankowiak and Kaszynski
JOCArticle
SCHEME 3
SCHEME 4
SCHEME 5
mercaptan with n-PrI and comparison of the NMR spectrum
for the propylation product 7 with that of 4e (Scheme 4). The
thiolation reaction of 4h was not reproducible and purifica-
tion of the thione was often difficult. Consequently, the
preparation of ester 1h was conducted on a crude, about
60% pure, thione 3h.
SCHEME 6
All sulfur-containing compounds exhibited sensitivity to
white light. Therefore, transformations of oxides 4c-h to the
corresponding esters 1 were conducted with careful exclusion
of ambient light, and reaction workup and product isolation
were performed in the presence of red light. For instance,
attempts to prepare thione 3g from N-oxide 4g in ambient
light (fluorescent tube) did not give the expected product;
instead a mixture of other products was observed by NMR.
N-Oxides 4b-d and 4f were prepared by oxidation of the
corresponding 2-chloropyridines 8b-d or bromopyridine 8f
with m-chloroperbenzoic acid (mCPBA, Schemes 1 and 2),
while the methyl ester 4g[Br] was prepared by oxidation of
the corresponding ester 8g[Br] with peracetic acid
(Scheme 3). The required ester was obtained from 2-bromoi-
sonicotic acid (9[Br]) prepared form 2-amino-4-methylpyr-
idine according to a literature procedure.16 The chloro ester
4g[Cl] was obtained by treatment of 2-chloroisonicotinic
acid N-oxide (10) with diazomethane (Scheme 3). The iso-
propyl ester N-oxide 6 was prepared from ester 11 according
to a modified patent procedure,14 using mCPBA instead of
peracetic acid. The acid 1017 was prepared by oxidation of
9[Cl] with AcOOH (Scheme 3).
The original13 one-step preparation of 2-chloro-4-hepty-
loxy N-oxide (4b) from 2-chloro-4-nitropyridine N-oxide (12)
turned out to be impractical in larger scales. Therefore,
we improved the two-step process in which 2-chloro-
4-nitropyridine18 (13) was alkoxylated with sodium heptano-
late in dry THF to give pyridine 8b, and subsequently oxidized
with mCPBA to 4b in an overall yield of 65% (Schemes 5
and 1). The use of AcOOH as the oxidant led to exces-
sive decomposition of pyridine 8b (presumably due to
cleavage of the R-O bond), which was avoided by using
mCPBA.
2-Chloro-4-propylsulfanylpyridine N-oxide (4e) was pre-
pared directly from nitro derivative 12 according to a
literature procedure described for the benzyl analogue.19
Thus, treatment of 12 with n-PrSH in ethanol in the presence
of AcONa gave 4e as the major product accompanied by
small amounts of the 2-propylsulfanyl derivative 14 (Scheme 6).
Experimental Electronic Absorption Spectra. Experimen-
tal absorption spectra for the purified esters 1a-h were
recorded in MeCN and cyclohexane, and results are shown
in Table 1 and Figures 1-3.20 With the exception of 1e, all
esters exhibit three absorption bands above 200 nm which
can be described as π f π* transitions. The absorption band
for the n f π* transition, expected above 400 nm, was not
observed even for high concentration solutions (10-3 M of 1e).
The observed lowest energy band exhibits significant
substituent dependence. For the electron-donating groups,
X, the maximum of absorption, λmax, is blue-shifted and for
the heptyloxy group the shift is 31 nm in 1b relative to the
parent 1a (Figure 1). Electron-withdrawing groups, X, de-
crease the excitation energy, and the λmax for the CN
derivative 1f is red-shifted by 51 nm relative to 1a. Analysis
demonstrated that the observed excitation energies Eex for
this absorption band in series 1 generally give poor correla-
tion with substituent constants, which is consistent with pre-
vious observations for other series of compounds.21 The best
(19) Bouillon, C.; Kalopissis, G.; Lang, G. U.S. patent 3,862,305, 1975.
(20) For details see the Supporting Information.
ꢀ
(16) Bonnet, V.; Mongin, F.; Trecourt, F.; Queguiner, G.; Knochel, P.
Tetrahedron 2002, 58, 4429–4438.
(17) Puszko, A.; Talik, Z. Pol. J. Chem. 1992, 66, 1427–1430. Puszko, A.
Pol. J. Chem. 1993, 67, 837–847.
(18) Talik, Z.; Talik, T. Rocz. Chem. 1962, 36, 417–422.
ꢀ
(21) (a) Shorter J. Correlation analysis in Organic Chemistry; Clarendon
Press: Oxford, UK, 1973; pp 53-55 and references cited therein. (b) Ford, G. P.;
Katritzky, A. R.; Topsom, R. D. In Correlation Analysis in Chemistry: Recent
Advances; Chapman, N. B., Shorter, J., Eds.; Plenum Press: New York, 1978; pp
298-299.
J. Org. Chem. Vol. 74, No. 19, 2009 7443

Jankowiak and Kaszynski
JOCArticle
FIGURE 1. Electronic absorption spectra of 1a, 1b, and 1f recor-
ded in MeCN.
FIGURE 2. Correlation of the relative excitation energy (ΔEex
=
:
-
EH - Ex) of 1 in MeCN and the substituent parameter σp
(excluding 1b) F = 0.41 ( 0.03, r2 = 0.95.
correlation of the relative Eex values (ΔEex) in MeCN is ob-
tained with the σp- substituent constant,22 if the data point for
1b is excluded (Figure 2). Trends in other series of substituent
constants such as σp, σpþ, and σR0 (from 19F NMR)21 do not
parallel those observed for Eex in series 1 in MeCN and the
corresponding correlations are of significantly lower quality.
Excitation energies obtained for 1in cyclohexane correlate well
with the σR0 parameter,21 if the data point for 1h is excluded
(Eex =3.21 - 1.03σR0 eV, r2 =0.97).
The middle absorption band for series 1 is located at about
290 nm, and its position is practically independent of the
substituent. The only exception is the 4-propylsulfanyl group
for which the band in 1e has an unusually large blue shift of
6 nm relative to the parent 1a and its intensity is over double
relative to other esters in the series.20 Considering that the
high-energy band at about 210 nm is largely absent in 1e, the
observed band at 284 nm apparently is a combination or
superposition of the two high-energy bands.
Analysis of the substituent effect on the K-band of en-
ones24 supports this suggestion. Alkyl and chlorine in the
β position exert a small bathochromic shift of þ12 nm on the
π f π* transition band of the enone,24 while for derivatives
1c and 1h, enthiones, the shift is þ8.5 and þ19 nm, respec-
tively, relative to the parent 1a in MeCN. For the OMe group
the effect is þ30 nm in enones, while þ27 nm is observed in
1b. The large increment of þ85 nm, reported for the SR
group in the β position of an enone,24 predicts the K-band for
1e at 293 nm, while the unusual peak in this derivative is
observed at 284 nm.
A comparison of spectra recorded for 1 in MeCN with
those obtained in cyclohexane demonstrates a modest nega-
tive solvatochromic effect for the two lowest energy absorp-
tion bands (Figure 3). The polar solvent exerts a blue shift
ranging from 4 nm for 1f to 16 nm for 1a for the lowest energy
absorption band, and about 6 nm for the band near
290 nm. The only exception is 1e for which the complex
absorption band at 281.5 nm with visible shoulder features in
cyclohexane is red-shifted by 2.5 nm in MeCN. A similar sol-
vatochromic effect was reported for the N-trimethylacetoxy
FIGURE 3. Electronic absorption spectra for 1g in MeCN (blue)
and cyclohexane (black).
analogue of 1a.25 Spectra of several compounds in cyclohex-
ane exhibit vibronic structures for the lowest energy transition.
Photochemical Stability. The substituent effect on the
stability of the Barton esters was assessed by exposing
samples of 1a, 1b, and 1f in dry acetone-d6 to a laboratory
standard fluorescent light. NMR analysis demonstrated that
after 45 min ∼95% of the cyano derivative 1f was trans-
formed to a major product. During the same time, the parent
ester, 1a, decomposed to the extent of 27%, while the
decomposition of the heptyloxy derivative, 1b, was at a level
of about 5%. After another 60 min, all 1f was completely
decomposed, the parent 1a was transformed by about 55%,
and the heptyloxy, 1b, was decomposed to the extent of 12%.
After 16 h all samples were completely decomposed giving a
single product (1b), two products in 2:1 ratio (1f), or an
equimolar mixture of two products (1a). In ambient daylight
the decomposition process is faster. The products of photo-
decomposition were not isolated, but the major component
is assumed to be the expected23 4-substituted 2-(adamant-
1-ylsulfanyl)pyridine.
(22) Hansch, C.; Leo, A.; Taft, R. W. Chem. Rev. 1991, 91, 165–195.
(23) Barton, D. H. R.; Crich, D.; Motherwell, W. B. Tetrahedron 1985, 41,
3901–3924.
(24) Scott, A. I. Interpretation of Ultraviolet Spectra of Natural Products;
Prergamon Press: New York, 1964; p 58.
(25) Aveline, B. M.; Kochevar, I. E.; Redmond, R. W. J. Am. Chem. Soc.
1995, 117, 9699–9708.
7444 J. Org. Chem. Vol. 74, No. 19, 2009

Jankowiak and Kaszynski
JOCArticle
The observed order of the rate of photodecomposition for
the three derivatives (1f>1a>1b) is consistent with the rela-
tive intensity of absorption in the visible range (>400 nm) in
the series (see Figure 1).
Computational Analysis. To gain a better understanding of
the nature of the electronic excitations and the substituent
effect on their energies in series 1, we conducted TD-DFT
and ZINDO calculations for the closely related N-acetates 2
at the equilibrium geometries obtained at the B3LYP/
6-311G(d,p) level of theory. TD-DFT vertical excitation
energies were calculated by using the B3LYP functional in
combination with either 6-311G(d,p) or cc-pVDZ basis sets.
Geometry optimizations demonstrated that the N-acetoxy
group is nearly orthogonal to the pyridine ring as a result of
repulsion of electron pairs on the adjacent N and O atoms.
This is consistent with the solid-state structure for 1-(4-tert-
butylcyclohexanecarbonyloxy)pyridine-2(1H)-thione26 and
results of our recent calculations for another series of pyri-
dine-2(1H)-thione derivatives.27 Most substituents in the 4
position are coplanar with the pyridine ring (Figure 4). The
only exception is the acetoxy group in 2l, which forms a
dihedral angle (defined by Cring-Cring-O-C(dO)) of 32°
with the pyridinethione ring. The CH3 and CF3 groups in 2c
and 2d adopt the eclipsed conformation.
Both TD-DFT and ZINDO//DFT calculations of series 2
reproduced the general trends observed in series 1, including
the substituent impact on the energies of the main absorption
bands. The results showed that the low-energy electronic
absorptions involve two highest occupied and two lowest
energy virtual MOs. The DFT results demonstrated that the
HOMO orbital is localized on the sulfur atom and contains
the lone pair (nCS). The remaining 3 orbitals are of π type
MOs and generally involve mainly the thiono group
(HOMO-1, πCS and LUMOþ1, π*CS) or the ring (LUMO,
π*ring, Figure 5). The substituents with strong electron-
donating or -accepting properties significantly perturb the
pyridinethione MO system. Thus, the NMe2 in 2k switches
order of the nCS and πCS, and the latter becomes the HOMO.
Similarly, for the strongly electron-withdrawing NO2 in 2n
the π*CS and π*ring MOs are altered and both have density
extending from the ring to the substituent.
FIGURE 4. B3LYP/6-311G(d,p) optimized geometry of 2g.
f e 1 ꢀ 10-4), the πCS f π*ring (λ=340 to 475 nm, f ≈ 5 ꢀ
10-2), the πCS f π*CS (λ ≈ 270 nm, f ≈ 0.2), and the πring f
π*ring (λ < 250 nm). In addition, all esters exhibit CT
excitations from the CdS group to the CdO of the N-acetoxy,
nCS f π*CO and πCS f π*CO (f ≈ 1 ꢀ 10-2), at about 280 nm.
The latter transition combines with the πCS f π*CS transi-
tion, and it constitutes a significant component of this
electronic excitation in several esters.
The nCS f π*ring transition is very weak and has not been
observed experimentally. The computed energies and their
substituent dependence for the remaining three main transi-
tions in 2 are in general agreement with the experimental
results obtained for 1. Thus, the computational results
reproduced the experimentally observed strong substitent
dependence for the πCS f π*ring transition and weak depen-
dence for the πCS f π*CS transition. Theoretical results for
the former transition correlate well with the experimental
data in cyclohexane except for the data points for the
derivatives with strongly electron-withdrawing substituents
CN and COOMe (2f and 2g). The correlation factor r2=0.98
for the cyclohexane data is higher than that for the MeCN
data (r2 = 0.96, Figure 6), presumably due to the solvent
effect that is not included in the TD-DFT model.
The calculated πring f π*ring excitations are less reliable. They
show, however, that certain substituents such as SMe, NMe2,
and CtCH exhibit strong electronic coupling with the ring,
which results in significant decrease of the excitation energy and
increase of the oscillator strength. The largest red shift for the
πring f π*ring excitations is calculated for the SMe derivative
(Eex = 259 nm and f ≈ 0.43), which is consistent with the
experimentally observed apparent merger of the πring f π*ring
and πCS f π*CS transitions at λmax =281.5 nm in cyclohexane
and at λmax =284 nm in MeCN for 1e (Table 1, vide supra).
The ZINDO calculations are generally in agreement with
the TD-DFT results. However, the ZINDO calculated en-
ergies for the nCS f π*ring and πCS f π*ring excitations have a
Substituents with lone pairs (OMe, SMe, Cl, NMe2, OAc)
and high-lying π electrons (CtCH) interact with the π
system of the pyridinethione ring. Consequently, the result-
ing πring orbital has density extending into the substituent
and becomes the HOMO-2, as shown for 2i in Figure 5. In
these esters, the HOMO-3 (nCO) and LUMOþ2 (π*CO) are
localized on the N-acetoxy group (Figure 5).
TABLE 1. Absorption Maxima for Esters 1 in MeCN
The two sets of TD-DFT results are similar. The cc-pVDZ
basis set computes the nCS f π*ring and πCS f π*ring
transitions at slightly lower energy (-0.06 and -0.03 eV,
or 9.5 and 3 nm, respectively) and the πCS f π*CS transition
at slightly higher energy (0.08 eV or -5 nm) than those
obtained with the 6-311G(d,p) basis set.
The TD-B3LYP/cc-pVDZ//B3LYP/6-311G(d,p) results
revealed that the electronic excitations are located in four
areas of the spectrum. On the basis of the major transitions,
these excitations can be classified as the nCS fπ*ring (λ>380 nm,
1
X
λ
max (log ε) (nm)
λmax (log ε) (nm) λmax (log ε) (nm)
a H a
b OC7H15 235 (4.36)
208 (4.05)
290 (4.11)
293 (4.11)
292.5 (4.15)
292.5 (4.16)
284 (4.45)
294.5 (4.12)
292 (4.09)
295 (4.17)
364 (3.74)
333 (3.81)
357.5 (3.80)
386 (3.73)
365 (3.75)
415 (3.69)
409 (3.68)
369 (3.73)
b
c Me
216.5 (4.22)
208 (4.1)
d CF3
e SC3H7
f CN
g COOMe 216(4.20)
h Cl 227 (4.22)
216 (4.1) and 230 (4.1)b
(26) Hartung, J.; Hiller, M.; Schwarz, M.; Svoboda, I.; Fuess, H. Liebigs
Ann. 1996, 2091–2097.
(27) Kaszynski, P. Phosphorus, Sulfur, Silicon Relat. Elem. 2009, 184,
1296–1306.
aReference 23: λmax (log ε) 364 (3.64) and 286 (4.15) in EtOH. See
Figure 1.
J. Org. Chem. Vol. 74, No. 19, 2009 7445

Jankowiak and Kaszynski
JOCArticle
FIGURE 5. B3LYP/cc-pVDZ generated contours of molecular orbitals for 2i relevant to the low-energy transitions.
FIGURE 7. Substituent effect on MO energy: B3LYP/cc-pVDZ
derived energy levels for 2a (black), 2f (red), and 2i (blue).
FIGURE 6. Correlation of the experimental (for 1 in MeCN) and
theoretical (for 2) excitation energies. Best fit lines: ZINDO (red
circles, all data): E(theor)=2.01 þ 0.372E(exp), r2=0.95; TD-B3LYP/
cc-pVDZ (black diamonds, excluding g and f): E(theor) = 1.55 þ
0.553E(exp), r2 =0.96.
The origin of the substituent effect on absorption spectra
was investigated by using the DFT results for esters 2a, 2f,
and 2i in Figure 7. Analysis shows that the OMe substituent
destabilizes all four orbitals with the largest effect on the
LUMO (þ0.49 eV). Conversely, the CN substituent stabi-
lizes all four orbitals, with the largest effect exerted again on
the LUMO (-1.02 eV). The remaining three orbitals have
lower energies by a similar value of about 0.55 eV relative to
2a. Thus, transitions involving excitation to the LUMO,
which is designated as π*ring, are expected to exhibit the
strongest substitent effect. Indeed, both low-energy transi-
tions, nCS f π*ring and πCS f π*ring, involve the LUMO
(π*ring) and both show strong dependence on the substituent,
while the band at about 290 nm involves the HOMO-
-LUMOþ1 transition (πCS f π*CS), which renders it
practically substituent independent. Consequently, the ob-
served trends in nCS f π*ring and πCS f π*ring excitation
energies follow those of the LUMO energy. Also, the calcu-
more narrow range, the order of nCS and πCS is reversed, and
the low-intensity excitations to the N-acetoxy π*CO are
absent. On the other hand, the ZINDO results give more
realistic values of the oscillator strength, f ≈ 0.2, for the two
observed excitations πCS f π*ring and πCS f π*CS, and
appropriate energy for the latter transition at about 300 nm.
In addition, the ZINDO calculated excitation energies cor-
relate relatively well with all the experimental data points
(Figure 6). The quality of the correlation for data obtained in
MeCN and cyclohexane is practically identical.
Analysis of the ZINDO results for 2 demonstrated that the
excitation energies calculated in MeCN dielectric medium
are generally blue-shifted by an average of 28 (nCS f π*ring),
7 (πCS f π*ring), and 8 nm (πCS f π*CO). This is consis-
tent with the experimentally observed negative solvato-
chromic effect in series 1 and an average blue shift of 9 and
5 nm for the two π f π* transitions. The only exception
is the πCS f π*ring transition for 2n (X=NO2) and the πCS f
π*CS transition for 2k (X = NMe2), for which a red shift
of 1 and 16 nm, respectively, is calculated in the polar
medium.
-
lated nCS f π*ring excitation energies correlate with the σp
parameter as shown in Figure 8.
Discussion and Conclusions
Experimental photophysical studies revealed that the
position of the observed lowest energy absorption band is
7446 J. Org. Chem. Vol. 74, No. 19, 2009

Jankowiak and Kaszynski
JOCArticle
Correlation between the experimental excitation energies
-
and the σp substituent parameter in Figure 2 permits an
estimate of the λmax for unknown derivatives of 1. Thus, λmax
for 1n (X=NO2) is predicted at 430 nm in MeCN. A value of
413 nm is predicted for the same derivative 1n on the basis of
ZINDO results and the correlation in Figure 6. However,
this value seems too low considering that λmax for 1f (X =
-
CN) is 415 nm. Derivative 1 with the MeSO2 group (σp
=
1.13)22 in the 4 position has a predicted λmax value of 421 nm,
and the CF3SO2 group (σp- = 1.63),22 which is one of the
most electron-withdrawing substituent, shifts the πCS
π*ring absorption to 452 nm, according to Figure 2.
f
The presented results demonstrate synthetic methods for
the preparation of substituted 1-hydroxypyridine-2(1H)-
thiones, offer basic understanding of the photophysical
processes, and provide a predictive tool for designing
N-acyloxy derivatives with desired light sensitivity.
FIGURE 8. Correlation of the excitation energy Eexc (B3LYP/
cc-pVDZ) for the nCS f π*ring transition in 2 and the substituent
parameter σp-. Eexc = 2.98 - 0.66σp-, r2 = 0.92.
Computational Details
Quantum-mechanical calculations were carried out with
the Gaussian 98 suite of programs.28 Geometry optimizations
were undertaken by using the B3LYP functional29,30 with the
6-311G(d,p) basis set,31 default convergence limits, and without
symmetry constraints. Vibrational frequencies were used to
characterize the nature of the stationary points. Vertical elec-
tronic excitation energies for esters 2 were obtained by employ-
ing the TD-DFT method32 and using the B3LYP functional29,30
with the 6-311G(d,p)31 or cc-pVDZ33 basis sets. Excitation
energies for 2 were also obtained at their DFT determined
geometries by using the INDO/2 algorithm (ZINDO)34 as
supplied in the Cerius2 suite of programs and including all
electrons and orbitals in the CI. ZINDO calculations with the
solvation model (Self Consistent Reaction Field) used cavity
radii derived from the DFT calculations with the VOLUME
keyword.
sensitive to the substituent in the 4-position and varies from
333 nm for RO to 415 nm for CN. As a consequence, the
esters have differential sensitivity to light: the cyano deriva-
tive 1f undergoes photoinduced transformation about
20 times faster than the alkoxy derivative 1b under the
same conditions. The position of the higher energy absorp-
tion band at about 295 nm is practically substituent inde-
pendent. Both bands exhibit a modest negative solvato-
chromic effect.
Analysis of the TD-DFT and ZINDO results revealed the
electronic structure and the nature of the observed electronic
excitations in series 1. In general, our results are consistent
with those of previous computational studies of nitro and
fluoro derivatives of 1-methoxypyridine-2(1H)-thione.12
They demonstrate that the substituent in the 4-position
strongly affects the energy level of the LUMO (π*ring) and
consequently the position of the absorption maxima of two
lowest energy transitions: the unobserved nCS f π*ring and
the observed πCS f π*ring. Out of the two transitions, the
latter appears to have the dominant effect on the photo-
Experimental Section
Melting points are uncorrected. 1H NMR and 13C NMR
spectra were recorded at 300 and 75 MHz, respectively, in
CDCl3, unless specified otherwise. Chemical shifts were refer-
enced to the solvent (CHCl3 set at 7.26 and 77.0 ppm). UV
spectra were recorded in CH3CN and cyclohexane. Molar
stability of the ester. The experimentally unobserved nCS
f
π*ring transition has a significantly lower absorption coeffi-
cient and is predicted to appear >390 nm for all esters in
series 1, which suggests similar weak sensitivity to visible
light for all members of the series. Further analysis of the
series 2 revealed that the 4-RO substituent exerts the nearly
maximum effect on increasing the excitation energy (largest
blue shift) and hence provides the maximum photostabiliza-
tion for esters in series 1. A similar position of the absorption
maximum and consequently photostability can be expected
for the 4-NMe2 derivative 1k.
Computational results are moderately useful as a tool for
the prediction of excitation energies. The TD-DFT method
has a problem with good reproduction of energies for the
electron-deficient species such as 1f and 1g (Figure 6) and the
magnitude of the oscillator strength. In contrast, the ZIN-
DO//DFT method performs better and provides a better
correlation with all experimental results, despite a more
narrow range of computed excitation energies (the slope of
the correlation line is 0.37). The best predictive tool, how-
ever, appears to be the Hammett-type correlation shown in
Figure 2.
(28) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
M. A.; Cheeseman, J. R.; Zakrzewski, V. G.; Montgomery, J. A., Jr.;
Stratmann, R. E.; Burant, J. C.; Dapprich, S.; Millam, J. M.; Daniels, A.
D.; Kudin, K. N.; Strain, M. C.; Farkas, O.; Tomasi, J.; Barone, V.; Cossi,
M.; Cammi, R.; Mennucci, B.; Pomelli, C.; Adamo, C.; Clifford, S.;
Ochterski, J.; Petersson, G. A.; Ayala, P. Y.; Cui, Q.; Morokuma, K.;
Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.;
Cioslowski, J.; Ortiz, J. V.; Baboul, A. G.; Stefanov, B. B.; Liu, G.;
Liashenko, A.; Piskorz, P.; Komaromi, I.; Gomperts, R.; Martin, R. L.;
Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.;
Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen, W.; Wong, M. W.;
Andres, J. L.; Gonzalez, C.; Head-Gordon, M.; Replogle, E. S.; Pople, J. A.
Gaussian 98, Revision A.9; Gaussian, Inc., Pittsburgh, PA, 1998.
(29) Becke, A. D. J. Chem. Phys. 1993, 98, 5648–5652.
(30) Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B 1988, 37, 785–789.
(31) (a) McLean, A. D.; Chandler, G. S. J. Chem. Phys. 1980, 72, 5639–
5648. (b) Krishnan, R.; Binkley, J. S.; Seeger, R.; Pople, J. A. J. Chem. Phys.
1980, 72, 650–654.
(32) (a) Bauernschmitt, R.; Ahlrichs, R. Chem. Phys. Lett. 1996, 256, 454–
464. (b) Stratmann, R. E.; Scuseria, G. E.; Frisch, M. J. J. Chem. Phys. 1998,
109, 8218–8224.
(33) (a) Woon, D. E.; Dunning, T. H. Jr. J. Chem. Phys. 1993, 98, 1358–
1371. (b) Dunning, T. H. Jr. J. Chem. Phys. 1989, 90, 1007–1023.
(34) Zerner, M. C. In Reviews of Computational Chemistry; Lipkowitz,
K. B. Boyd, D. B., Eds.; VCH Publishing: New York, 1991; Vol. 2,
pp 313-366 and references cited therein.
J. Org. Chem. Vol. 74, No. 19, 2009 7447

Jankowiak and Kaszynski
JOCArticle
extinction coefficients were obtained by fitting maximum ab-
sorbance against concentration in agreement with Beer’s law.
The synthesis, isolation, and handling of 1c-h and all its
sulfur-containing intermediates were performed in the presence
of red light only!
Adamantane-1-carboxylate Esters 1a-e: General Procedure.
According to a literature procedure,13 adamantane-1-carbonyl
chloride (1.1 mmol) was added to a suspension of crude sodium
salt 3[Na] (prepared from 1.0 mmol of N-oxide 4) in dry MeCN
(2 mL) and the reaction mixture was stirred overnight. The
solvent was evaporated and the residue was passed trough a
silica gel plug with use of CH2Cl2 as the eluent. The resulting
product was purified by repeated recrystallization from MeCN
with a few drops of AcOEt.
1-(Adamantane-1-carbonyloxy)-4-trifluoromethylpyridine-2(1H)-
thione (1d). Crude product was recrystallized from MeCN to give 1d
in 50% yield as a yellow solid: mp 156-158 °C; 1H NMR (CDCl3)
δ 1.78 (br s, 6H), 2.09-2.14 (m, 3H), 2.16 (br s, 6H), 6.66 (dd, J1=
7.3 Hz, J2=2.3 Hz, 1H), 7.55 (d, J=7.2 Hz, 1H), 7.86 (br s, 1H);
13C NMR (CDCl3) δ 27.5, 36.0, 38.5, 41.1, 107.4, 121.9 (q, J =
275 Hz), 134.1 (q, J = 35 Hz), 134.5 (q, J = 4.4 Hz), 139.1, 172.4,
175.4. Anal. Calcd for C17H18F3NO2S: C, 57.13; H, 5.08; N, 3.92.
Found: C, 57.05; H, 4.95; N, 4.02.
Analytical data for other compounds 3 are provided in the
Supporting Information.
2-Chloropyridine N-Oxides 4b-d and 4f: General Procedure.
A solution of appropriate 2-chloropyridine 7 (1.0 mmol) and
m-chloroperoxybenzoic acid (mCPBA, 2.0 mmol) in dry CH2-
Cl2 (2 mL) was stirred at rt for 48 h. A saturated solution of
Na2CO3 was added, the mixture was stirred for 0.5 h and
extracted with CH2Cl2, the extracts were dried (Na2SO4), and
solvent was evaporated. The residue was purified on a silica
gel plug (CH2Cl2 then acetone) to give N-oxide 4 as a colorless
oil.
2-Chloro-4-trifluoromethylpyridine N-Oxide (4d). Yield 37%;
mp 42-44 °C; 1H NMR (CDCl3) δ 7.43 (dd, J1 =6.8 Hz, J2 =
2.1 Hz, 1H), 7.75 (d, J=2.0 Hz, 1H), 8.40 (d, J=6.8 Hz, 1H);
HRMS calcd for C6H4ClF3NO m/z 197.9934, found m/z
197.9917.
Analytical data for other compounds 4 are provided in the
Supporting Information.
Acknowledgment. Financial support for this work was
received from the National Science Foundation (DMR-
0606317). We thank Mr. Jakub Jakowiecki, a visiting student
from the laboratory of Prof. M. Makosza (IchO, Warsaw,
Poland), for his help with the preparation of 2-bromoisoni-
cotinic acid 9[Br].
Analytical data for other compounds 1 are provided in the
Supporting Information.
1-Hydroxypyridine-2(1H)-thiones 3b-e: General Procedure.
Crude 1-hydroxypyridine-2(1H)-thione sodium salt 3[Na] was
treated with dilute HCl and the product was extracted with
CH2Cl2. The solvent was evaporated to leave thione 3 as a yello-
wish viscous oil, with purity >90%, based on NMR analysis.
1-Hydroxy-4-trifluoromethylpyridine-2(1H)-thione (3d). 1H
NMR (CDCl3) δ 6.89 (dd, J1 = 7.1 Hz, J2 = 2.2 Hz, 1H), 7.91
(d, J = 2.0 Hz, 1H), 8.19 (d, J = 7.1 Hz, 1H); HRMS calcd for
C6H5F3NOS m/z 196.0044, found m/z 196.0062.
Supporting Information Available: Synthetic details and
characterization data for all compounds, details of photostabil-
ity experiment, tabularized NMR spectra for 1, 3, 4, and 5,
tabularized results of TD-DFT and ZINDO calculations, ar-
chive of calculated equilibrium geometries for 2, and selected 1H
and 13C NMR spectra. This material is available free of charge
via the Internet at http://pubs.acs.org.
7448 J. Org. Chem. Vol. 74, No. 19, 2009