Thiol Reactivity toward Atomic Oxygen Generated during the Photodeoxygenation of Dibenzothiophene S-Oxide

Article abstract of DOI:10.1021/acs.joc.7b02428

Aromatic heterocyclic oxides, such as dibenzothiophene S-oxide (DBTO), have been suggested to release ground state atomic oxygen [O(³P)] upon irradiation, and as such, they have been used to create a condensed phase reactivity profile for O(³P). However, thiols, which are highly reactive with O(³P) in the gas phase, were not previously investigated. An earlier study of O(³P) with proteins in solution indicated a preference for thiols. A further investigation of the apparent thiophilicity provided the subject for this study. DBTO was employed as a putative O(³P)-precursor. However, the effective rate of O(³P) formation was found to be dependent on reactant concentrations in certain cases. All reactants were found to increase the rate of deoxygenation to some extent, but in the presence of reactants containing an alcohol linked to a reactive functional group, deoxygenation occurred substantially more rapidly. The rate enhancement was quantified and attributed to the reaction of activated O atom within the solvent cage prior to escape into the bulk solution. Through competition experiments, the relative rate constants of O(³P) with thiols and other functional groups were found. A small preference for primary thiols was observed over other thiols, sulfides, and alkenes. A much larger preference was observed for thiols, sulfides, and alkenes over aromatic groups. In summary, DBTO was successfully used as an O(³P)-precursor, and the thiophilicity of O(³P) was confirmed and quantified.

Full text of DOI:10.1021/acs.joc.7b02428

Subscriber access provided by UNIVERSITY OF TOLEDO LIBRARIES
Article
Thiol reactivity towards atomic oxygen generated during
the photodeoxygenation of dibenzothiophene S-oxide
Sara M. Omlid, Miao Zhang, Ankita Isor, and Ryan D. McCulla
J. Org. Chem., Just Accepted Manuscript • Publication Date (Web): 27 Nov 2017
Downloaded from http://pubs.acs.org on November 28, 2017
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
The Journal of Organic Chemistry is published by the American Chemical Society.
1155 Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 10
The Journal of Organic Chemistry
1
2
3
4
5
6
Thiol reactivity towards atomic oxygen generated during the photo-
deoxygenation of dibenzothiophene S-oxide
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Sara M. Omlid, Miao Zhang, Ankita Isor, and Ryan D. McCulla*
Department of Chemistry, Saint Louis University, 3501 Laclede Ave. St. Louis, MO 63108
KEYWORDS. Reactive oxygen species, bimolecular photoreduction, aromatic sulfoxides, sulfoxide-alcohol hydrogen bonds,
2
-mercaptoethanol, 1-propanethiol.
ABSTRACT: Aromatic heterocyclic oxides, such as dibenzothiophene Sꢀoxide (DBTO), have been suggested to release ground
3
state atomic oxygen [O( P)] upon irradiation, and as such, they have been used to create a condensed phase reactivity profile for
3
3
O( P). However, thiols, which are highly reactive for O( P) in the gas phase, were not previously investigated. An earlier study of
3
O( P) with proteins in solution indicated a preference for thiols. A further investigation of the apparent thiophilicity provided the
3
3
subject for this study. DBTO was employed as a putative O( P)ꢀprecursor. However, the effective rate of O( P) formation was
found to be dependent on reactant concentrations in certain cases. All reactants were found to increase the rate of deoxygenation to
some extent, but in the presence of reactants containing an alcohol linked to a reactive functional group, deoxygenation occurred
substantially more rapidly. The rate enhancement was quantified and attributed to the reaction of activated Oꢀatom within the solꢀ
3
vent cage prior to escape into the bulk solution. Through competition experiments, the relative rate constants of O( P) with thiols
and other functional groups were found. A small preference for primary thiols was observed over other thiols, sulfides, and alkenes.
A much larger preference was observed for thiols, sulfides, and alkenes over aromatic groups. In summary, DBTO was successfully
3
3
used as an O( P)ꢀprecursor, and the thiophilicity of O( P) was confirmed and quantified.
3
ence by O( P) for the thiol containing residue, cysteine, and
INTRODUCTION
Due to its role in combustion chemistry and production in
the upper atmosphere, atomic oxygen, O( P), is an important
gas phase reactive oxygen species (ROS). However, little is
known of the condensed phase chemistry, primarily because
3
more significantly, O( P) oxidation was shown to specifically
11
target the regulatory cysteine residues. Given the conceivable
3
applications, it is important to understand the fundamental
1
,2
3
nature of the reactivity of O( P) with thiols.
Dibenzothiophene Sꢀoxide (DBTO) is the most commonly
3
clean and mild methods to produce O( P) in solution have
3
3
2–5
used O( P)ꢀprecursor for the study of O( P) in solution. UV
irradiation of DBTO is posited to undergo unimolecular S—O
3
been limited by the high energies associated with O( P) generꢀ
3
ation. Some heterocyclic oxides are reported to deoxygenate
bond cleavage through a dissociative triplet (T ) state, leading
2
upon UV irradiation to give the parent heterocycle and the
to dibenzothiophene (DBT) and an oxidant, which is observed
3
2,4–8
presumptive O( P),
and in this way, heterocyclic oxides
2,12
indirectly in the form of secondary reaction products. While
3
have been used as photoactivatable O( P)ꢀprecursors.
3
most evidence indicates O( P) as the putative oxidant, definiꢀ
A few studies have employed heterocyclic oxides as
tive identification of the oxidizing species has not yet been
3
3
3
O( P)ꢀprecursors to address whether O( P) has a reactivity in
solution similar to its reactivity in the gas phase. In the gas
achieved. Since direct spectroscopic detection of O( P) is not
3
feasible in solution, the presumptive assignment of O( P) is
3
phase, O( P) is highly reactive, yet, surprisingly selective for
based on indirect evidence involving an observed reactivity
3
such a highly energetic oxidant; this is illustrated with its
strong preference for alkenes, sulfides, and thiols (rate conꢀ
consistent with expectations for O( P) and other energetic
2
considerations. However, distinguishing between freely difꢀ
1
2
13
3
ꢀ1 ꢀ1
3
stants, k = 10 to 10 cm mol s ) over other functional
fusing O( P) and a viable “oxenoid” alternative, such as a nonꢀ
10
3
ꢀ1
groups such as alkanes, ethers, or alcohols (k = 10 cm mol
In solution, the reactivity of O( P) with a variety of
functional groups has been investigated.
covalent DBTꢀO atom complex or some DBTO excited state,
ꢀ1 9,10
3
2,6
s ).
has not been conclusively resolved.
2,4,5,8
Missing from
3
In this work, we sought to gauge the reactivity of O( P) in
these studies are thiols, which represent one of the most reacꢀ
solution with thiols relative to other functional groups. To that
end, we employed competition experiments to measure the
3
tive functional groups with O( P) in the gas phase.
3
3
In recent years, a waterꢀsoluble heterocyclic oxide O( P)ꢀ
relative reaction rates with O( P) and various reactants, which
3
3
precursor was used to probe the reactivity of O( P) with a parꢀ
allowed us to construct a more complete O( P) reactivity proꢀ
ticular test protein to gauge its selectivity, if any, for particular
file in solution. Furthermore, we sought to better understand
the photodeoxygenation process of DBTO and to establish
guidelines for its use in studying the kinetics of O( P) in soluꢀ
11
3
amino acids. Considering the reactivity of O( P) in the gas
phase, selectivity for methionine and cysteine residues was
anticipated. The results of the protein study indicated a preferꢀ
3
tion.
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 2 of 10
RESULTS & DISCUSSION
Products of O( P) Oxidation. An analysis of the oxidation
Thiol (RSH) oxidation resulted in the observed disulfide
3
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
3
(RSSR) product. The presumed reaction mechanism of O( P)
with RSH proceeds by way of a sulfenic acid intermediate
(RSOH), which will react with a second RSH to undergo a
condensation reaction (Scheme 1). In one gas phase study, an
products of eleven reactants was conducted using DBTO as a
3
photoactivatable O( P)ꢀprecursor. Solutions of each reactant
and DBTO were dissolved in acetonitrile, degassed and irradiꢀ
ated with broadly emitting fluorescent bulbs centered at
3
addition mechanism, whereby O( P) adds to the sulfur atom
followed by tautomerization, was shown to account for 90% of
oxidation to RSOH, while a mechanism involving hydrogen
3
50nm (fwhm 325 – 375nm). Product analysis was conducted
by injecting photolyzed solutions on a GCꢀMS, and product
identities were further verified by comparison to retention
times of authentic samples. The results of our analysis are
depicted in Table 1 and more details are provided in the SI.
abstraction followed by recombination accounted for the other
9
1
0%. Conversely, a computational study came to the opposite
1
4
conclusion.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
Scheme 1. Thiol oxidation pathway by O( P).
Table 1. Oxidation products & relative abundances ob-
served upon photolysis of DBTO in the presence of select
reactants.
a
b
No. Reactants
1
Oxidation Products (Product Ratios)
3
In our product analysis (Table 1), an additional O( P) reacꢀ
tive site was observed for phenylmethanethiol, more commonꢀ
ly known as benzyl mercaptan. The benzylic position was
susceptible to oxidation, resulting in the formation of benzalꢀ
dehyde as a minor product in addition to benzyl disulfide. A
plausible mechanism leading to benzaldehyde involves oxidaꢀ
tion of the benzylic carbon to the thiohydrate followed by
desulfurization via release of hydrogen sulfide; however, no
evidence for the mechanism was obtained. Precedence for
2
3
3
O( P) reactivity at the benzylic position is evident in the case
4
5
of benzyl alcohol, which is oxidized by a similar benzylic
15,16
hydrogen abstraction mechanism to give benzaldehyde.
3
The oxidation of alkenes by O( P) have been previously reꢀ
ported to occur via a stepwise addition mechanism to the
3.7
1
1
2,4
epoxide and aldehyde. The oxidation of styrene was previꢀ
4
ously reported. Here we observed similar product ratios to
6
7
those previously reported; although, we did not observe the
very minor product, acetophenone. Styrene oxide and pheꢀ
3
nylacetaldehyde were attributed to oxidation by O( P), while
benzaldehyde is the result of ozone formation, which is
3
4
formed via O( P) oxidation of residual O in solution. The
2
3
2.1
1.3
reaction rate of O( P) and O is competitive with that of alꢀ
kenes, permitting ozone formation even at residual levels of
2
8
dissolved O that could not be completely removed by argon
2
4
8
9
sparging.
3
Similarly, O( P) oxidation of benzene primarily occurs via
2,8
an addition mechanism to form phenol. Biphenyl is formed
as a minor product (<2%), indicating a possible competing
1
.1
1
13
mechanism.
c
c
3
57
1
O( P) was found to oxidize sulfides to sulfoxides. In the
synthesis of sulfoxide via oxidation of sulfide, sulfones are
often an unavoidable byproduct. However, in our studies of
1
1
0
1
3
sulfide oxidation by O( P), sulfone was not detected.
Effect of Reactant Concentration on Photodeoxygena-
tion Quantum Yields. The initially presumed process resultꢀ
ing in the observed oxidations involved the irradiation of
3
DBTO prompting the release of O( P) as the primary oxidant.
a
Based on this presumed mechanism, the quantum yield of
photodeoxygenation (Φ+DBT) was expected to be independent
of the concentration of the reactant molecule. However, it was
observed that Φ+DBT increased when the concentration of the
reactant was increased (Figure 1). Thus, the dependence of
Φ+DBT on reactant concentration was investigated.
Solutions of each reactant (5mM – 2M) and DBTO (10ꢀ
20mM) were photolyzed in acetonitrile using broad spectrum
UVA light (centered at 350nm). Photocontrols (no DBTO) and
thermal controls (no hν) were performed for each reactant. The
b
percent conversion for each reactant to product was less than
1
0%. DBTO deoxygenation was less than 15%. Products were
identified by GCꢀMS and confirmed with authentic sample injecꢀ
c
tions via GCꢀFID. Ref 13.
ACS Paragon Plus Environment

Page 3 of 10
The Journal of Organic Chemistry
0
.010
0.03
1
2
3
4
5
6
7
8
9
(A)
(B)
0
.02
.01
0.006
0
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
0
.002
0.00
0
1000
2000
0
1000
2000
Concentration of Reactant (mM)
Concentration of Reactant (mM)
5
ꢀHexenꢀ1ꢀol
ꢀ(Methy lt h io )ꢀ1ꢀbutanol
6ꢀMercaptoꢀ1ꢀhexanol
3ꢀButenꢀ1ꢀol
Meth io nol
1ꢀButanol
Butyl Sulfide
2
1ꢀOctene
4
1ꢀPropanethiol
Benzene
ꢀMercaptoethanol
2ꢀMercaptoethanol
Figure 1. The effect of reactant concentration on Φ+DBT by irradiating solutions of DBTO and reactant in for 5h with broadly emitting fluoꢀ
rescent bulbs centered at 350nm. Part (A) shows the analysis of some of the reactants used in competition studies. Φ in acetonitrile
neat) was 0.0024, similar to reported values. Benzene is given using previously reported values. Part (B) depicts the analysis of HOꢀ
containing bifunctional reactants. Lines represent linear regression fits. The error in Φ+DBT averaged 0.0003 (at 95% confidence). A greater
error in Φ+DBT was observed in Part (B), especially at higher reactant concentrations (average error at 95% confidence = 0.0008).
+DBT
2
2
(
To find Φ+DBT at varying concentrations of reactant, soluꢀ
tions of DBTO (17mM) and reactant (0ꢀ2M) in acetonitrile
were irradiated with broadly emitting fluorescent bulbs (fwhm
the 1ꢀpentanol concentration was increased, which indicated
that both thiol and alcohol must be in close proximity (or
linked) in order to elicit a similar enhancement in Φ+DBT as
observed in the presence of BME.
3
25 – 375nm) for each reactant shown in Figure 1, (see Experꢀ
imental section for further details). The Φ+DBT in the absence
of reactant (neat acetonitrile) was found to be 0.0024, which is2
within experimental error of the reported Φ+DBT at 320nm.
Additionally, Φ+DBT were measured using monochromatic light
(305nm) in the presence of three select reactants at several
concentrations. No significant difference in Φ+DBT measured
under broadꢀspectrum and monochromatic light was observed.
When DBTO was photolyzed in the presence of longer
chain alcoholꢀthiols or reactants containing an alcohol linked
to an alkene or sulfide, a rate enhancement was also observed
(Figure 1B). We hypothesized that the alcohol moiety of the
reactants forms a hydrogen bond to the sulfoxide of DBTO,
thus increasing the “effective concentration” of the reactant
Figure 2). No considerable deoxygenation rate enhancement
was observed in the presence of 1ꢀpentanol or 1ꢀbutanol,
which could be reasonably explained by the relatively slow
rate of O( P) reactions with alkanes. If free O( P) were solely
responsible for oxidation, then the concentration of the reacꢀ
tant should have little to no effect on the rate of DBT forꢀ
mation (measured as Φ+DBT). Therefore, we deduced that reacꢀ
tants containing a highly O( P) reactive functional group
linked to an alcohol by (CH ) , which will be referred to as
(
As shown in Figure 1A, there was an increase in DBT forꢀ
mation with increasing reactant concentration for all reactants
considered. The observation was consistent with previous reꢀ
ports of a general correlation between solvents containing
functional groups that would be expected to react rapidly with
3
3
3
2,17
O( P) and an increase in Φ+DBT.
For alkene, sulfide, and
3
most alcohol and thiol containing reactants, only a small inꢀ
crease in Φ+DBT was observed. However, in the presence of 2ꢀ
sulfanylethanꢀ1ꢀol, more commonly known as 2ꢀ
mercaptoethanol (BME), a dramatic increase in the Φ+DBT was
observed. The Φ+DBT in BME (neat) was measured and found
to be 0.15 at 305nm, or 15 to 60ꢀfold greater than in any other
solvent previously reported, including cyclohexene (Φ+DBT
.010), which was previously considered to be “the most effiꢀ
cient externally trapping solvent”.
2 n
HOꢀcontaining bifunctional reactants, allow for an additional
oxidation pathway that is also capable of competing with deꢀ
activation pathways.
=
0
2,17
Although BME was found to dramatically increase the rate
of DBTO deoxygenation, photolysis of DBTO in the presence
of both 1ꢀpentanol and 1ꢀpropanethiol resulted in very little
rate enhancement. For example, when DBTO was photolyzed
in the presence of 1ꢀpentanol at a constant concentration (1M)
and 1ꢀpropanethiol with increasing concentrations (from 0 to
1M), Φ+DBT was found to increase at a similar rate as in the
absence of 1ꢀpentanol (see SI, Figure S1). The same result was
observed when 1ꢀpropanethiol was held constant at 1M and
Figure 2. Sulfoxideꢀalcohol hydrogen bonding of DBTO and
HOꢀcontaining bifunctional groups: (A) thiols, n = 1, 2, or 5;
B) alkenes, n = 1 or 3; (C) sulfides, n = 2 or 3.
(
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 4 of 10
The chain length between the alcohol and reactive functionꢀ photoreduction mechanism were operative, BME disulfide
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
al group was found to be consequential (Figure 1B). For thiols,
the optimal chain length was found to be three carbons. For
alkenes, 3ꢀbutenꢀ1ꢀol was found to be an effective deoxygenaꢀ
tion rate enhancer, while the longer chain alkene, 5ꢀhexenꢀ1ꢀ
ol, enhanced the deoxygenation rate only to the extent that
nonꢀalcoholic reactive reactants increased the rate. For sulꢀ
fides, a threeꢀcarbon chain length was slightly more effective
than the fourꢀcarbon.
formation would be expected to result from the BME radical
species formed upon electron or hydrogen atom transfer to
DBTO. However, when a solution of BME and 9,10ꢀ
anthracenedicarbonitrile, a photoꢀinitiated electron acceptor,
dissolved in acetonitrile was photolyzed, no significant change
in BME disulfide concentration was observed (Figure 3B).
2
4
Increased Φ+DBT have been observed with the use of sensiꢀ
23
tizers such as carbazoles. Carbazoles are thought to act as
electron donors in the bimolecular photoreduction of sulfoxꢀ
ides with oxygen leaving as hydroxyl radical and/or hydroxide
Sulfoxide-Alcohol Hydrogen Bonding. Generally, alcoꢀ
holic hydrogen atoms are found much farther upfield when
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
23
dissolved in chloroformꢀd compared to DMSOꢀd , due to sulꢀ
ion. To consider whether BME alters the deoxygenation of
DBTO by similar means or results in a similar oxidant, carbaꢀ
zole was photolyzed in the presence of DBTO and 1ꢀ
propanethiol or BME at a wavelength outside the absorption
spectra for DBTO but within that for carbazole (λ=352nm)
(Figure 3C). At 352nm, deoxygenation of DBTO in the abꢀ
sence of carbazole was insignificant as expected. With carbaꢀ
zole, deoxygenation of DBTO occurred; however, little to no
increase in disulfide was observed compared to the thermal
control for either 1ꢀpropanethiol or BME. This indicated that
even if bimolecular photoreduction was operative, it did not
generate an oxidant that produces disulfide.
6
1
8
foxide alcohol hydrogen bonding. To confirm that hydrogen
bonding occurs between alcoholic reactants and DBTO,
changes in chemical shifts for the hydroxyl proton were measꢀ
1
ured using H NMR. Observed changes in the chemical shifts
with and without DBTO ranged from 4 to 8Hz in acetonitrileꢀ
d3 and 21 to 68 Hz in chloroformꢀd for the alcoholic (OH)
hydrogen atoms, and an average of ꢀ1.5Hz in acetonitrileꢀd3
and ꢀ3Hz in chloroformꢀd for the hydrogen atoms of the αꢀ
1
carbon (additional data can be found in the SI). For control, H
NMR analysis of nonꢀalcoholic reactants with and without
DBTO was performed, and only minor changes in chemical
shifts (less than 1Hz) were observed. The magnitude of change
in chemical shift was highly sensitive to alcohol concentration;
The anthraquinoneꢀsensitization of DBTO is thought to reꢀ
sult in a DBTO triplet that leads to a different deoxygenation
1
25
thus, standard H NMR was not considered to be an adequate
pathway and different ROS. None of the three triplet
quenchers (isoprene, cyclopentadiene, and cyclohexadiene)
tested resulted in a decrease in Φ+DBT with or without BME
(Figure 3D), suggesting a longꢀlived triplet was not involved.
approach for quantitative evaluations of hydrogen bond
strength.
IR spectroscopy, the traditional tool used to measure sulfoxꢀ
1
9
ideꢀalcohol hydrogen bond strength, was not applicable here
due to the overlapping of frequencies of interest and intramoꢀ
Bimolecular Photoreduction
20
lecular hydrogen bonding of thiol linked alcohols. Other
methods of quantitative measurement of hydrogen bond
O
A
S
S
RO
1
strength, including variable temperature control H NMR,
were rejected due to similar interference concerns.
Thiols, like alcohols, are also known to participate in hyꢀ
drogen bonding, though forming much weaker hydrogen
OH
hv
S
2
1,22
RO
bonds.
We compared the change in chemical shifts of ꢀSH
and ꢀOH hydrogen atoms from chloroformꢀd to DMSOꢀdd6:
averaging a change of +0.85 ppm and +2.94 ppm, respectively
more data is available in Table S2 in SI). The data indicates
ROH
RO
O
S
OH
S
(
thiolꢀsulfoxide hydrogen bonds are much weaker than alcoholꢀ
sulfoxide hydrogen bonds. No significant chemical shift was
observed for 1ꢀpropanethiol with and without DBTO while
large chemical shifts were observed for alcohols. Therefore,
the effect of hydrogen bonding of thiols to DBTO is likely
negligible compared to alcohols, which is consistent with the
large increases in quantum yield for HOꢀcontaining bifuncꢀ
tional reactants. Nonetheless, we cannot rule out a small or
OH
CN
HO
B
SH
No disulfide
hv
in ACN
CN
+
H
N
C
D
negligible effect of sulfoxideꢀthiol hydrogen bonding on Φ+DBT
RSH
No disulfide
in the presence of thiols.
DBTO
hv
in ACN
Mechanism for Quantum Yield Enhancement. One poꢀ
tential mechanism that would account for the increase in Φ+DBT
with increasing reactant concentration is the known bimolecuꢀ
R = ꢀPr or ꢀEtOH
HO
SH
2
3
No decrease
in Φ+DBT
lar photoreduction mechanism. In this mechanism, the sulꢀ
foxide accepts an electron from a reducing agent, such as
methoxide in basic conditions, and subsequently releases oxyꢀ
DBTO
+
hv, in ACN
23
gen as a hydroxide ion or hydroxyl radical (Figure 3A).
Figure 3. Possible mechanisms for the increase in Φ+DBT in the
presence of BME. (A) Bimolecular photoreduction mechanism.
(BꢀC) Experimental test for bimolecular photoreduction mechaꢀ
nism. (D) Triplet quenching test for a mechanism that involves
BME as a triplet sensitizer.
2
3
At increasing BME concentrations, the formation of oxiꢀ
dized BME (BME disulfide) increased as DBT increased, with
yields relative to DBT that are comparable to other reacꢀ
tants (Figure S2 in accompanying SI). Thus, if a bimolecular
%
ACS Paragon Plus Environment

Page 5 of 10
The Journal of Organic Chemistry
2
6
Thiols are not particularly good electron donors. They are,
In summary, our results support an observed concentration
dependence on Φ+DBT that is caused by oxidation with caged
Oꢀatom.
2
7
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
however, adequate hydrogen atom donors, but aromatic sulꢀ
foxides are generally not considered suitable hydrogen atom
acceptors. However, an approximation of the photoinduced
change in free energy for the electron transfer (∆G°et) from
BME (standard redox potential, E° = 1.28V vs. NHE) or 1ꢀ
propanethiol (1.43V vs. NHE) to DBTO indicated that bimoꢀ
lecular photoreduction in the presence of thiols could not be
ruled out on energetic grounds (see SI for detailed calculaꢀ
tions). Conversely, in the presence of alkenes (for 1ꢀoctene, E°
3.08V vs. NHE), bimolecular photoreduction is energetiꢀ
cally unfavorable. However, Φ+DBT increases with increasing
ꢀbutenꢀ1ꢀol concentrations to a similar extent as was obꢀ
served for BME. The enhancement in Φ+DBT cannot solely be
due to bimolecular photoreduction since it is energetically
unfavorable for alkenes.
28
Measuring kcageX and kescape. The photolysis of DBTO in
the presence of a given reactant (X) results in the formation of
the oxidized product (XO) and DBT. Under the presumed
unimolecular S—O bond cleavage mechanism, once a DBTO
molecule absorbs a photon and is excited into a singlet state, it
can either follow a deactivation pathway or undergo intersysꢀ
29
30
3
1
tem crossing (ISC) to a S—O dissociative T state followed by
2
3
2
12
=
S—O bond cleavage.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
In order to align the presumption that caged Oꢀatom (i.e.
3
*
3
[DBTꢀꢀꢀO] ) can generate free O( P) with previous observaꢀ
tions of DBTO photodeoxygenation, there must exist a compeꢀ
tition between three pathways: “solvent cage oxidation,” i.e.
the oxidation of a nearby molecule by caged Oꢀatom within its
Despite their similar standard reduction potentials (E°), in
neat BME, the Φ+DBT was found to be an order of magnitude
higher than in neat 1ꢀpropanethiol. If bimolecular photoreducꢀ
tion were the sole cause of the enhancement in Φ+DBT that is
observed in the presence of BME, then a much greater differꢀ
ence in E° for BME and 1ꢀpropanethiol might be expected.
original solvent cage (kcageX), recombination or deactivation
3
(collectively, k ), and “diffusive separation” of O( P) away
d
2
from DBT (k
) (Scheme 2).
escape
3
Only free O( P) will react with X at a rate equal to k [X].
X
In contrast, the rate of XO formation via caged Oꢀatom is a
function of [X] and kcageX, where kcageX is likely a function of
kX and other factors such as hydrogen bonding affinity and
polarity. When kescape/kcageX is large, the formation of XO
through oxidation by caged Oꢀatom is insignificant relative to
Although we cannot rule out bimolecular photoreduction as
a possible contributing mechanism, none of the results support
bimolecular photoreduction as the primary cause for the conꢀ
centration dependence of Φ+DBT. However, the results are conꢀ
sistent with a unimolecular S—O bond cleavage mechanism,
whereby rate enhancement is achieved in the presence of HOꢀ
containing bifunctional reactants through an increase in “efꢀ
fective concentration” of the reactant, which is achieved
through hydrogen bonding.
3
XO formed by free O( P) oxidation. The oxidation reaction
rate of caged Oꢀatom and X can also be insignificant when
kescape/kcageX is small, provided the oxidation is run at low [X]
3
and assuming the rate limiting step for the O( P) pathway to
3
XO is the formation of O( P) rather than the oxidation of X.
kescape/kcageX was quantified using the relationship between
reactant concentration and Φ+DBT (Figure 1). The quantum
yield for the formation of DBT (Φ+DBT) is the product of the
quantum efficiencies of each reaction step. Upon simplificaꢀ
tion, the Φ+DBT is equal to the total probability of DBT forꢀ
mation (eq 1), where d is equal to the rate of all pathways that
do not result in the formation of DBT. Effectively, d is kd,
assuming unimolecular or pseudoꢀunimolecular deactivation
Scheme 2. Proposed pathways of photolysis of DBTO lead-
ing to DBT.
3
pathways and noting that the oxidation of DBT by O( P)
2
should be neglibile. It is important to note that DBT can only
form by way of the common intermediate, i.e. caged Oꢀatom,
and thus, all d processes occur from caged Oꢀatom and are in
competition with both kescape and kcageX[X] processes. From eq
1
, a linear relationship between X and Φ+DBT/(1ꢀ Φ+DBT) is conꢀ
*
For the purposes of this discussion, [DBTꢀꢀꢀO] in Scheme
structed (eq 2). Derivations for eq 1 and eq 2 can be found in
the SI.
2, is a common intermediate that will be referred to as “caged
Oꢀatom”. Oxidations by caged Oꢀatom are those that occur
within the original DBTO solvent cage, where caged Oꢀatom
forms, and are presumed to be the cause of the observed enꢀ
hancements in Φ+DBT. Caged Oꢀatom could be any one of the
ꢄ
ꢅꢆꢇꢈꢉꢅꢀꢄꢇꢈꢊꢅꢋꢌꢍꢎ
Φꢀꢁꢂꢃ = ꢄ
(eq 1)
(eq 2)
ꢅꢆꢇꢈꢉꢅꢀꢄꢇꢈꢊꢅꢋꢌꢍꢎꢀꢏꢐꢏ
ꢑ
ꢒꢓꢔꢕ
(ꢖꢗꢑ
ꢄ
ꢇꢈꢊꢅꢋ
ꢅꢆꢇꢈꢉꢅ
ꢌꢙꢎ +
ꢘ ꢘ
ꢄ
3
=
)
following or combination thereof: (i) O( P) inside a solvent
ꢒꢓꢔꢕ
cage with DBT, (ii) a nonꢀcovalent complex between the Oꢀ
atom and DBT, or (iii) a DBTO excited state capable of directꢀ
ly transferring Oꢀatom to a given reactant. Experimentally, the
differentiation of these species was not possible; however, the
important distinction is that caged Oꢀatom have the potential
for oxidation of the reactants before escaping the solvent cage
The concentration of X was plotted against Φ+DBT/(1ꢀ Φ+DBT
)
for each reactant. Using a linear fit, kescape/kcageX was calculated
by taking the ratio of the yꢀintercept (kescape/d) and the slope
(kcageX/d). The kescape/kcageX values for a selection of reactants
are presented in Table 2.
3
as free O( P) or before deactivation or recombination can ocꢀ
For reactants containing an alcohol, hydrogen bonding of
the alcohol to the sulfoxide of DBTO seemingly resulted in
lower kescape/kcageX values relative to their nonꢀalcohol counterꢀ
parts. For example, the kescape/kcageX value for BME was found
to be low (271mM) while its nonꢀalcoholic counterpart, 1ꢀ
3
cur. From here on forward, O( P) will only be used to describe
a free oxygen atom, which arises from the caged Oꢀatom (i.e.
*
[DBTꢀꢀꢀO] ) and diffuses through the bulk solution.
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 6 of 10
propanethiol, had a value of 1365mM. The calculated kescape
kcageX values for each reactant represent the concentrations that
/
Experimental solutions containing DBTO, reactant A, reacꢀ
tant B, and an internal standard dissolved in acetonitrile were
photolyzed under broad spectrum UVA light (centered at
350nm). The concentrations of reactants were varied from
competition to competition, from 5mM to 2M for more and
less reactive reactants respectively.
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
are necessary for the rate of oxidation by caged Oꢀatom to
3
equal the rate of oxidation by free O( P). Φ
as Φ+DBT from oxidation of X, and Φ
is defined
is defined as
+
DBTcage
+DBTescape
3
Φ+DBT from diffusion of O( P) away from DBT. The ratio of
the Φ+DBTcage to Φ+DBTescape is equal to the ratio of [X] to the
value of kescape/kcageX. Therefore, oxidation of 1ꢀpropanethiol
within the solvent cage is only competitive at high concentraꢀ
tions (>150mM), but for BME, oxidation within the solvent
cage is competitive even at low concentrations (≥30mM) when
Scheme 3. Elementary reactions for competition experiments.
kA
3
A
B
+
O( P)
AO
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
kB
3
defining a competitive reaction as one accounting for ≥10% of
Φ+DBT (see SI for derivation and sample calculations).
+
O( P)
BO
ꢌꢚꢛꢎ
ꢌꢂꢛꢎ
ꢄꢜꢌꢚꢎ
ꢝ
6
Within the nonꢀalcoholic (1ꢀpropanethiol, 2ꢀpropanethiol,
butyl sulfide, and 1ꢀoctene in Table 2) and HOꢀcontaining
bifunctional reactants, the kescape/kcageX values followed a trend
=
ꢏ
(eq 3)
ꢄꢔꢌꢂꢎꢝ
Two reactants were selected from amongst all others to act
as standard competitors, which would allow for comparison
across all reactants. Benzene was selected as one of the standꢀ
3
within the expectations for the reactivity of O( P). For examꢀ
ple, the kescape/kcageX values of nonꢀalcohol containing subꢀ
strates increased from thiols at 1.3M, to sulfides at 3.1M and
alkenes at 4.9M. As will be discussed later in this work, the
3
ard competitors because its absolute rate constant with O( P)
8
in solution is known. Diphenyl sulfide, which was expected to
3
trend follows the observed order of reactivity with O( P).
have a large rate constant relative to benzene, was selected to
be the second standard competitor. Initially, 1ꢀoctene, whose
rate constant is also known, was used as the second standard
competitor until a side reaction with thiols was discovered.
Table 2. The kescape/kcageX values for some reactants.
Reactant
Bridgea
Length
Isolating kescape from kcageX in Competition Experiments
with HO-containing Bifunctional Molecules. For most reacꢀ
tants, eq 3 is effective in calculating the relative rate constants
kescape/kcageX (mM)
b
2
ꢀMercaptoethanol
2
3
6
271
177
3
with O( P) because most of the oxidation occurs in the bulk
c,d
3ꢀMercaptoꢀ1ꢀpropanol
3
solution by O( P). However, for HOꢀcontaining bifunctional
e
6
1
2
ꢀMercaptoꢀ1ꢀhexanol
ꢀPropanethiol
322
molecules, the oxidation of reactant A to AO can occur within
3
the solvent cage as well as in the bulk solution by O( P), and
1365
1358
574
effectively, for A, there becomes more than one reactive inꢀ
termediate. Thus, eq 3 no longer holds in the presence of the
additional oxidative intermediate.
d
ꢀPropanethiol
Methionol
3
4
4ꢀ(Methylthio)ꢀbutanol
Butyl Sulfide
836
For the purposes of this analysis, we define k as the relaꢀ
rel
3
tive rate constant for the reactions of O( P) with reactants A
3068
576
and B (k /k , shown in eq 4, and illustrated in Scheme 4) and
A
B
3
5
ꢀButenꢀ1ꢀol
ꢀHexenꢀ1ꢀol
2
4
the observed relative rate constant (k ) as shown in eq 5.
obs
1528
4934
3744
ΦAOe and ΦBOe are defined as the quantum yields for the forꢀ
mation of oxidized reactants arising from the reaction with
1ꢀOctene
3
O( P) in the bulk solution.
1
ꢀButanol
ꢄ
ꢄ
ꢜ
ꢔ
ꢌꢂꢎ∗ꢑꢜꢢꢅ
ꢌꢚꢎ∗ꢏꢑꢔꢢꢅ
a
Bridge length defined as the number of carbons between the –
ꢞ
ꢟꢠꢡ =ꢏ
=
(eq 4)
(eq 5)
b,c,e
OH and functional group.
Commercial/common names given
ꢌꢚꢛꢎꢌꢂꢎ
=
ꢌꢂꢛꢎꢌꢚꢎ
for 2ꢀsulfanylethanꢀ1ꢀol, 3ꢀsulfanylpropanꢀ1ꢀol, and 6ꢀ
ꢞꢣꢤꢥ
d
sulfanylhexanꢀol, respectively. Concentration dependence data
for these reactants can be found in Figure S3 in the SI.
Scheme 4. Proposed pathways for the oxidation of reac-
tants, via the oxidant(s) produced in the photodeoxygena-
tion of DBTO.
Competition Experiments. For the majority of the eleven
reactants, kcageX was negligible relative to kescape. In the absence
of oxidation within the solvent cage (kcageX), we presume oxiꢀ
dation of the reactant occurs exclusively through the reaction
3
with O( P). To determine the relative reaction rates of free
3
O( P) with various reactants, competition experiments were
performed, whereby, two reactants were allowed to compete
3
for a common intermediate, O( P), in solution (Scheme 3).
3
The relative O( P) reactivity can be quantified for a given pair
of reactants in the form of a relative rate constant (krel or
k /k ), which is proportional to [AO]/[BO] provided low conꢀ
A
B
6
sumption of A or B (eq 3). Further, if one of the two reacꢀ
tant’s rate constant is known, then absolute rate constants may
be extracted.
ACS Paragon Plus Environment

Page 7 of 10
The Journal of Organic Chemistry
calculated indirectly using k /k and k /k and found to be
SPh2
Using the calculated kescape/kcageX values, an equation relating
x
Bz
x
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
kobs to krel was devised. Herein, we highlight the key points of
this derivation; however, a more detailed derivation is availaꢀ
ble in the accompanying supporting information document.
within experimental error of the directly measured kSPh2/k . 1ꢀ
Bz
Octene was found to be the only exception; however, koctene
was calculated within experimental error of its reported value
8
ꢀ1 ꢀ1
(8.7 × 10 M s ) by averaging the k
values calculated
For the following derivation, terminology and steps are conꢀ
sistent with the reaction model in Scheme 4. X and XO will be
used to describe a generic reactant and its oxidation prodꢀ
uct(s), where X can be A or B. A/AO will be used to describe
a reactant/product with a small kescape/kcageX value (<1M) and
B/BO for a reactant/product with a large kescape/kcageX value
octene
8
ꢀ1 ꢀ1
8
from koctene/k (11.2 × 10 M s ) and koctene/k
(5.9 × 10
Bz
SPh2
ꢀ
1
ꢀ1
M s ).
3
Table 3. Relative rate constants for the reactions of O( P)
with various reactants.
a
b
c
(>1M).
Entry Reactant X
1
k
X
/ kBz
k
X
/ kSPh2
k
X
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
The total concentration of XO formed is equal to the sum of
95.9 ± 4.2
91.5 ± 9.5
1.95 ± 0.10 28.8
1.68 ± 0.37 27.5
XO resulting from the oxidation of X by caged Oꢀatom and
3
O( P). The change in XO concentration is shown in eq 6.
2
3
Φꢍꢛ = ꢏꢦꢧꢦꢠꢥꢨꢧꢩꢠꢦꢍ + ꢦꢧꢦꢨꢧꢪꢠꢍ
(eq 6)
71.0 ± 8.9
66.7 ± 7.5
1.29 ± 0.25 21.3
1.18 ± 0.14 20.0
η is used to indicate the quantum efficiency of all photoꢀ
chemical processes leading to S—O bond cleavage (η ), of the
d
a
4
diffusion of Oꢀatom away from DBT (ηescape), of the reaction
3
between O( P) and X (η ), and of the reaction of X with Oꢀ
X
atom within its original solvent cage (ηcageX). For reactants
with large kescape/kcageX value, ηaηcageX is very small relative to
ηaηescapeηX, with more than 96% of oxidation occurring by
5
6
55.9 ± 7.1
53.9 ± 2.5
1
16.8
0.95 ± 0.02 16.2
0.86 ± 0.14 15.3
3
O( P) in the bulk solution at typically concentrations of X used
in competition experiments. Thus, ΦBO is equal to ηaηescapeηB,
while ΦAO is given by eq 6.
7
50.9 ± 6.0
Therefore, using eq 4-6, an equation for krel is given as a
function of k , the kescape/kcageA value, and the concentrations
obs
of A and B (eq 7). Note that eq 7 only allows for competitions
in which at least one reactant has a high kescape/kcageX value,
which is the case for the standard competitors.
8
9
46.9 ± 8.8
37.2 ± 2.5
10.2 ± 3.8
0.7 ± 0.2
14.1
e
0.33 ± 0.06 8.6
ꢌꢔꢎ
ꢮꢅꢆꢇꢈꢉꢅꢯꢮ
ꢌꢜꢎ
ꢫꢄꢬꢭꢆ
ꢖꢀ
ꢗ
ꢰ
ꢇꢈꢊꢅꢜ
ꢞꢟꢠꢡ
=
(eq 7)
10
0.19 ± 0.04 3.1
ꢮꢅꢆꢇꢈꢉꢅꢯꢮ
ꢇꢈꢊꢅꢜ
a
3
b
krel for O( P) with reactant X relative to benzene. krel for
3
c
O( P) with reactant X relative to diphenyl sulfide. Absolute rate
Competition Results. For all reactants except BME, krel is
9
ꢀ1 ꢀ1
constant
✕10 M s of reactant X, calculated using k / k and
equal to k , and thus, eq 3 - 5 was used interchangeably for
X
Bz
ꢀ1 ꢀ1
obs
8
the known absolute rate constant for benzene (3 × 10 M s ) (ref
). The relative rate constant reflects total reactant reactivity,
most reactants to calculate k . Note that reactions were run to
rel
d
8
low conversions, therefore, the changes in concentrations of A
and B over the course of a given reaction were negligible.
with the total concentration of XO formed including benzaldeꢀ
e
hyde formation in addition to dibenzyl disulfide. koctene was calꢀ
For BME, the krel value was calculated using eq 7. A sumꢀ
mary of these results is provided in Table 3. To verify the
model used to calculate krel for BME, a series of competitions
with BME and each standard were performed at varying BME
concentrations (Table 4). The calculated krel values at the two
different BME concentrations were found to be 94 and 98
culated by averaging the values calculated using kX / k , k
/
Bz
X
kSPh2, k , and k
(from entry 5).
Bz
SPh2
3
Table 4. Competition experiments for O( P) with BME and
standard competitors.
Concentration (mM)
(
Table 4, entry 1 & 2), which gave an average krel value which
Diphenyl
Sulfide
a
was slightly above that of 1ꢀpropanethiol (Entries 1 & 2, Table
Entry BME
Benzene
kobs
117
krel
3). To demonstrate the ability to predict kobs from k , an addiꢀ
rel
tional competition between BME (10 mM) and benzene (1.12
M) was performed. At these concentrations of BME and benꢀ
zene, using a krel of 95.9, the predicted kobs value was 103.3
and found to be 102.9. As expected, krel and kobs were found to
be equal for reactants with high kescape/kcageX values.
1
2
3
4
49
203
26
1680
2240
ꢀꢀ
ꢀꢀ
94
98
ꢀꢀ
179
2.2
2.6
35
50
1.9
2.0
58
ꢀꢀ
The absolute rate constants (k ) for each reactant were calꢀ
X
culated using krel to benzene and the known condensed phase
a
Calculated krel using eq 7 from kobs, initial reactant concentraꢀ
tions, and kescape/kcageA values.
3
8
rate constant for O( P) and benzene in acetonitrile (k , 3 × 10
Bz
ꢀ
1
ꢀ1 8
M s ). To validate each calculated k , kSPh2/kBz values were
X
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 8 of 10
3
Trends in O( P) Reactivity. In solution, the reactions of
CONCLUSIONS
3
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
O( P) with primary thiols were found to have the greatest rate
3
The putative O( P), produced upon DBTO photodeoxygenaꢀ
tion, is selective for primary thiols over other functional
groups. Additionally, moderate selectivity is observed for othꢀ
er thiols, sulfides, and alkenes. Thus, DBTO, modified to
achieve desired solvent solubility, could be a useful preꢀ
oxidant for selective oxidation of cysteine residues in proteins.
Additionally, the preꢀoxidant feature could be particularly
attractive for cellular studies where an equilibration time
might be desirable.
constants, in line with previous observations of thiol selectiviꢀ
1
1
ty in proteins. Also of note, there was no significant differꢀ
ence in reactivity between the two primary thiols considered,
suggesting that local polarity has little to no influence on reacꢀ
tivity.
3
In the gas phase reactions of O( P) and thiols, only rate conꢀ
3
3
stants for 1° thiols have been reported. In solution, the obꢀ
3
served reactivity of O( P) with thiols was found to be 1° >
benzylic > 2° > 3°. The order was found in contrast to the reꢀ
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
This work highlights the effectiveness of DBTO as a photoꢀ
3
activity of O( P) with alkanes, in which a 3° preference is obꢀ
3
activatable precursor for the study of O( P) kinetics in soluꢀ
2
,3,6
served due to the greater reactivity of 3° CꢀH.
tion. At low concentrations of reactant like that which might
be observed in biological settings, this work suggests that oxiꢀ
dation occurs almost exclusively in the bulk solution by free
Generally, functional groups were found to be more reactive
when adjacent to a benzylic position, implying either a steric
effect or that an intermediate is formed capable of being stabiꢀ
lized by the benzene ring. A stabilization effect would be in
3
O( P) rather than in the DBTO solvent cage, which can be
competitive at very high concentrations of reactants (>300mM
on average) and at moderate concentrations for certain alcohol
containing reactant (>40mM, on average).
3
line with previous reports of O( P) oxidation mechanisms,
where abstraction occurs with a radical intermediate and addiꢀ
2,4,8
tion occurs in a stepwise fashion.
Benzyl alcohol was found
to have a rate constant similar to that of alkenes, although alꢀ
EXPERIMENTAL SECTION
3
cohols are usually relatively poor reactants for O( P) oxidaꢀ
General. All reactants and authentic samples of oxidation products
were purchased from Sigma Aldrich, TCI America, or Fisher Scienꢀ
tific and used without further purification unless otherwise noted.
Dibenzothiophene Sꢀoxide was prepared from dibenzothiophene using
10
tion. Styrene was also found to have a greater reaction rate
3
with O( P) compared to its counterpart 1ꢀoctene. In contrast,
benzyl mercaptan was found to have a lower rate constant than
primary thiols. The oxidations of alkenes and alcohols by
2
previously reported methods.
3
HPLC analysis was conducted on an Agilent 1200 series instruꢀ
ment equipped with a quaternary pump, diodeꢀarray detector, and an
Agilent Eclipse XDBꢀC18 column (5 ꢁm, 150 × 4.6 mm). HPLC
methods were developed using a 0.1% trifluoroacetic acid waꢀ
ter/acetonitrile solvent system. At least three injections were perꢀ
formed for all solutions analyzed on the HPLC. GCꢀMS analysis was
carried out on a Shimadzu instrument equipped with a Shimadzu
GCMSꢀQP2010S mass spectrometer and a 30m RDX column (0.25
mm ID 9 0.25 lm film thickness). A HewlettꢀPackard 5900 Series GC
equipped with a flame ionization detector and a Sciencix CTSꢀ5 colꢀ
umn (0.25 mm × 30 m × 0.25 ꢁm) was used for all GCꢀFID analysis.
We note that some compounds (most commonly, sulfoxides and
epoxides), upon GC analysis, decompose in the column before reachꢀ
ing the detector. Therefore, we only confirm the presence or absence
of compounds for which peaks are observed that give identical retenꢀ
tion times as of authentic samples via GCꢀFID analysis.
O( P) are known to occur by different mechanisms, and yet,
they both exhibit a rate enhancement while benzyl mercaptan
does not. It is plausible that the oxidation of the benzylic ꢀ
CH ꢀ of benzyl mercaptan is competitive with oxidation of the
2
thiol, which effectively diminishes any rate enhancement efꢀ
fect.
Among the two sulfides investigated, there was very little
3
difference in observed O( P) reaction rates. Direct competiꢀ
tions of butyl sulfide with diphenyl sulfide in this study proꢀ
vide sufficient validation of our results.
Similar trends are observed in the order of reactivity of nonꢀ
alcohol containing substrates and respective kescape/kcageX valꢀ
ues. This correlation is echoed in the HOꢀcontaining bifuncꢀ
tional reactants kescape/kcageX values. Thus, kcageX would seem to
1
H NMR data for hydrogen bonding analyses were collected on a
be a function of k and the equilibrium of the reactant hydroꢀ
X
400 MHz Bruker instrument.
gen bonding with DBTO, though we were unable to measure
the equilibrium.
Photolysis. All solutions to be photolyzed were prepared using
acetonitrile as the solvent in 5ꢀ, 10ꢀ or 25ꢀmL volumetric flasks. 4ꢀmL
of prepared solution was transferred into a quartz test tube (~1cm x
Implications for DBTO. The assumption that an Oꢀatom
was capable of diffusing away from DBT made during modelꢀ
1
0cm) equipped with a stir bar, stoppered with a rubber septum, and
sealed with parafilm. Solutions were degassed via bubbling with arꢀ
gon (≥ 25 mins). Following sparging, an additional septum was added
and wrapped with parafilm to further deter oxygen leakage. All phoꢀ
tolyses were carried out in a Luzchem LZCꢀ4C photoreactor using 14ꢀ
broadly emitting fluorescent bulbs centered at 350nm (fwhm: 325ꢀ
375nm; primarily UVA and some UVB and visible light), unless
otherwise noted.
ing was seemingly validated with the k
competition results. Freely diffusing O( P) is also consistent
/kcageX analysis and
escape
3
with the percent yields of oxidized reactant relative to Φ+DBT
,
which were found to increase at higher reactant concentraꢀ
tions. Furthermore, the loose association between reaction rate
constants and Φ+DBT is consistent with expectations of a mechꢀ
3
Calibration Curves. All stock and standard solutions were preꢀ
pared fresh daily. Dodecane (1mM) was used as an internal standard.
For each calibration curve, three to ten standard solutions were preꢀ
pared with different concentrations of analyte. Calibration curves
were performed on a GCꢀFID for all the reactant oxidation products,
and peak areas were recorded relative to the peak area for dodecane.
anism resulting in a free O( P).
3
With the large rate constants observed for O( P), there was
3
some concern that O( P) could not exit the solvent cage before
reacting with either a reactant, DBT, or the solvent. The diffuꢀ
3
sion distances of O( P) were estimated to be about 50 Å, on
2
average, in acetonitrile with typical reactant concentrations
used (see SI for calculations and a list of calculated diffusion
distances is given in Table S8). At the estimated diffusion
R values of 0.98 or greater were observed.
Calibration curves for DBT were performed on an HPLC and the
peak areas of DBT were recorded using an absorbance wavelength of
2
3
305nm and plotted against concentration. R values were found to be
distances, it is reasonable to conclude that O( P) is capable of
0
.998 or better.
escaping its solvent cage in tact.
ACS Paragon Plus Environment

Page 9 of 10
The Journal of Organic Chemistry
Reactant Concentration Dependence. For each reactant, six soluꢀ
tions were prepared containing 0, 25ꢀ100, 500, 1000, 1500, and
000mM of reactant and 17mM DBTO (>99.9% purity), and photoꢀ
lyzed for 5 h. Following photolysis, solutions were injected on the
HPLC and the concentration of DBT was calculated using calibration
curves of DBT. For each reactant, at least two trials were performed.
To convert DBT concentration to Φ+DBT, the concentration of DBT
formed in the solution with 0mM reactant was used to calculate the
This work was supported by grants CHEꢀ1255270 from the Naꢀ
tional Science Foundation and donors to the Herman Frasch
Foundation.
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
2
ACKNOWLEDGMENT
We thank Emra Bosnjack and Kathryn Sulkowski with help in the
preparation of this manuscript.
photon flux using its previously report quantum yield in acetonitrile
2
(
0.0024). This method was considered acceptable given that the
Φ
+DBT for the photolysis of DBTO in acetonitrile under the conditions
ABBREVIATIONS
used herein was also determined in the traditional manner (i.e. using
chemical actinometry) and found to be within experimental error of
the reported value.
DBTO, dibenzothiophene Sꢀoxide; DBT, dibenzothiophene;
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
O( P), atomic oxygen; BME, 2ꢀmercaptoethanol or βꢀ
mercaptoethanol; GCꢀFID, gas chromatography – flame ionizaꢀ
tion detection; GCꢀMS, gas chromatography – mass spectrometry;
ACN, Acetonitrile.
Quantum Yield Measurements. A number of Φ+DBT measureꢀ
ments were determined at the specific wavelength of 305nm (± 6nm)
using a 75 W Xe lamp focused directly on a monochromator (Photon
Technologies International). Photolysis of azoxybenzene to yield the
rearranged product, oꢀhydroxyazobenzene, was used as a chemical
REFERENCES
34
actinometer. Using this method, photon flux was measured frequentꢀ
ly to limit error due to drift.
(
1) Chapman, S. Philos. Mag. J. Sci. 1930, 10, 369–383.
(2) Gregory, D. D.; Wan, Z.; Jenks, W. S. J. Am. Chem. Soc. 1997,
119, 94–102.
More than thirty Φ+DBT measurements using this method were
made in the presence of varying concentrations of
mercaptoethanol, and few measurements were made for 1ꢀ
propanethiol and 1ꢀoctene at varying concentrations. Additionally,
+DBT measurements were made in neat 2ꢀmercaptoethanol, 1ꢀ
2ꢀ
(
(
(
3) Lucien, E.; Greer, A. J. Org. Chem. 2001, 66, 4576–4579.
4) Thomas, K. B.; Greer, A. J. Org. Chem. 2003, 68, 1886–1891.
5) Korang, J.; Grither, W. R.; Mcculla, R. D. J. Am. Chem. Soc.
a
Φ
2
010, 132, 4466–4476.
6) Wan, Z.; Jenks, W. S. J. Am. Chem. Soc. 1995, 117, 2667–
2668.
propanethiol, and 1ꢀoctene.
(
Competition Studies. Stock solutions were prepared in 10ꢀ or 25ꢀ
mL volumetric flasks. Solutions containing DBTO (10 – 20mM),
competitor A (5 – 1000mM), competitor B (20 – 2500mM), and doꢀ
decane (1mM) in acetonitrile were used for the “experimental” and
“thermalꢀcontrol” solutions. Solutions containing DBT (0 or 10 –
20mM), competitor A (5 – 1000mM), competitor B (20 – 2500mM),
and dodecane (1mM) in acetonitrile were used for the “photocontrol”
solutions. Stock solutions were transferred to quartz tubes: two quartz
tubes for the photocontrol, two or more for the experimental, and at
least one for the thermalꢀcontrol, which was wrapped in foil. The
solutions were photolyzed for 5 to 24 h in a photoreactor. Three or
more trials were conducted for each competition pair.
Immediately following photolysis, solutions were analyzed by GCꢀ
FID. Peak areas relative to dodecane were recorded and later convertꢀ
ed to concentrations using calibration curves. In addition to GCꢀFID
injections, most solutions were also subjected to HPLC analysis
where areas for DBT were recorded and analyzed for completeness.
Competition pairs with large product yields in either the photoꢀ or
thermalꢀcontrols were excluded. The minor product yields observed in
some controls were accounted for in the calculations of the associated
relative rate constants.
(
(
7) Nag, M.; Jenks, W. S. J. Org. Chem. 2004, 69, 8177–8182.
8) Bucher, G.; Scaiano, J. C. J. Phys. Chem. 1994, 98, 12471–
1
2473.
9) Nip, W. S.; Singleton, D. L.; Cvetanovic, R. J. J. Am. Chem.
(
Soc. 1981, 103, 3526–3530.
(10) Herron, J. T.; Huie, R. E. J. Phys. Chem. Ref. Data 1973, 2,
4
67–518.
11) Zhang, M.; Ravilious, G. E.; Hicks, L. M.; Jez, J. M.;
McCulla, R. D. J. Am. Chem. Soc. 2012, 134, 16979–16982.
12) Stoffregen, S. a; Lee, S. Y.; Dickerson, P.; Jenks, W. S.
(
(
Photochem. Photobiol. Sci. 2014, 13, 431–438.
(13) Zheng, X.; Baumann, S. M.; Chintala, S. M.; Galloway, K. D.;
Slaughter, J. B.; Mcculla, R. D. Photochem. Photobiol. Sci. 2016, 15,
7
91–800.
14) Cardoso, D. V. V.; de Araújo Ferrão, L. F.; Spada, R. F. K.;
RobertoꢀNeto, O.; Machado, F. B. C. Int. J. Quantum Chem. 2012,
(
1
12, 3269–3275.
(15) McCulla, R. D.; Jenks, W. S. J. Am. Chem. Soc. 2004, 126,
1
6058–16065.
(
16) Nicovich, J. M.; Gump, C. A.; Ravishankara, A. R. J. Phys.
Chem. 1982, 30332, 1684–1690.
(17) Nag, M.; Jenks, W. S. J. Org. Chem. 2005, 70, 3458–3463.
ASSOCIATED CONTENT
Supporting Information
(
18) Lazic, V.; Jurkovic, M.; Jednacak, T.; Hrenar, T.; Vukovic, J.
P.; Novak, P. J. Mol. Struct. 2015, 1079, 243–249.
The Supporting Information is available free of charge on the
ACS Publications website.
(19) Alfassi, Z. B.; Andersen, K. K.; Ashworth, M. R. F.;
Braverman, S.; Brot, N.; Chanon, M.; Chatgilialoglu, C.; Drabowicz,
J.; Fujihara, H.; Furukawa, N.; Gavezzotti, A.; Grossert, J. S.;
Hargittai, I.; Herron, J. T.; Hoyle, J.; Kaji, A.; Kielbasinski, P.;
Mikolajczyk, M.; Oae, S.; Pihlaja, K.; Posner, G.; Samat, A.; Schank,
K.; Shorter, J.; Simonet, J.; Still, I. W. J.; Tanaka, K.; Uchida, Y.;
Weissbach, H.; Zoller, U. Sulphones and Sulphoxides (1988); Patai,
S.; Rappoport, Z.; Stirling, C., Eds.; John Wiley & Sons, Ltd:
Chichester, UK, 1988.
3
Experimental information of O( P) oxidation product analyses;
auxillary figures (S1 & S2); NMR data; estimations of ∆G°et;
derivations of kescape/kcageX; derivation of k ; relative ratios for
rel
kescape to kcageX and derivations and calculations for diffusion
;
distances (PDF)
(
20) Haase, D.; Ruch, E.; Acta, T. C.; Sung, E.; Harmony, M. D. J.
Am. Chem. Soc. 1977, 5603, 5603–5608.
(21) Kruusma, J.; Rhodes, A.; Bhatia, R.; Williams, J. A. G.;
Benham, A. M.; Kataky, R. J. Solution Chem. 2007, 36, 517–529.
AUTHOR INFORMATION
Corresponding Author
(
22) Minkov, V. S.; Boldyreva, E. V. J. Phys. Chem. B 2014, 118,
513–8523.
23) Cubbage, J. W.; Tetzlaff, T. A.; Groundwater, H.; McCulla, R.
*Eꢀmail: ryan.mcculla@slu.edu
8
(
Funding Sources
D.; Nag, M.; Jenks, W. S. J. Org. Chem. 2001, 66, 8621–8628.
(24) Freccero, M.; Mella, M.; Albini, A. Tetrahedron 1994, 50,
2115–2130.
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 10 of 10
(
25) Rockafellow, E. M.; McCulla, R. D.; Jenks, W. S. J.
(31) Surdhar, P. S.; Armstrong, D. A. J. Phys. Chem. 1987, 91,
6532–6537.
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Photochem. Photobiol. A Chem. 2008, 198, 45–51.
(
(
26) Burner, U.; Obinger, C. FEBS Lett. 1997, 411, 269–274.
27) Denes, F.; Pichowicz, M.; Povie, G.; Renaud, P. Chem. Rev.
(32) Margrey, K. A.; Nicewicz, D. A. Acc. Chem. Res. 2016, 49,
1997–2006.
(33) Singleton, D. L.; Cvetanovic, R. J. J. Phys. Chem. Ref. Data
1988, 17, 1377–1437.
(34) Bunce, N. J.; Lamarre, J.; Vaish, S. P. Photochem. Photobiol.
1984, 39, 531–533.
2
014, 114, 2587–2693.
(
(
28) Guo, Y.; Jenks, W. S. J. Org. Chem. 1997, 62, 857–864.
29) Klán, P.; Wirz, J. Photochemistry of Organic Compounds:
From Concepts to Practice; Wiley: Chichester, UK, 2009.
30) Wardman, P. J. Phys. Chem. Ref. Data 1989, 18, 1637–1755.
(
Authors are required to submit a graphic entry for the Table of Contents (TOC) that, in conjunction with the manuscript title,
should give the reader a representative idea of one of the following: A key structure, reaction, equation, concept, or theorem,
etc., that is discussed in the manuscript. Consult the journal’s Instructions for Authors for TOC graphic specifications.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ACS Paragon Plus Environment