Rapid in Situ Generation of Benzene Diazonium Ions under Basic Aqueous Conditions from Bench-Stable Triazabutadienes

Article abstract of DOI:10.1055/s-0036-1588797

Aryl diazonium ions are useful across a range of chemical/biochemical areas. These species are generally made using strongly acidic conditions, which can be detrimental to acid-sensitive functional groups. The ability to generate benzene diazonium ions under basic aqueous conditions is reported herein along with newfound understanding on the aqueous reactivity of solubilized triazabutadienes whereby π-interactions appear to confer a large degree of stability.

Full text of DOI:10.1055/s-0036-1588797

SYNLETT
0
9
3
6
-
5
2
1
4
1
4
3
7
-
2
0
9
6
© Georg Thieme Verlag Stuttgart · New York
2017, 28, A–D
letter
A
en
J. He et al.
Letter
Synlett
Rapid in Situ Generation of Benzene Diazonium Ions under Basic
Aqueous Conditions from Bench-Stable Triazabutadienes
Jie Hea
Flora W. Kimanib
John C. Jewett*a
a Chemistry and Biochemistry, University of Arizo-
na, 1306 E. University Blvd, Tucson, AZ 85721,
USA
jjewett@email.arizona.edu
b Current address: 1102 Natural Sciences 2, Univer-
sity of California, Irvine, California, 92697-2025,
USA
Received: 08.02.2017
Accepted after revision: 27.03.2017
Published online: 27.04.2017
chemistry is that it generally requires liberation under
mildly acidic media or with the aid of light and is slow in
the absence of those two conditions.10 Herein, we report a
new benchstable triazabutadiene that can readily liberate
benzene diazonium ions with remarkable ease in neutral
and even basic environments.
DOI: 10.1055/s-0036-1588797; Art ID: st-2017-r0097-l
Abstract Aryl diazonium ions are useful across a range of chemi-
cal/biochemical areas. These species are generally made using strongly
acidic conditions, which can be detrimental to acid-sensitive functional
groups. The ability to generate benzene diazonium ions under basic
aqueous conditions is reported herein along with newfound under-
standing on the aqueous reactivity of solubilized triazabutadienes
whereby π-interactions appear to confer a large degree of stability.
H2N
N2
a)
NaNO2
HCl
R
b)
Key words diazonium ions, acids, cleavage, bioorganic chemistry,
protecting groups
N
~ neutral
pH
N2
N
N
N
N
R'
While one of the oldest transformations in organic
chemistry, the generation of aryl diazonium ions has largely
remained the same since Griess’ work in 1858. The strongly
acidic reaction conditions used limits the diversity of func-
tionality that could otherwise be found connected to the
aryl diazonium species (Scheme 1, a). In spite of this chal-
lenge, the products of aryl diazonium ions have been a
mainstay in photoswitching chemistry,1 dye chemistry,2
and had a foundational role in the area of bioconjugation.3
We recently reported water-soluble triazabutadiene probes
that release benzene diazonium ions when exposed to
mildly acidic physiologically relevant conditions (Scheme 1,
b).4–6 The design of these probes was initially inspired by
literature reports on the factors contributing to the stability
and reactivity of the triazabutadiene scaffold.7 The value of
the triazabutadiene moiety is manifest when it is consid-
ered as a protecting group for aryl diazonium ions.6 In this
role, it enables acid-sensitive or highly functionalized cargo
be attached to the diazonium ion.8 Such was the case when
we labeled proteins with triazabutadienes and liberated re-
active aryl diazonium species that went on to react with ty-
rosine mimics.9 A limitation of the existing protection
Scheme 1 Aqueous syntheses of aryl diazonium ions. (a) Historically
aryl diazonium ions have been made using a nitrosonium ion generated
under strongly acidic conditions. (b) More recently, triazabutadienes
have been shown to be a source of aryl diazonium ions that can be
formed under mild conditions.
In our previous work we had determined that having a
mesityl group on the imidazole-derived left half (R in
Scheme 1, b) provided a good balance between stability and
reactivity for release of the diazonium below pH 7.4 Indeed,
the majority of our earlier studies were focused on the me-
sityl-substituted triazabutadiene due to the stability it af-
forded when compared to alkyl-substituted counterparts.
We hypothesized that the mesityl stability was primarily
due to steric consideration; but upon irradiating the com-
pound with light, we found that our understanding was in-
complete. We later showed that light can isomerize the tri-
azabutadiene about the N1–N2 bond, and while trying to
understand the photoinduced conformational changes we
came to recognize that 1 sits contrasteric, such that the me-
sityl is on the same side of the molecule as the phenyl ring
© Georg Thieme Verlag Stuttgart · New York — Synlett 2017, 28, A–D

B
J. He et al.
Letter
Synlett
(Figure 1).4 This orientation was counterintuitive based on
the steric implications of having these bulky groups in close
proximity, but we hypothesized that it was the result of sta-
bilization from π-interaction. It was within this context that
we pursued the synthesis of tert-butyl derivative 2 to an-
swer the question of what effect a non-π-interaction steri-
cally large group would have on the reactivity of these
uniquely pH-responsive molecules.
Table 1 Relative Reaction Rates of Triazabutadienes 1–3 in Phosphate-
Buffered Solutionsa
R
N
N
pH
N2
N
N
N
3
O3S
pH
8
R
Half-consumed
Rate (1/s)
R = Mes, 1
t-Bu, 2
N3
R
N1
N
0.03·10–3
2·10–3
N2
Me,
3
R
Mes
t-Bu
Me
20% at 1 h
2 min
N
N
N
N
N
N
N
N
0.9·10–3
0.009·10–3
0.6·10–3
0.3·10–3
N
9 min
3
O3S
3
9
Mes
t-Bu
Me
<5% at 1 h
10 min
30 min
O3S
E
Z
Figure 1 Triazabutadienes can sit in two conformations, as R becomes
large, the steric repulsion should favor the E form (to the left).
a See Figures S1–S3 in the Supporting Information.
The synthesis of 2 was straightforward and it was stable
both in organic solution and as a solid at room temperature
(see Scheme S1 in the Supporting Information).11 Once dis-
solved into water the remarkable instability of the com-
pound became apparent. We were surprised by the reactiv-
ity of 2; it had the shortest diazonium ion releasing half-life
in water solutions observed to date. Indeed, tert-butyl-sub-
stituted 2 was more reactive than the methyl-substituted 3.
To better quantify the differences in reactivity we prepared
a series of buffered solutions and used NMR analysis to
track the degradation of the compounds to phenyl diazoni-
um salts. At pH 7, triazabutadiene 2 reacted at a much fast-
er rate (the time to half-consumption was less than 1 min-
ute, and not detectable by 8 minutes), when compared to
mesityl triazabutadiene 1 (half-life of roughly 1 hour).
While much more reactive than 1, trace amounts of 3 were
still visible at 8 minutes. In order to better assess relative
rates, compounds 1–3 were dissolved into more basic buf-
fers. Both the tert-butyl and methyl triazabutadienes (2 and
3, respectively) degraded quickly at elevated pH 8 and even
pH 9, with half the starting material being consumed with-
in 10 and 30 minutes, respectively (Table 1). Compared with
1, these rates are nearly 60-fold faster. To our surprise, 2
and 3 were even able to react at pH 10 and 11. At pH 10,
90% of 2 was consumed after 18 hours as compared with
67% of 3 (Figure S4 in the Supporting Information). Similar
levels of conversion took over 90 hours at pH 11 (Figure S5
in the Supporting Information). Based on the dependence
of pH, it appears that the mechanism of diazonium ion re-
lease remains the same across these compounds, however,
somehow 2 and 3 are significantly more basic than 1.
formation (Figure 2, a).10 The NOESY data obtained for 2
showed that the tert-butyl group was oriented in the E con-
formation (Figure 2, d). This is consistent with a sterically
dominated model lacking the energy stabilization from π-
interaction. The methyl-substituted triazabutadiene 3 ap-
peared to have a mixture of the two conformers in solution
with the major conformer being more similar to the mesityl
(that is, positioned away from the alkyl sulfonate as in Fig-
ure 2, a and c).
To better understand the origin of increased basicity we
performed 2D NMR experiments on these compounds (see
Figures S9–S11 in the Supporting Information). As noted
above, we had previously established through NOESY ex-
periments that the mesityl on 1 sits in a contra-steric Z con-
Figure 2 Structures of 1 (a and b) and 2 (c and d) are shown with ar-
rows to indicate NOE signals observed (solid green) or not observed
(dotted red). Compound 1 sits in the Z conformation (b), whereas 2 sits
as the E conformer (d).
© Georg Thieme Verlag Stuttgart · New York — Synlett 2017, 28, A–D

C
J. He et al.
Letter
Synlett
We hypothesize that the stabilizing effect that drove the
two aryl groups to be on the same side of the molecule
stemmed from a π–π interaction. This interaction explained
the conformational preference of this derivative where the
attractive interactions outweighed the repulsive forces.5 In
the absence of this stabilizing phenomenon tert-butyl triaz-
abutadiene 2 sits with the phenyl ring situated far away
from the bulky tert-butyl moiety. The 60-fold rate differ-
ence observed (Table 1) correlates to ~10 kJ/mol difference
in stability (assuming the same reaction mechanism). This
number falls within standard values for π–π interactions,
and as such we ascribe the majority of the rate difference to
these interactions or lack thereof.12 We cannot rule out the
possibility that both conformers are in rapid equilibrium
and that one somehow reacts faster than the other. Indeed,
the methyl-substituted triazabutadiene 3 offers evidence
that such an equilibrium can occur. Simple proton NMR of 3
does not show two species, but NOESY experiments suggest
that 3 exists as a mixture of both E and Z conformations.
This indicates that these two isomers likely interconvert
rapidly on the NMR timescale in water and DMSO so as to
give a time-averaged spectrum. Neither 1 nor 2 have cross-
peaks coming from the other isomer in the NOESY, indicat-
ing that they spend the vast majority of the time as Z or E,
respectively. Again, 3 is unique, presumably due to the min-
imal size difference between the methyl and extended
chain substituents. That said, the subtle, twofold, rate dif-
ference between 2 and 3 is best attributed to the differenc-
es in electron-donating capabilities of tert-butyl versus
methyl and casts doubt on the idea that the E conformer is
somehow intrinsically more reactive/basic.
Me
Me
Br
Me
Me
Me
N
N
N
N
N
N
N
N
N
N
3
KO3S
1
4
3
KO3S
Figure 3 Bromine-substituted compound 4 maintains a similar steric
bulk to 1 while changing the electronic environment.
In conclusion, we reported a bench-stable compound
that can be readily synthesized and handled, but releases
aryl diazonium ions with unprecedented ease in neutral
water. A better understanding of which structural and elec-
tronic components lead to the most dramatic effects on rate
will help to design reagents with tailored release dynamics.
Finally, combining these release rates with our recently re-
ported triazabutadiene protection strategy15 should greatly
expand the utility for those wishing to exploit the power of
these unique aryl diazonium protecting groups.
Funding Information
NSF-CAREER (to J.C.J.) (CHE-1552568)
NSF (departmental instrumentation grant for the NMR facility) (CHE-
0840336)
Supporting Information
While the rates of the two alkyl-substituted compounds
were different, we also investigated the effect of changing
the electronic nature of the mesityl ring. We synthesized
triazabutadiene (4,13 Figure 3), which maintained steric
bulk, but had an inductively electron-withdrawing bromine
atom in place of the mildly electron-donating p-methyl of
the mesityl ring on the imidazolium core of the scaffold.14
Despite the significant electronic change, the rate of hydro-
lysis for this compound was found to be quite similar to that
for 1 (Figures S6 and S7 in the Supporting Information).
They were so close that a direct competition experiment
was used to further confirm that p-bromo-substituted 4
was indeed mildly more stable than 1 (Figure S8 in the Sup-
porting Information). Upon inspection by NOESY, it was
confirmed that 4 exists as the Z isomer in solution just like
1 (Figure 2, b; see Figure S12 in the Supporting Informa-
tion). These data further support the proposal that π-inter-
actions play a more significant role in stabilizing the overall
HOMO than the electron-withdrawing nature of the mesityl
ring.
Supporting information for this article is available online at
https://doi.org/10.1055/s-0036-1588797. Included are details about
the synthesis and characterization of all new compounds, and experi-
mental conditions for rate studies.
S
u
p
p
ortiInfogrmoaitn
S
u
p
p
o
nrtIo
g
f
rmoaitn
References and Notes
(1) Beharry, A. A.; Woolley, G. A. Chem. Soc. Rev. 2011, 40, 4422.
(2) Bafana, A.; Devi, S. S.; Chakrabarti, T. Environ. Rev. (Ottawa, ON,
Can.) 2011, 19, 350.
(3) (a) Pauly, H. Hoppe-Seyler's Z. Physiol. Chem. 1915, 94, 284.
(b) Higgins, H. G.; Fraser, D. Aust. J. Sci. Res., Ser. A 1952, 5, 736.
(c) Hooker, J. M.; Kovacs, E. W.; Francis, M. B. J. Am. Chem. Soc.
2004, 126, 3718.
(4) Kimani, F. W.; Jewett, J. C. Angew. Chem. Int. Ed. 2015, 54, 4051.
(5) Triazabutadienes are
a subclass of triazazenes sometimes
referred to as π-conjugated triazenes, see: Patil, S.; Bugarin, A.
Eur. J. Org. Chem. 2016, 860.
(6) Consistent with our ongoing work in the area,6b the acid sensi-
tivity of alkyl triazabutadienes was recently reported, see:
(a) Barragan, E.; Bugarin, A. J. Org. Chem. 2017, 82, 1499.
(b) Kimani, F. W. PhD dissertation: Triazabutadiene Chemistry in
Organic Synthesis and Chemical Biology; University of Arizona:
Tucson, 2016.
© Georg Thieme Verlag Stuttgart · New York — Synlett 2017, 28, A–D

D
J. He et al.
Letter
Synlett
(7) (a) Fanghänel, E.; Hohlfeld, J. J. Prakt. Chem. 1981, 323, 253.
(b) Khramov, D. M.; Bielawski, C. W. J. Org. Chem. 2007, 72,
9407.
(8) Cornali, B. M.; Kimani, F. W.; Jewett, J. C. Org. Lett. 2016, 18,
4948.
(9) Jensen, S. M.; Kimani, F. W.; Jewett, J. C. ChemBioChem 2016, 17,
2216.
(10) He, J.; Kimani, F. W.; Jewett, J. C. J. Am. Chem. Soc. 2015, 137,
9764.
round-bottom flask was added 1,3-propane sultone (1.8 g, 14.9
mmol). The mixture was then brought to reflux overnight. After
cooling the reaction to r.t., the resulting precipitate was filtered
and washed with ether to give S2b as a white solid (54%). The
product was used for the next step without further purification.
1H NMR (400 MHz, DMSO-d6) δ = 9.41 (1 H, t, J = 1.6 Hz), 8.13 (1
H, dd, J = 2.0, 1.5 Hz), 7.95 (1 H, t, J = 1.8 Hz), 7.62 (2 H, t, J = 0.7),
4.44 (2 H, t, J = 7.0 Hz), 2.46 (2 H, t, J = 7.0 Hz), 2.22 (2 H, quint,
J = 7.0 Hz), 2.08 (6 H, s).
(11) No degradation was noted after greater than 1 month at r.t.
(12) Sinnokrot, M. O.; Sherrill, C. D. J. Phys. Chem. A 2004, 108, 10200.
(13) Experimental Procedure for the Synthesis of 4
3-[1-(4-Bromo-2,6-dimethylphenyl)-1H-imidazol-3-ium-3-
yl]propane-1-sulfonate (S2b)
3-{(E)-3-[4-bromo-2,6-dimethylphenyl]-2-[(E)-phenyltriaz-
2-en-1-ylidene]-2,3-dihydro-1H-imidazol-1-yl}propane-1-
sulfonate (4)
To a solution of S2b (200. mg, 0.54 mmol) in THF (10 mL), was
added phenyl azide (64 mg, 0.54 mmol) in a THF (0.5 mL).
The mixture was stirred for 5 min in an ice bath at which time
KOt-Bu (78 mg, 0.70 mmol) was added in one portion. The
mixture was slowly allowed to warm up to r.t. and stirred over-
night. The reaction mixture was filtered through a plug of Celite,
and the solvent was removed under vacuum to give the product
as yellow solid (53%). 1H NMR (400 MHz, DMSO-d6): δ = 7.47 (2
H, s), 7.32 (1 H, d, J = 2.5 Hz), 7.09 (2 H, dd, J = 8.3, 7.2 Hz), 7.02–
6.97 (1 H, m), 6.87 (1 H, d, J = 2.5 Hz), 6.51–6.38 (2 H, m), 4.06 (2
H, t, J = 7.1 Hz), 2.50–2.47 (2 H, m), 2.07–2.05 (2 H, m), 1.97 (6
H, s). 13C NMR (100 MHz, DMSO-d6): δ = 152.1, 151.1, 138.2,
137.5, 131.3, 128.7, 125.5, 121.3, 120.8, 118.0, 116.8, 48.5, 45.0,
25.7, 17.7.
To a solution of 4-bromo-2,6-dimethylaniline (4 g, 20 mmol) in
MeOH (5 mL) in a round-bottom flask was added glyoxal (2.9
mL, 20 mmol, 40% solution in water) with stirring. After a few
minutes, a yellow sticky solid formed to which NH4Cl (2.14 g, 40
mmol) and formaldehyde (3.2 mL 37 wt% in water) were added.
The resulting mixture was diluted with 40 mL MeOH. The
mixture was then refluxed for 1 h before H3PO4 (2.8 mL, 85 wt%
in water) was added dropwise. The reaction was allowed to stir
at reflux overnight. The solvent was removed under vacuum,
and the residue was poured into 30 g ice. The aqueous mixture
was neutralized with 40% KOH to pH 9 and extracted with
CH2Cl2. The organic layer was separated, dried over MgSO4, and
concentrated to give compound S1b as light brown solid (50%).
1H NMR (400 MHz, DMSO-d6): δ = 7.69 (1 H, dd, J = 1.2, 1.0 Hz),
7.53–7.48 (2 H, m), 7.24 (1 H, t, J = 1.3 Hz), 7.14 (1 H, dd, J = 1.2,
1.0 Hz), 1.97 (6 H, t, J = 0.7 Hz).
(14) This constitutes a change in sigma value of roughly 0.5 as com-
pared with the methyl group.
(15) Guzman, L. E.; Kimani, F. W.; Jewett, J. C. ChemBioChem 2016,
17, 2220.
To a solution of S1b (2.5 g, 10.0 mmol) in toluene (100 mL) in a
© Georg Thieme Verlag Stuttgart · New York — Synlett 2017, 28, A–D