S
1
Supporting Information for
Intermolecular
Proton
-
coupled Electron Transfer
Reactivity from a Persistent
Charge
-
Transfer State for Reductive Photoelectrocatalysis
Pablo Garrido
-
Barros,
‡
Catherine G. Romero,
‡
Jay R. Winkler,* and Jonas C. Peters*
Division of Chemistry and Chemical Engineering, California Institute of Technology (Caltech);
Pasadena, California 91125, United States
Table of Contents
S1.
General
experimental information...................................................
2
-
4
S2.
Synthetic details
.......................................................................
5
-
11
S
3
.
Electrochemical data
................................................................
12
-
16
S
4
.
U
V
-
v
is data
..............
...
..........................................................
17
-
19
S
5
.
p
K
a
calculation
.......................................................................
20
-
21
S
6
.
Stoichiometric photochemical
reactions..........................................
22
-
31
S7.
Fluorescence
data............................................................
.......
3
2
-
3
5
S8.
Transient absorption data...........................................................
36
-
67
S9.
Reaction q
uantum yield determination............
..
..............................
68
-
69
S10.
Spectroelectrochemistry of
{Fc‒N‒an}
.....
..................
.
..................
70
-
71
S11.
Isotope scrambling experiment.......................................................
72
S12.
Photoelectrocatalytic reactions...................................................
73
-
80
S13.
H
2
Quantification for CPE..............
..............................................
81
S14.
Stoichiometric reaction
with W(dppe)
2
(NNH
2
)(OTf)
2
...
.......
..............
82
-
8
7
S1
5
.
DFT calculations..................................................................
8
8
-
95
S1
6
.
References..................................................................
..
...
...
9
6
-
9
7
S
2
S1. General Experimental Information
S1.1 Chemical/Reagent Considerations:
All manipulations were carried out using standard Schlenk or
glovebox techniques under
an N
2
or Ar atmosphere as specified. Solvents were deoxygenated and dried by thoroughly
sparging with N
2
followed by passage through an activated alumina column in a solvent
purification system by SG Water, USA LLC. Subsequently, the solvents were further dried and
stored under N
2
atmosphere inside a glove box with molecular sieves obtained from Sigma Aldrich
that were activated at 200ºC overnight under vacuum. Non
-
halogenated solvents were tested with
sodium benzophenone ketyl in
tetrahydrofuran (THF) in order to confirm the absence of oxygen
and water. Deuterated DMSO
-
d
6
, CD
3
CN,
and CDCl
3
solvent
s
(D, 99.9% with a purity of 99.5%)
w
ere
purchased from Cambridge Isotope Laboratories, Inc., and use
d
as received.
N
2
gas in a NEXUS glovebox (Vacuum Atmospheres Company) was purified by a Nexus
modular purification system. The purity of N
2
gas was assessed via colorimetric, gas
chromatography and NMR methods, with regard to NH
3
, N
2
O, NO
2
−
and NO
3
−
impurities
.
1
T
riflic acid
(HOTf)
,
ferrocene,
ferrocenecarboxaldehyde, b
is(
p
-
toluenesulfonyl)amine
(HNTs
2
)
,
2
-
picoline
, t
etrabutylammonium
triflate ([TBA][OTf]), and methyl triflate (MeOTf)
were all used as purchased from Sigma Aldrich.
Tetrabutylammonium hexafluorophosphate
([TBA][PF
6
]) from Sigma Aldrich was recrystallized from ethanol prior to use. Silver triflate
(AgOTf) was purchased from Strem and used without further purification. Whenever water was
specified as solvent, deionized water OmniSolv (Supelco, Sigma Aldich) was used t
o pre
pare the
solutions.
Teflon
-
coated magnetic stir bars were soaked in concentrated nitric acid for at least 1 h,
washed repeatedly with deionized water then acetone, and dried in an oven prior to use. In air
-
or
moisture
-
sensitive reactions, solvents
were deoxygenated and dried by thoroughly sparging with
N
2
followed by passage through an activated alumina column in a solvent purification system by
SG Water, USA LLC.
S1.2 NMR Spectroscopy:
1
H
,
31
P
, and
15
N
spectra were recorded with Varian
or Bruker
400 MHz spectrometers. Spectra
were internally referenced to solvent signals
when deuterated solvents were used
or an internal
standard.
15
N
-
NMRs are referenced to liquid NH
3
scale.
S1.
3
UV
-
V
is
Spectroscopy:
Measurements were taken on a Cary 50 UV
-
V
isible spectrophotometer using a 1 cm quartz cell
sealed with a Teflon stopcock.
S1.
4
Electrochemistry:
A CH
I
instrument 600B
and Biologic VSP potentiostat
w
ere
used for all electrochemical data
collection. Cyclic voltammetry (CV)
experiments were carried out in a one
-
compartment three
-
electrode cell using a boron doped diamond (BDD) disk as the working electrode (3 mm diameter),
S
3
a
glassy carbon (GC) disk
as the counter electrode, and a Ag/AgOTf (5 mM) reference electrode.
Details for the CVs are noted as they appear. E
1/2
values for the reversible waves were obtained
from the half potential between the oxidative and reductive peaks. For all measurements
,
IR
compensation was applied accounting for 85% of the total resistance. All the reported potentials
are referenced to the ferrocenium/ferrocene couple (Fc
+/0
).
The BDD disk electrode for cyclic
voltammetry was polished using a PK
-
3
polishing kit with 6 μm diamond (Biologic), while the
GC electrode was polished using alumina 0.05 μm powder.
S1.5 Spectroelectrochemistry
Spectroelectrochemistry was performed in a nitrogen
-
filled glove box with a Biologic SP
-
300
potentiostat. Measurements were performed with a quartz spectroelectrochemical cell with a 0.17
cm path length from Pine Research Instrumentation (AKSTCKIT3), an Ag
wire pseudo
-
reference
electrode (Kurt J. Lesker), a gold honeycomb electrode (Pine Instruments), and a platinum wire
counter electrode (Kurt J. Lesker). Potentiostatic electrochemical impedance spectra (PEIS) were
recorded to obtain Nyquist plots to deter
mine the uncompensated resistance. 95% of the
uncompensated resistance was accounted for using electronic compensation. Measurements were
recorded using a StellarNet SL4 deuterium and tungsten halogen UV
-
vis
-
NIR light source coupled
to StellarNet Black Com
et UV
-
vis and DWARF
-
Star NIR spectrometers. All spectra were
acquired in THF solution with 0.7 M TBAPF
6
electrolyte.
S1.
6
Gas chromatography:
Gas chromatography coupled to a thermal conductivity detector (GC
-
TCD) was used for H
2
quantification. A 100 μL Hamilton syringe was used to sample the headspace and to inject into the
GC
-
TCD. GC
-
TCD was performed in the Environmental Analysis Center (Caltech) using a HP
5890 Series II instrument with N
2
as the carrier gas. Calibration was determined by direct injection
of known volumes of H
2
according to previous work with this instrument.
2
S1.7
Luminescence and Transient Absorption
Time
-
resolved luminescence measurements and transient absorption
measurements were carried
out in the Beckman Institute Laser Resource Center at Caltech. All measurements were performed
with samples under an N
2
atmosphere at room temperature. Samples
were prepared in air
-
tight 1
cm path length quartz cuvettes in a dark, N
2
filled glovebox. Prior to measurement, all samples
were protected from light by wrapping in aluminum foil.
Steady
-
state Luminescence
Fluorescence spectra were recorded using a modified Jobin Yvon Fluor
o
log
instrument. Excitation
was provided by a 450 W Xe arc lamp, wavelength
-
selected with a 0.25 m monochromator.
Luminescence was collected at
90°
with reflective optics, focused onto an optical fiber bundle and
fed to an Ocean Optics QE Pro cooled CCD spectrometer. The instrument was controlled with
software written in MATLAB.
All spectra were corrected for instrument response.
S
4
Time
-
resolved
Fluorescence
Laser excitation was provided by regeneratively amplified (Continuum) pulses from a diode
-
pumped passively mode
-
locked Nd:YAG laser (Spectra Physics Vanguard 2000). The output from
the regenerative amplifier was tripled (355 nm,
~
10 ps) and directed on to the sample held in a
stirred 1
-
cm fluorescence cuvette. Fluorescence was collected at 90
°
using reflective optics and
focused
onto the entrance slit of a 0.275 m spectrograph (Acton SpectraPro 275). A fiber bundle
at the spectrograph image plane collect
ed
the fluorescence and directed it to the entrance slit of a
picosecond streak camer
a
(Hamamatsu C5680). The streak camera was operated in photon
-
counting mode using High Performance Digital Temporal Analyzer software (Hamamatsu). The
instrument time resolution is
~
20 ps.
Transient Absorption
Samples in stirred air
-
tight cuvettes were excited with 355
-
nm pulses (
~
8 ns) from a Q
-
switched
Nd:YAG laser (Spectra
-
Physics Quanta
-
Ray PRO
-
Series) or a tunable Nd:YAG
-
pumped optical
parametric oscillator (Opotek Radiant QX8130U), both operating at a 10 Hz repetition rate. Probe
light from a current
-
pulsed (1 ms) 75
-
W Xe arc
lamp was directed with all
-
reflective optics
colinearly with the laser excitation through the cuvette and wavelength selected using a double
monochromator (ISA DH10). Wavelength
-
select
ed probe light was detected by a photomultiplier
tube (PMT, Hamamatsu R955) wired for 5 gain stages. The PMT output was amplified (Femto
DHPCA
-
100), offset in a wideband differential amplifier
, and digitized at speeds up to 1 GS/s with
a transient digitizer (GageScope). For luminescence decay measurements the probe light was
blocked and sample
fluorescence
was detected by the PMT. For timescales >400
μ
s the Xe arc
lamp was operated in CW mode. Data collection was controlled by a PC with software written in
LabView (National Instruments). Signals were averaged for several hundred laser shots to
optimize signal
-
to
-
noise levels.
S1.8 Computational details
All DFT calculations were performed in the Gaussian 09
,
3
using the TPSS (meta
-
GGA)
4
functional
with def2
-
TZVPP
5,6
on all atoms, Grimme
-
d3 dispersion correction
,
7
and, when explicitly
indicated, SMD
8
implicit solvation modelling acetonitrile for thermochemical parameters (for
direct comparison with experimental available data). Geometry optimizations were computed in
solution without symmetry restrictions. All calculated structures were stationary poi
nts as
confirmed by single
-
point vibrational frequency calculations. Free energy corrections were
calculated at 29
8.15 K and 105 Pa pressure, including
zero
-
point
energy corrections (ZPE). Unless
otherwise mentioned, all reported energy values are free energies in solution under standard state
conditions. Reduction potentials were determined via calculated exchange reactions with
ferrocene/ferrocenium. The p
K
a
values were likewise determined via exchange reactions with 2
-
chloroanilinium/2
-
chloroaniline
.
9
The bond dissociation free energies (BDFEs) were calculated
directly
based on the energy for the different species i
nvolved in the H
-
atom transfer reaction
SubH·
à
Sub + H·, using
the free energy of H∙. This DFT methodology has already
been proven
to successfully reproduce the thermodynamic data associated
with
this catalytic system as reported
in previous publications.
S
5
S2. Synthetic details
{Fc‒N‒an}
,
10
W(N
2
)
2
(dppe)
2
,
11
diphenylfumarate
,
12
and
{
N‒an
}
13
were all synthesized according
to literature procedures.
2
-
picolinium triflate (
[PicH][OTf]
and
[
PicD
][
OTf
]
)
1 mL (0.94 g, 10 mmol) 2
-
picoline was
dissolved in 3 mL diethyl ether. 1 mL (1.7 g
, 11 mmol) trifluoromethanesulfonic acid
(HOTf)
was
dissolved separately in 2 mL diethyl ether, and both were cooled to
-
78 °C for 20 minutes. At
-
78
°C, the
HOTf
solution was added dropwise to the solution of 2
-
picoline which resulted in the
immediate formation of a white precipitate. The suspension was stirred at room temperature for an
additional 30 minutes before washing the solid with 20 mL of diethyl ether and
drying
in vacuo
.
The resulting solid was dissolv
ed in 2 mL THF, layered with 3 mL diethyl ether, and kept
undisturbed at
-
35 °C overnight, forming large crystals of
[PicH][OTf]
.
(2.4 g, 99
% yield).
1
H
-
NMR
(400 MHz, DMSO
-
d
6
)
d
2.73 (s, 3H),
d
7.91 (m, 2H),
d
8.48 (td, 1H),
d
8.80 (dd, 1H),
d
15.5 (br s, 1H).
[
PicD
][
OTf
]
, used for the KIE study, was prepared in the same manner using
DOTf, and resulting in an 85% D labe
l
by
1
H
-
NMR
.
1,2
-
dimethylpyridinium
([PicMe][OTf])
200 μ
L (
0.19
g,
2
mmol) 2
-
picoline was dissolved in
2
mL diethyl ether.
2
2
0
μ
L (
0.33
g
,
2
mmol)
m
ethyl trifluoromethanesulfonate
(MeOTf)
was
dissolved separately in
1
mL diethyl ether, and both were cooled to
-
78 °C for 20 minutes. At
-
78
°C, the
MeOTf
solution was added dropwise to the solution of 2
-
picoline which resulted in the
immediate formation of a white precipitate. The suspension was stirred at room temperature for an
additional 30 minutes before washing the solid with 20 mL of diethyl ether an
d drying
in vacuo
.
(
0.5
0
g, 96 % yield)
.
1
H
-
NMR
(400 MHz,
MeCN
-
d
3
)
d
2.73 (s, 3H),
d
4.13
(s, 3H),
d
7.
83
(m,
2H),
d
8.
34
(td, 1H),
d
8.
56
(dd, 1H)
{Fc‒N
H
+
‒an}
OTf
The monoprotonated complex was synthesized by pronation with HOTf as
follows: 50 mg (0.12 mmol)
{Fc‒N‒an}
was dissolved in 2 mL anhydrous diethyl ether. A solution
of 10.5 μL HOTf (0.12 mmol) was dissolved separately in 1 mL anhydrous diethyl ether. While
stirring, the solution of HOTf was added dropwise to that of
{Fc‒N‒an}
and stirred at room
temperature
for 15 minutes, after which a yellow precipitate formed. The precipitate was washed
with 10 mL anhydrous diethyl ether and 5 mL pentane before being ful
ly dried
in vacuo
. (61 mg,
90
% yield).
1
H
-
NMR
(400 MHz,
DMSO
-
d
6
)
d
2.60 (d, 3H),
d
4.32 (s, 5H),
d
4.56 (m, 6H),
d
5.13
(m, 1H),
d
5.39 (d, 1H),
d
7.63 (m, 4H),
d
8.19 (d, 2H),
d
8.78 (s, 1H)
19
F
-
NMR
(400 MHz,
DMSO
-
d
6
)
d
-
79.35
{Fc‒NMe
+
‒an}
OTf
The methylated complex was synthesized with MeOTf as follows: 10 mg
(0.024 mmol)
{Fc‒N‒an}
was dissolved in 2 mL anhydrous diethyl ether. 2.6 μL MeOTf (0.024
mmol
)
was dissolved separately in 1 mL anhydrous diethyl ether. While stirring, the solution of
MeOTf was added dropwise to that of
{Fc‒N‒an}
and stirred at room temperature for 15 minutes,
after which a yellow precipitate formed. The precipitate was washed with 10 mL anhydrous diethyl
ether and 5 mL pentane before being fully dried
in vacuo
. (12
mg, 86
% yield).
1
H
-
NMR
(400
MHz, MeCN
-
d
3
)
d
2.72 (s, 6H),
d
4.19 (s, 5H),
d
4.4
1
(s, 2H),
d
4.4
9
(s, 2H),
d
4.5
5
(s, 2H),
d
5.53
(s, 2H),
d
7.6
1
(
t
,
2
H)
,
d
7.
72
(
t
,
2
H),
d
8.2
0
(
d
,
2
H),
d
8.
31
(
d
,
2
H),
d
8.8
5
(s, 1H)
19
F
-
NMR
(400
MHz, MeCN
-
d
3
)
d
-
79.36
S
6
W(NNH
2
)(dppe)
2
(OTf)
2
W(N
2
)
2
(dppe)
2
(
2
5 mg,
0.024
mmol) was dissolved in
5
mL dry benzene
in an N
2
filled glovebox.
HOTf
(
4.3 μL,
0.048 mmol) was
diluted with 300 μL benzene and added
dropwise
with stirring. The resulting
buff
-
colored solution was stirred for 1
5
minutes at room
temperature
and then lyophilized
(
31 mg, 99 % yield)
.
1
H
-
NMR
(400 MHz, C
6
D
6
)
d
2.49 (m
, 4H)
,
d
2.62 (m, 4H),
d
5.17 (brs, 2H
, NN
H
2
),
d
6.83 (m, 8H),
d
6.96 (m, 16H),
d
7.11 (m, 8H)
,
d
7.36
(m, 8H)
19
F
-
NMR
d
-
78.
3
,
d
-
76.5
31
P
-
NMR
d
37.2
.
The
15
N
-
labeled compound was synthesized
using W(
15
N
2
)
2
(dppe)
2
which was synthesized according to literature procedure
14
under
15
N
2
and
handled using standard Schlenk techniques under Ar. The spectral
features are
analogous to the
unlabeled compound except the following:
1
H
-
NMR
(400 MHz, C
6
D
6
)
d
5.39 (d, 2H,
15
N
15
N
H
2
J
(
15
N‒
1
H) = 92 Hz)
31
P
-
NMR
d
37.2 (d,
J
(
15
N‒
31
P)
= 4.4 Hz
.
S
7
NMR spectra
Fig. S1
.
1
H NMR spectr
um
(
400 MHz,
DMOS
-
d
6
) of
{Fc‒NH
+
‒an}
ge
nerated
via
protonation
with 1
equivalent
HOTf.
S
8
Fig. S2
.
1
H NMR spectr
um
(
400 MHz, MeCN
-
d
3
) of
{Fc‒N
Me
+
‒an}
ge
nerated with 1
equivalent
MeOTf.
S
9
Figure S
3
.
1
H
-
NMR spectrum (400 MHz, C
6
D
6
)
of
W(NNH
2
)(dppe)
2
(OTf)
2
.
S
10
Figure S
4
.
19
F
-
NMR
of
W(NNH
2
)(dppe)
2
(OTf)
2
in
C
6
D
6
.
S
11
Figure S
5
.
31
P
-
NMR of
W(NNH
2
)(dppe)
2
(OTf)
2
in C
6
D
6
.
S
12
S3. Electrochemical data
Fig. S
6
.
Schematic representation of the anodic redox processes observed for
{Fc‒N‒an}
and
{Fc‒
NH
+
‒an}
and the corresponding cyclic voltammetry of these compounds in a 1 mM solution in
DME containing 0.1 M [TBA][PF
6
] using a glassy carbon disk working electrode, a Ag/AgOTf
reference electrode and a Pt disk counter electrode at 100 mV·s
-
1
.
S
13
Fig. S
7
.
Schematic representation of the anodic and cathodic redox processes observed for
{Fc‒
N‒an}
and
{
N‒an}
and the corresponding cyclic voltammetry of these compounds in a 1 mM
solution in DME containing 0.1 M [TBA][PF
6
] using a glassy carbon disk working electrode, a
Ag/AgOTf reference electrode and a Pt disk counter electrode at 100 mV·s
-
1
.
S
14
Fig. S
8
.
C
yclic voltammetry of a 1 mM
{
N‒an}
(red) and
{
N
H
+
‒an}
(blue) solution in DME
containing 0.1 M [TBA][PF
6
] using a glassy carbon disk working electrode, a Ag/AgOTf reference
electrode and a Pt disk counter electrode at 100 mV·s
-
1
.
S
15
Fig. S
9
. (A) C
yclic voltammetry of a 1 mM
{
N
H
+
‒an
}
(blue trace) and
{
N‒an
}
(red trace) solution
in DME containing 0.1 M [TBA][PF
6
]. (B)
C
yclic voltammetry of a 1 mM
{Fc‒NH
+
‒an}
(blue
trace) and
{Fc‒N‒an}
(red trace) solution in DME containing 0.1 M [TBA][PF
6
]. For both sets of
CVs, we employed a glassy carbon disk working electrode, a Ag/AgOTf reference electrode and
a Pt disk counter electrode at 100 mV·s
-
1
. The schematic representation of the redox processes is
shown above the CV figures.
S
16
Fig. S
10
. C
yclic voltammetry of a 1 mM
{
N‒an
}
(red trace) and 1 mM
{Fc‒N‒an}
with 10
equivalents
of [PicH][OTf] (red trace) solution in DME containing 0.1 M [TBA][PF
6
] using a
glassy carbon disk working electrode, a Ag/AgOTf reference electrode and a Pt disk counter
electrode at 100 mV·s
-
1
. Figure A shows an enlargement of the more cathodic region while Figure
B displays the anodic processes associated to
{Fc‒NH
+
‒an}
. The larger peak at around
-
1.8 is
associated to electrode mediated hydrogen evolution by the [PicH][OTf]. These CVs show the
absence of any shift in the redox potential of the most anodic process associated to an
à
an
·
-
and
thus the lack of chemical reacti
on with protons upon reduction. The availability of protons in the
double layer is evidenced by the larger current obtained at potentials <
‒
2 V in the presence of
[PicH][OTf] and the potential of the anodic wave associated to the protonated form
{Fc‒NH
+
‒
an}
.
S
17
S
4
.
UV
-
vis data
In situ preparation of the protonated
{Fc‒NH
+
‒an}
complex with different acids
Fig. S
11
. UV
-
vis spectra of a 0.5 mM solution of
{Fc‒N‒an}
in DME (black trace) and with the
addition of 10
equivalents
of either [PhNH
3
][OTf] (red trace) or (Tos)
2
NH (blue trace) acids. The
small redshift observed upon addition of both acids evidences the protonation at the amine group
in
{Fc‒N‒an}
to form in situ
{Fc‒NH
+
‒an}
.
S
18
Fig. S
12
. UV
-
vis spectra of a 0.5 mM solution of
{Fc‒N‒an}
in DME (black trace) and with the
addition of 10
equivalents
of either BzOH
(benzoic acid,
red trace) or
2,4
-
CF3
BzOH (
2,4
-
bis(trifluoromethyl)benzoic acid,
blue trace) acids. The absence of any shift in the maximum
absorption bands upon addition of both acids reveals the lack of protonation of
{Fc‒N‒an}
.
Fig. S1
3
. UV
-
vis spectra of a 0.5 mM solution of
{Fc‒NH
+
‒an}
in DME (red trace) and with the
addition of 20
equivalents
of acetophenone showing the absence of any relevant change.
S
19
Fig. S1
4
. UV
-
vis spectra of a 0.05 mM solution of
{Fc‒N
Me
+
‒an}
in DME.
{Fc‒NMe
+
‒an}
S
20
S
5
. pK
a
calculation
The p
K
a
of the
{Fc‒N‒an}
complex to form
{Fc‒NH
+
‒an}
was calculated based on
1
H NMR
titration using anilinium triflate [PhNH
3
][OTf] acid in CD
3
CN.
{Fc‒N‒an}
is dissolved in CD
3
CN
and an initial NMR is taken. Subsequently, aliquots of a CD
3
CN stock solution
containing
[PhNH
3
][OTf] (20 mM) are added to monitor changes in the chemical shifts upon protonation. The
relative concentration of the protonated and deprotonated species
is
determined from the chemical
shift of the representative peaks relative to the chemical shifts of the protonated and deprotonated
species analyzed independently,
using eq
uations
S1
and
S2
:
푓
(
{
퐹푐
‒
푁
‒
푎푛
}
)
=
훿
−
훿
{
"#
,
%
&
!
,
'(
}
훿
{
"#
,
%
,
'(
}
−
훿
{
"#
,
%
&
!
,
'(
}
퐸푞
.
푆
1
푓
5
푃
ℎ
푁
퐻
*
+
9
=
훿
−
훿
,
-
%
&
"
훿
,
-
%
&
#
!
−
훿
,
-
%
&
"
퐸푞
.
푆
2
Having the fraction of each component for the different NMR samples, the concentration of each
component was calculated assuming a normalized concentration for [
{Fc‒N‒an}
+ [
{Fc‒NH
+
‒
an}
] of 1 and using the relative integral of the peaks for [PhNH
3
+
+ PhNH
2
] to calculate the relative
concentration of the latter. Using this total normalized concentration of [PhNH
3
+
+ PhNH
2
] and
[
{Fc‒N‒an}
+ [
{Fc‒NH
+
‒an}
] and the fractions calculated previously, the concentration for each
species and thus the equilibrium constant for the protonation process can be calculated according
to the following equations:
[
{
퐹푐
,
푁
,
푎푛
}
]
=
푓
(
{
퐹푐
,
푁
,
푎푛
}
)
[
{
퐹푐
‒
푁
퐻
+
‒
푎푛
}
]
=
1
−
[
{
퐹푐
‒
푁
‒
푎푛
}
]
>
푃
ℎ
푁
퐻
*
+
?
=
푓
5
푃
ℎ
푁
퐻
*
+
9
·
[
푃
ℎ
푁
퐻
*
+
+
푃
ℎ
푁
퐻
.
+
]
[
푃
ℎ
푁
퐻
.
]
=
>
푃
ℎ
푁
퐻
*
+
+
푃
ℎ
푁
퐻
.
+
?
−
>
푃
ℎ
푁
퐻
*
+
?
퐾
/0
=
[
푃
ℎ
푁
퐻
.
]
[
{
퐹푐
‒
푁
퐻
+
‒
푎푛
}
]
>
푃
ℎ
푁
퐻
*
+
?
[
{
퐹푐
‒
푁
‒
푎푛
}
]
퐸푞
.
푆
3
A p
K
a
of 10.62 for PhNH
3
+
was used
,
15
resulting in a p
K
a
of
14.3
for
{Fc‒NH
+
‒an}
.
S
21
Fig. S1
5
.
1
H NMR spectra
(400 MHz)
in CD
3
CN for the titration experiment used in the pK
a
determination of
{Fc‒NH
+
‒an}
. The purple trace is the spectrum of
{Fc‒N‒an}
, the blue trace the
spectrum of
{Fc‒NH
+
‒an}
, and the cyan, green, yellow and red spectra correspond to subsequent
additions of [PhNH
3
][OTf] to a CD
3
CN solution of
{Fc‒N‒an}
.