Organotrifluoroborate hydrolysis: Boronic acid release mechanism and an acid-base paradox in cross-coupling

Article abstract of DOI:10.1021/ja300236k

The hydrolysis of potassium organotrifluoroborate (RBF₃K) reagents to the corresponding boronic acids (RB(OH)₂) has been studied in the context of their application in Suzuki-Miyaura coupling. The "slow release" strategy in such SM couplings is only viable if there is an appropriate gearing of the hydrolysis rate of the RBF₃K reagent with the rate of catalytic turnover. In such cases, the boronic acid RB(OH)₂ does not substantially accumulate, thereby minimizing side reactions such as oxidative homocoupling and protodeboronation. The study reveals that the hydrolysis rates (THF, H₂O, Cs₂CO ₃, 55 °C) depend on a number of variables, resulting in complex solvolytic profiles with some RBF₃K reagents. For example, those based on p-F-phenyl, naphthyl, furyl, and benzyl moieties are found to require acid catalysis for efficient hydrolysis. This acid-base paradox assures their slow hydrolysis under basic Suzuki-Miyaura coupling conditions. However, partial phase-splitting of the THF/H₂O induced by the Cs₂CO ₃, resulting in a lower pH in the bulk medium, causes the reaction vessel shape, material, size, and stirring rate to have a profound impact on the hydrolysis profile. In contrast, reagents bearing, for example, isopropyl, β-styryl, and anisyl moieties undergo efficient "direct" hydrolysis, resulting in fast release of the boronic acid while reagents bearing, for example, alkynyl or nitrophenyl moieties, hydrolyze extremely slowly. Analysis of B-F bond lengths (DFT) in the intermediate difluoroborane, or the Swain-Lupton resonance parameter (R) of the R group in RBF₃K, allows an a priori evaluation of whether an RBF₃K reagent will likely engender "fast", "slow", or "very slow" hydrolysis. An exception to this correlation was found with vinyl-BF ₃K, this reagent being sufficiently hydrophilic to partition substantially into the predominantly aqueous minor biphase, where it is rapidly hydrolyzed.

Full text of DOI:10.1021/ja300236k

Organotrifluoroborate Hydrolysis: Boronic Acid Release Mechanism
and an Acid-Base Paradox in Cross-Coupling
Alastair J. J. Lennox and Guy C. Lloyd-Jones*
*
Guy.lloyd-jones@bristol.ac.uk
Supporting Information
Contents:
General experimental details
S2
S4
1
0
B labelling experiments
Hydrolysis of 1a under basic conditions
Hydrolysis of 1a under buffered conditions
S8
S14
Hydrolysis of 1a under base free conditions
S17
S19
S27
S29
S34
Hydrolytic equilibration of 1a with 2a under base and glass free conditions
Effect of Cs CO on THF/H O phase splitting
2
3
2
Suzuki-Miyaura couplings
Hydrolysis of R-BF K reagents (1a-1s)
3
Survey of B-F bond lengths from single crystal X-ray diffraction
DFT calculations and Swain Lupton () analysis
Partitioning study
S63
S65
S80
S82
S84
Suzuki vs Sonogashira in potassium alkynyltrifluoroborate couplings
References
S1

General experimental details
Techniques
Manipulations involving air and moisture sensitive materials were conducted
employing standard Schlenk-line techniques, using vacuum lines attached to a double
manifold with greaseless J. Youngs valves equipped with an oil pump (0.1 mmHg)
under an atmosphere of dry nitrogen. Where air/moisture sensitive reactions were
conducted, glassware and needles were placed in an oven (200 °C) prior to use and
allowed to cool under an atmosphere of dry nitrogen or under vacuum at 0.1 mmHg
(oil pump). Liquid reagents, solutions or solvents were added via syringe through
rubber septa; solid reagents were added via Schlenk type adapters. The removal of
solvents in vacuo was achieved using a Büchi rotary evaporator (bath temperatures up
to 40 °C) at a pressure of 15 mmHg (diaphragm pump), or at 0.1 mmHg (oil pump) on
a vacuum line at room temperature.
Solvents
Dry solvents were obtained by passing solvent through a column of anhydrous
alumina using equipment from Anhydrous Engineering situated in the University of
Bristol chemistry department. Strauss flasks fitted with a greaseless J. Youngs valve
were used to collect anhydrous solvents, they were dispensed using gas-tight syringes
balanced by a nitrogen inlet. Solvents that required degassing were subjected to three
cycles of freeze-pump-thaw with the exception of THF and water. THF was distilled
under N , sodium and benzophenone, and water was saturated with N via bubbling
2
2
for at least 30 min. Anhydrous, degassed solvents were withdrawn from Strauss flask
through rubber septa using gas-tight syringes balanced by a nitrogen inlet.
Commercial grade solvents were used for chromatography and extraction. Deuterated
solvents for NMR analysis were purchased from Cambridge Isotopes Limited.
Analysis
NMR spectra were recorded on Varian 400, Varian 500, JEOL GX 300 or Eclipse 300
1
13
spectrometers. All chemical shifts were quoted in parts per million (ppm); H and C
1
9
11
NMR spectra were referenced to TMS as an internal standard, F to CFCl and B to
3
BF .OEt with a deuterated solvent unless otherwise stated. Coupling constants, J,
3
2
S2

were calculated using ACDLabs to the nearest 0.1 Hz. The following abbreviations
(
and their combinations) were used to label the multiplicities: s (singlet), d (doublet), t
triplet), q (quartet), m (multiplet) and br (broad). Mass spectra were determined by
(
the University of Bristol mass spectrometry service using a Fisons VG Analytical
Autospec spectrometer in electron impact (EI) ionisation mode.
Chromatography
TLC analysis was performed on Merck Silica gel 60F254 glass backed plates.
Visualisation was achieved by UV fluorescence (254 nm). Flash column
chromatography was conducted using Fluorochem 60 silica: 230-400 mesh (40-63
μm).
Reagents
Reagents and materials purchased from commercial suppliers were used without
further purification unless stated. 1,3-Bis(trifluoromethyl)-5-bromobenzene,
potassium cyclopropyltrifluoroborate (1d), potassium benzyltrifluoroborate (1h),
potassium 2-furyltrifluoroborate (1b), potassium vinyltrifluoroborate (1c), potassium
phenylethynyltrifluoroborate (1e), potassium 4-trifluoromethylphenyltrifluoroborate
(1q), potassium 3,5-bistrifluoromethylphenyltrifluoroborate (1s), potassium 4-
methylphenyltrifluoroborate (1m), potassium phenyltrifluoroborate (1n), potassium 2-
2
naphthyltrifluoroborate
(1p),
[ H ]-bromobenzene,
diisopropylethylamine,
5
triethylamine, 3-(N-morpholino)propanesulfonic acid (MOPS) and boric acid (99%
1
0
B) were purchased from Sigma Aldrich. Potassium 4-fluorophenyltrifluoroborate
1a), benzotrifluoride, 1,8-diazabicycloundec-7-ene (DBU), di-n-butylamine, 4-
(
bromobenzonitrile, potassium 3-nitrophenyltrifluoroborate (1r), copper(I)iodide were
purchased from Alfa Aesar. Trimethylsilylacetylene, cesium carbonate, sodium
sulphate and potassium carbonate were purchased from Fisher Scientific. 4-
Fluorophenylboronic
acid
(2a)
was
purchased
from
Maybridge.
Tris(hydroxymethyl)aminomethane (TRIS) was purchased from Melford.
S1
Bis(triphenylphosphine)palladium(II) chloride [PdCl (PPh ) ] , dichloro-[1,l’-
2
3 2
S2
bis(diphenylphosphino)ferrocene]palladium(II)
[PdCl (dppf)] ,
potassium
2
S3
2
S4
cyclobutyltrifluoroborate and [ H ]-4-fluorophenylboronic acid were prepared by
4
known methods. Potassium 1-naphthyltrifluoroborate (1o), potassium 4-
S3

methoxyphenyltrifluoroborate (1l), potassium cyclohexyltrifluoroborate (1j) and
potassium phenylethenyltrifluoroborate (1k) were prepared via an unpublished route;
1
13
19
11
S5
material identical ( H, C, F and B NMR) to that previously published.
Potassium isopropyltrifluoroborate (1g) and 1,3-diphenylpropyl (1i) were kindly
donated by Prof. V. K. Aggarwal and Dr. T. G. Elford (University of Bristol).
Accurate masses of these reagents and products were measured on a calibrated
Sartorius BP 211 balance.
10B labeling experiments
1
0
10
Synthesis of [ B]-potassium 4-fluorophenyltrifluoroborate, [ B]-1a
10
[
B]-1a
10
Preparation of [ B]-B(Oi-Pr)
3
1
0
To an oven dried round-bottomed flask was charged [ B] – boric acid (0.03 mol, 1.83
g) and calcium hydride (0.12 mol, 5.04 g) and then purged with nitrogen. Anhydrous
isopropanol (0.09 mol, 5.4 g) was added dropwise with stirring down a condenser and
over a stream of N . After bubbling had ceased the reaction mixture was heated to 90
2
°
C for 12 hours, after which the volatile compounds were pumped off in vacuo and
trapped at -78 °C. The contents of the trap were then purified by reduced pressure
distillation (47-50 °C at 30 mm Hg) to give the title compound as a clear colourless
1
3
oil (3.62g, 64.5 %) H NMR (300 MHz, CDCl ) δ = 4.34 (sept. J = 6 Hz 3H), 1.13
3
3
10
(
d, J = 6 Hz 18H). B incorporation was not determined for this material, but
analysis of the 1a derived from it, vide infra, confirmed high abundance.
Data is consistent with that expected based on published data for the same compound
1
0
11
S6
with natural abundance B/ B (20/80)
S4

10
10
S7
Preparation of [ B]-potassium 4-fluorophenyltrifluoroborate, [ B]-1a
10
[
B]-1a
To a two necked round bottomed-flask, purged with nitrogen and a magnetic stirrer
bar, was added 4-fluorobromobenzene (0.023 mol, 4.067 g) and dry diethylether (23
mL). A solution of n-butyllithium (10.4 mL of a 2.22 M solution in THF) was added
dropwise at -78 °C and stirred for 30 mins before being allowed to warm to room
temperature and stirred for a further 15 mins. This solution was then cannulated into a
1
0
flask charged with [ B]-triisopropylborate (0.019 mol, 3.623 g) in dry diethylether
76 mL) under an N atmosphere at -78 °C and was stirred for one further hour. The
(
2
reaction mixture was again allowed to warm to room temperature before a saturated
aqueous solution of KHF (0.087 mol, 6.8 g) was added dropwise and solvent was
2
removed in vacuo. The solids were extracted with acetone (2 x room temperature, 2 x
boiling), combined and filtered. The solvent was evaporated before being taken up in
the minimal volume of hot acetone, filtered and cooled before diethylether was added.
The white solid was filtered off, washed with diethylether and dried in vacuo, to give
1
3
the title compound (3.45 g, 88%) H NMR (300 MHz, d -acetone) δ = 7.45 (t J =
6
3
19
6
.97 Hz 2H), 6.81 (t J = 8.62 Hz 2H). F NMR (282 MHz, d -acetone) δ = -120.45
6
1
0
(
1F), 142.3 (3H). MS (EI) analysis of the boronic acid [ B]-2a derived by hydrolysis
1
0
10
indicated 94.8% B incorporation. Isotope cluster m/z (%) calc.: (for 94.5% B)
39.1 (100); 140.1 (12.4); 141.1 (0.96), observed: 139.1 (100); 140.1 (12.5); 141.1
8).
1
(
Data is consistent with that expected based on published data for the same compound
1
0
11
S5
with natural abundance B/ B (20/80).
S5

Exchange study
10
10
-5
[
B] Potassium 4-fluorophenyltrifluoroborate ([ B]-1a) (1.76 x 10 mol, 3.5 mg)
2
2
-5
and [ H ] 4-fluorophenylboronic acid ([ H ]-2a) (1.76 x 10 mol, 2.5 mg) were added
4
4
as solids to a Schlenk tube. To this was added THF (2 mL) and cesium carbonate
-4
(
1.06 x 10 mol, 34 mg) in water (0.2 mL). This stirring solution was then heated at
1
9
5
5 °C for five hours after F NMR confirmed complete conversion to the 4-
fluorophenylboronic acid had occurred. A sample was removed and immediately
analysed by MS (EI). MS(EI) m/z (%): 139 (93.7), 140 (8.9), 141 (4.8), 142 (2.2), 143
(30.1), 144 (100), 145 (18.8).
B(OH)F K study
3
Conditions were used that had previously been found to produce detectable levels of
B(OH)F K, with a substrate stable to protodeboronation on the relevant time scales.
3
10
10
-5
[
B] Potassium 4-fluorophenyltrifluoroborate [ B]-1a (6.00 x 10 mol, 12.1 mg)
was added as a solid to a PTFE lined NMR tube. To this was added D O (0.3 mL)
2
-5
followed by Cs CO (0.5 equiv., 3.00 x 10 , 9.78 mg) in D O (0.06 mL). The tube
2
3
2
1
9
was shaken for approximately 10 min before being analysed by F NMR (128 scans,
5 °C).
2
This procedure was repeated (2 x scale) but stirred in a glass test tube for 10 min
before a sample was removed and placed in a NMR tube without a PTFE liner.
S6

1
9
Figure S1. F NMR spectra of 1a in basic (0.5 eq. Cs CO ) D O in a PTFE vessel (upper spectrum) and a
2
3
2
glass vessel (lower spectrum). Two extra peaks are clearly shown in the lower spectrum.
1
9
Figure S2. An expansion of F NMR Figure S1 showing the superimposed septet and quartet couplings
1
1
10
arising from natural abundance material ( B 80%, spin = 3/2, and B 20%, spin = 3).
S7

Hydrolysis of 1a under basic conditions
Large volume reactions
The reaction vessel (50 mL rbf, 15 mm wide conical-shaped base Schlenk tube, 15
mL wide round shaped base Schlenk tube, 15 mm wide PTFE test tube and 15 mm
wide PTFE test tube with added glass powder (20 mg grade 3)) was charged with
-5
potassium 4-fluorophenyltrifluoroborate (1a) (4.36 x 10 mol, 8.8 mg), cesium
-4
carbonate (1.3 x 10 mol, 43 mg) and an appropriate magnetic stirring bar for the
vessel. A preheated (55 °C) solution of THF:water (10:1, 5 mL: 0.5 mL) was added
and samples (ca. 10 over a 2/3 hour period) were removed with a plastic syringe and
placed immediately in PTFE lined NMR tubes, pre-cooled to 0 °C. The samples were
kept at that temperature until they were placed directly into the Eclipse 300 NMR
1
9
probe. Each sample was analysed by F NMR and subject to 128 pulses (4 min 17 s)
at 25 °C.
S8

A Standard 5 mm OD NMR tube
B - 50 mL round-bottomed flask
9
8
7
6
5
4
3
2
1
0
8
7
6
5
4
3
2
1
0
0
0.5
1
1.5
0
1
2
3
Time / s x 10-4
Time / s x 10-4
D - Schlenk tube with hemispherical
9
base, ≤ 100 rpm stirring
C - Schlenk tube with conical base
9
8
7
6
5
4
3
2
1
0
8
7
6
5
4
3
2
1
0
0
1
2
3
0
0.5
1
1.5
2
Time / s x 10-4
Time / s x 10-4
S9

F - PTFE test tube
9
8
7
6
5
4
3
2
1
0
E - Schlenk tube with hemisphericalbase,
1
0
9
8
7
6
5
4
3
2
1
0
0
1
2
3
4
0
1
2
3
Time / s x 10-4
G - PTFE test tube with glass powder
Time / s x 10-4
9
8
7
6
5
4
3
2
1
0
0
0.5
1
1.5
2
Time / s x 10-4
Figure S3. Hydrolysis of 1a (8 mM) to 2a in THF:water (10:1) with 3 equiv. of Cs CO at 55 °C in reaction
2
3
vessels A-G, see Figure 1 in main text.
NMR tube reactions
The following experiments were undertaken in a Norell S502 NMR tube in a Varian
00 spectrometer probe.
5
S10

Cs
2
CO
3
-6
Potassium 4-fluorophenyltrifluoroborate (1a) (5.28 x 10 mol) in dry THF (0.1 mL)
was added to a Norell 502 NMR tube containing dry THF (0.5 mL). To this was
-5
added Cs CO (1.58 x 10 mol) in D O (0.06 mL), bringing the total THF:water ratio
2
3
2
to 10:1, the volume to 0.66 mL, the Cs CO concentration to 24 mM and the 1a
2
3
concentration to 8 mM. The tube was briefly shaken then immediately placed into the
1
9
Varian 500 probe which was preheated to 55 °C. The reaction was monitored by F
NMR using a time delayed array to separate the acquisitions. These acquisitions were
roughly every 300 seconds for the first 1.5 hours of reaction then every 900 seconds
for the remaining hour.
DBU
-5
1
,8-Diazabicycloundec-7-ene (DBU) (1.584 x 10 mol) in dry THF (0.3 mL) was
added to the Norell 502 NMR tube containing D O (0.06 mL). This was followed by
2
-6
the addition of 1a (5.28 x 10 mol) in dry THF (0.3 mL), bringing the total
THF:water ratio to 10:1 and volume to 0.66 mL, the DBU concentration to 24 mM
and the 1a concentration to 8 mM. The tube was briefly shaken then immediately
placed into the Varian 500 probe which was preheated to 55 °C. The reaction was
1
9
monitored by F NMR using a time delayed array to separate the acquisitions. These
acquisitions were roughly every 300 seconds for the first 1.5 hours of reaction then
every 900 seconds for the remaining hour.
NEt
3
-5
Triethylamine (1.584 x 10 mol) in dry THF (0.5 mL) was added to the Norell 502
NMR tube containing D O (0.06 mL). This was followed by the addition of 1a (5.28 x
2
-6
1
0 mol) in dry THF (0.1 mL), bringing the total THF:water ratio to 10:1 and volume
to 0.66 mL, the DBU concentration to 24 mM and the 1a concentration to 8 mM. The
tube was briefly shaken then immediately placed into the Varian 500 probe which
1
9
was preheated to 55 °C. The reaction was monitored by F NMR using a time
delayed array to separate the acquisitions. These acquisitions were roughly every 300
seconds for the first 1.5 hours of reaction then every 900 seconds for the remaining
hour.
S11

N,N-Diisopropylethylamine (Hünig’s base)
-5
Hünig’s base (1.584 x 10 mol) in dry THF (0.5 mL) was added to the Norell 502
NMR tube containing D O (0.06 mL). This was followed by the addition of 1a (5.28 x
2
-6
1
0 mol) in dry THF (0.1 mL), bringing the total THF:water ratio to 10:1 and volume
to 0.66 mL, the DBU concentration to 24 mM and the 1a concentration to 8 mM. The
tube was briefly shaken then immediately placed into the Varian 500 probe which
1
9
was preheated to 55 °C. The reaction was monitored by F NMR using a time
delayed array to separate the acquisitions. These acquisitions were roughly every 300
seconds for the first 1.5 hours of reaction then every 900 seconds for the remaining
hour.
0
0
0
0
0
0
0
0
0
.009
.008
.007
.006
.005
.004
.003
.002
.001
0
DBU
Ceasium carbonate
Triethylamine
Hunigs base
0
5000
10000
15000
Time / sec
Figure S4. Hydrolysis of 1a (8 mM) to 2a in THF:water (10:1) with 3 equiv. of base (see above) at 55 °C in a
Norell S502 NMR tube.
Sonication
-5
Cesium carbonate (1.58 x 10 mol ) in D O (0.06 mL) was added to the Norell 502
2
NMR tube containing THF (0.54 mL) and then sonicated for 20 seconds. Potassium
-6
4
-fluorophenyltrifluoroborate (1a) (5.28 x 10 mol) in dry THF (0.06 mL) was added
to the pre-sonicated solution in the NMR tube, bringing the total THF:water ratio to
1
0:1 and volume to 0.66 mL, the cesium carbonate concentration to 24 mM and the
1
a concentration to 8 mM). The solution was then sonicated for a further 20 seconds
before being placed directly into the preheated (55 °C) Varian 500 probe. The
1
9
reaction was monitored by F NMR using a time delayed array to separate the
acquisitions. These acquisitions were every 180 seconds for the first 30 mins then
S12

every 5 mins for the following 50 mins then every 15 min for the remaining two
hours.
0
0
0
0
0
0
0
0
0
.009
.008
.007
.006
.005
.004
.003
.002
.001
0
1
2
- Trifluoroborate
- Boronic acid
0
5000
10000
15000
Time / sec
Figure S5. Hydrolysis of 1a (8 mM) to 2a with 3 equiv. Cs CO in THF:water (10:1) at 55 °C, after 20
2
3
seconds sonication, in an NMR tube. After approximately 5000 seconds hydrolysis proceeds predominantly
by the direct rather than acid catalysed pathway, due to the increase in pH caused by the reduced rate of
hydrolysis.
pH monitoring
A pre-heated (55 °C) solution of THF:water (10:1, 6.6 mL) was added to a solid
-5
mixture of potassium 4-fluorophenyltrifluoroborate (1a) (5.28 x 10 mol, 10.6 mg)
-4
and Cs CO (3 equiv., 1.58 x 10 mol, 51.6 mg) in a 15 mm diameter Schlenk tube.
2
3
The stirring (100 rpm) solution was heated at 55 °C and the pH was read from an
uncalibrated Hanna HI 98103 pH probe every 10 seconds for the first 10 min and
every 30 seconds for the following 2 hours. Samples (0.3 mL) were removed
throughout the reaction using a plastic syringe and placed in pre-cooled PTFE lined
NMR tubes at 0° C. The samples were kept at that temperature until they were placed
1
9
directly into the Eclipse 300 NMR probe. Each sample was analysed by F NMR and
subject to 128 pulses (4 min 17 s) at 25 °C.
S13

8
7
6
5
4
3
2
1
0
12
11
1
0
9
8
7
6
5
[
Trifluoroborate]
pH
0
1000
2000
3000
4000
5000
6000
7000
Time / sec
Figure S6. Hydrolysis of 1a (8 mM) to 2a with 3 equiv. Cs CO in THF:water (10:1) at 55 °C.
2
3
Hydrolysis of 1a under buffered conditions
Tris buffer
-5
Tris(hydroxymethyl)aminomethane (Tris) (6.6 x 10 mol, 8 mg) was added as a solid
to the Norell 502 NMR tube containing THF:water (0.5 mL:0.06 mL). Potassium 4-
-6
fluorophenyltrifluoroborate (1a) (5.28 x 10 mol) in dry THF (0.1 mL) was then
added, bringing the total THF:water ratio to 10:1 and volume to 0.66 mL, the Tris
concentration to 100 mM and the 1a concentration to 8 mM. The tube was briefly
shaken then immediately placed into the Varian 500 probe which was preheated to 55
1
9
°
C. The reaction was monitored by F NMR using a time delayed array to separate
the acquisitions. These acquisitions were roughly every 300 seconds for the first 1.5
hours of reaction then every 900 seconds for the remaining hour.
MOPS buffer
-5
3
-(N-morpholino)propanesulfonic acid (MOPS) (6.6 x 10 mol, 13.7 mg) was added
as a solid to the Norell 502 NMR tube containing THF:water (0.5 mL:0.06 mL). 1a
-6
(
5.28 x 10 mol) in dry THF (0.1 mL) was then added, bringing the total THF:water
ratio to 10:1 and volume to 0.66 mL, the MOPS concentration to 100 mM and the 1a
concentration to 8 mM. The tube was briefly shaken then immediately placed into the
S14

1
9
Varian 500 probe which was preheated to 55 °C. The reaction was monitored by F
NMR using a time delayed array to separate the acquisitions. These acquisitions were
roughly every 300 seconds for the first 1.5 hours of reaction then every 900 seconds
for the remaining hour.
-5
This experiment was repeated with half the concentration of MOPS (3.3 x 10 mol,
6
.85 mg, 50 mM overall)
MOPS buffer
TRIS buffer
0.009
0
0
0
0
0
0
0
0
0
.009
.008
.007
.006
.005
.004
.003
.002
.001
0
0
0
0
0
0
0
0
0
.008
.007
.006
.005
.004
.003
.002
.001
0
5
1
0 mM MOPS
00 mM MOPS
0
1000
2000
3000
0
2000
4000
6000
8000
Time / sec
Time / sec
Figure S7. Hydrolysis of 1a (8 mM) to 2a under buffered (Tris - 100mM right graph, MOPS - 50 and 100
1
9
mM, left graph) THF:water (10:1). An unidentified signal in the F NMR accompanied 2a formation in the
TRIS buffered solution that is possibly the boronic ester of TRIS.
Phosphazene base P
4
-t-Bu
-5
Phosphazene base P t-Bu (6.6 x 10 mol, 66 μL 1 M solution in hexane) in dry THF
4
(0.3 mL) was added to the Norell 502 NMR tube containing potassium 4-
-6
fluorophenyltrifluoroborate (1a) (5.28 x 10 mol) in dry THF (0.3 mL). This was
followed by the addition of D O (0.06 mL), bringing the total THF:water ratio to
2
1
0:1, the P base concentration to 100 mM and the 1a concentration to 8 mM. The
4
tube was briefly shaken then immediately placed into the Varian 500 probe which
1
9
was preheated to 55 °C. The reaction was monitored by F NMR using a time
delayed array to separate the acquisitions. These acquisitions were roughly every 600
seconds for the first 2.5 hours of reaction then every 1800 seconds for the remaining 4
hours.
S15

DBU
DBU (6.6 x 10 mol, 9.9 μL) in dry THF (0.3 mL) was added to the Norell 502 NMR
-5
-6
tube containing potassium 4-fluorophenyltrifluoroborate (1a) (5.28 x 10 mol) in dry
THF (0.3 mL). This was followed by the addition of D O (0.06 mL), bringing the
2
total THF:water ratio to 10:1, the DBU concentration to 100 mM and the 1a
concentration to 8 mM. The tube was briefly shaken then immediately placed into the
1
9
Varian 500 probe which was preheated to 55 °C. The reaction was monitored by F
NMR using a time delayed array to separate the acquisitions. These acquisitions were
roughly every 600 seconds for the first 2.5 hours of reaction then every 1800 seconds
for the remaining 2.5 hours.
0
0
0
0
0
0
0
0
0
.009
.008
.007
.006
.005
.004
.003
.002
.001
0
P4 base buffered hydrolysis of 1a
0
0
0
0
0
0
0
0
0
.009
.008
.007
.006
.005
.004
.003
.002
.001
0
DBU buffered hydrolysis of 1a
0
5000
10000
15000
20000
25000
0
5000
10000
Time / s
15000
20000
Time / s
Figure S8 Hydrolysis of 1a to 2a under buffered conditions (left plot is P t-Bu base (100 mM), right plot is
4
DBU (100 mM)).
pH measurement
The pH of each buffer, MOPS (100 mM) and TRIS (100 mM), no buffer and base,
Et N (24 and 100 mM), Hunigs base (24 and 100 mM), DBU (100 mM) and P -t-Bu
3
4
(100 mM) were measured in THF:water (10:1) using a Hanna HI 98103 pH probe
calibrated only to pH 7 (phosphate buffer) and pH 4 (phthalate buffer). The probe was
given five aqueous washes in between measurements followed by a single wash of
THF:water (10:1). The pH of each solution was measured twice and normalized to
“no buffer” at pH 7.
S16

Table S1. pH of each buffer and base used for homogeneous rate measurements of the
hydrolysis of 1a to 2a.
Buffer (100 mM)
MOPS
pH 1
4.50
pH 2
4.05
pH av.
4.28
a
None
TRIS
7.00
8.90
7.00
8.30
7.00
8.60
Et N
8.00
7.75
7.86
3
Hünigs base
DBU
7.70
7.65
7.68
11.05
13.70
10.80
13.35
10.93
13.53
P4 t-Bu
a
MOPS solution was not completely homogeneous.
Buffer (24 mM)
pH
7.9
Et N
3
Hünigs base
7.65
Hydrolysis of 1a under base-free conditions
Addition of glass powder to pre-equilibrated system
-5
Potassium 4-fluorophenyltrifluoroborate (1a) (4.4 x 10 mol, 8.8 mg) was heated at
5 °C in a stirring solution of THF:water (10:1, 5.5 mL) in a PTFE lined test tube.
5
After 280 min, glass powder (20 mg grade 3) was added. Samples (0.5 mL) were
removed throughout the reaction with a plastic syringe and placed immediately in pre-
cooled (0 °C) PTFE lined NMR tubes. The samples were kept at that temperature
until they were placed directly into the Eclipse 300 NMR probe. Each sample was
1
9
analysed by F NMR and subject to 128 pulses (4 min 17 s) at 25 °C.
S17

0
0
0
0
.008
.006
.004
.002
0
0
5000
10000
15000
20000
25000
Time / s
Figure S9. Equilibration of 1a with 2a at 55 °C in THF:water (10:1) in a PTFE lined test tube. After 16800
seconds glass powder (20 mg) was added.
Effect of glass surface areaS8
-2
Glass powder (Grades 1, 2 and 3, 2.00 × 10 g) was added to a PTFE lined test tube
loaded with a stirrer bar and sealed with a rubber septum under a N atmosphere.
2
Degassed THF:H O (10:1, 8.25 mL) was then added and the test tube was placed into
2
a preheated oil bath at 55 °C. After 10 minutes of stirring, potassium 4-
-5
-2
fluorophenyltrifluoroborate (1a) (6.6 × 10 mol, 1.33 × 10 g) was added and
samples (0.3 mL) were removed into pre-cooled (0 °C) PTFE lined NMR tubes at
1
9
regular time intervals using a plastic syringe. These were stored at 0 °C prior to F
NMR analysis where they were subject to a maximum of 5 min at room temperature.
Spectra were acquired by accumulating 128 scans at 25 °C.
S18

4
.5
3
1
0
.9
3
2
1
0
grade 3 glass
grade 2 glass
grade 1 glass
0.8
grade 3 glass
grade 2 glass
grade 1 glass
0
.7
.6
.5
2
0
0.5
0.4
0.3
0.2
0.1
0
.5
1
.5
0
0
5000
10000
0
200
400
600
800
1000
Time/ s
Time/ s
Figure S10. First order log plots of the hydrolysis of 1a (8 mM) to 2a in THF:water (10:1) at 55 °C with the
addition of three different grades of glass powder. The right hand graph is an expansion of the initial period
of the left graph showing the rapid pre-equilibrium.
Hydrolytic equilibration of 1a with 2a under base and glass free
conditions
Rate of equilibration of 1a with 2a
-6
Potassium 4-fluorophenyltrifluoroborate (1a) (4.36 x 10 mol) was added as a solid to
a PTFE lined NMR tube. To this was added firstly dry THF (0.5 mL) and lastly D O
2
(0.05 mL) before being briefly shaken and placed immediately into the preheated (55
1
9
°
C) Varian 500 NMR probe. The reaction was monitored by F NMR using a time
delayed array to separate the acquisitions. Each acquisition was 64 scans (93 s) long
with a time delay between separate acquisitions of 0 seconds. To ensure accuracy this
experiment was repeated a further two times.
S19

0
0
0
0
0
0
0
0
0
.009
.008
.007
.006
.005
.004
.003
.002
.001
0
0
5
10
15
20
Time / min
Figure S11 The equilibration between 1a and 2a for the three separate runs performed in THF:water (10:1)
at 55 °C in PTFE lined NMR tubes. The fit is derived from the equilibration model outlined below, Scheme
1
.
Equilibration Model
Concentration data, Figure S11, were analysed using the model shown below, Scheme
1
:
Scheme 1. Model used for equilibration between 1a and 2a.
S9
Rate constants (k and k ) were derived from parallel iterations of five data sets. The
1
-1
-8
calculated equilibrium constant (K = k /k = 5.46 x 10 ) has then been used to model
1
1
-1
the effect of [1a] and [H O] on the equilibrium between 1a and 2a, Scheme 2.
2
Effect of [1a] on equilibrium
-7
-6
-6
Potassium 4-fluorophenyltrifluoroborate (1a) (5.5 x 10 , 2.75 x 10 , 4.4 x 10 , 8.25 x
-6
-5
1
0 and 1.1 x 10 mol) was added as a stock solution in dry THF (0.5 mL) to a PTFE
lined NMR tube. To this was added D O (0.05 mL), bringing the total THF:water
2
ratio to 10:1, the volume to 0.55 mL and [1a] to 1, 5, 8, 10, 15 and 20 mM
respectively. The tube was briefly shaken and placed immediately into the preheated
(55 °C) Varian 500 NMR probe. Each sample was given 15 mins of heating before
1
9
being analysed by F NMR (2000 – 200 scans depending on the concentration, 55
C)
°
S20

0
0
0
.025
1
a trifluoroborate
0.02
2a - boronic acid
.015
0.01
.005
0
0
0.005
0.01
0.015
0.02
0.025
[ArB]TOT /M
Figure S12. Equilibrium concentrations of 1a and 2a as a function of different initial 1a concentrations in
THF:water (10:1) in PTFE NMR tubes. Solid lines come from the same model (1a + 2H O <-> 2a + KHF +
2
2
-
3
-2 -1
-4
-2 -1
HF) as that shown in Figure S11 and Scheme 2, where k = 2.00 x 10 M s , k = 3.66 x 10 M s therefore
1
-1
-
8
K = 5.46 x 10 .
1
Effect of [H O] on equilibrium
2
-6
Potassium 4-fluorophenyltrifluoroborate (1a) (3.9 x 10 mol, 0.8 mg) was added to a
PTFE lined NMR tube in THF:water (20:1, 10:1, 7:1, 3:1, 1.66:1, 1:1, 1:1.66, 1:3, 1:7,
0
:1) (0.5 mL). The tube was heated at 55 °C in a water bath for 20 min before being
1
9
placed into a preheated (55 °C) Eclipse 300 NMR probe where it was analysed by F
NMR (200 scans, 55 °C). This procedure was repeated at with one hour equilibration
1
9
time at 25 °C before being analysed by F NMR (200 scans, 25 °C).
Scheme 2. Model used for simulating data for the effect of [1a] concentration and [H O] on the equilibrium
2
between 1a and 2a. It was found that 1a undergoes a stabilising solvation at high [H O]. The experimental
2
data was best fit, at the two temperatures (273 K and 326 K), by using 4 H O molecules.
2
S21

7
6
5
4
3
2
1
0
3
2
28 K
98 K
0
10
20
30
H2O] / M
40
50
60
[
Figure S13. The effect of [H O] on concentration of 2a at hydrolytic equilibrium at 298 K and 328 K. Solid
2
lines are concentration data from the model shown in Scheme 2. Dashed lines are how the model predicts
the system without any solvation of 1a.
Van’t Hoff analysis
-6
Potassium 4-fluorophenyltrifluoroborate (1a) (3.20 x 10 mol) in dry THF (0.4 mL)
was added to a PTFE lined NMR tube containing a D O (0.04 mL), bringing the total
2
ratio to 10:1, the volume to 0.44 mL and [1a] to 8 mM. The tube was immediately
1
9
placed into the Varian 500 probe at 25 °C. It was analysed by F NMR at four
different temperatures (25°C, 35°C, 45°C and 55°C). The reaction was subjected to
1
5 min equilibration time at each temperature before the acquisition (120 scans)
began. Measurements were performed twice at each temperature and equilibrium
2
3
analysed as follows Kapp = [2a][HF][KHF ]/[1a][H O] = [2a] /25*[1a]. Van’t Hoff
2
2
-1
-1 -1
analysis (Figure S14) indicates that ΔH = 0.4 kJmol , ΔS = -0.5 Jmol K . However,
we emphasize that this analysis will be highly dependent on the accuracy of the initial
concentrations (1a and H O) and should be interpreted accordingly.
2
S22

1
/T / K-1
-
16
0
.003
0.00305 0.0031 0.00315 0.0032 0.00325 0.0033 0.00335 0.0034
-
-
-
-
16.2
16.4
16.6
16.8
y = -3847.3x - 4.346
R² = 0.9873
-
17
-
17.2
-17.4
Figure S14. Van’t Hoff plot of equilibrium between 1a (Initial conc. = 8 mM) and 2a at four temperatures
3
(
25, 35, 45 and 55 °C) in THF:water (10:1). K = Kapp = [2a] /25*[1a].
Equilibration of R-BF K (1a-i) with boronic acids (2a-i).
3
-6
The potassium organotrifluoroborate (1a-i) (4.4 x 10 mol) was added to a PTFE
lined Quartz NMR tube. THF:water (10:1, 0.55 mL) was added after which the tube
was shaken and placed immediately into the preheated (55 °C) Eclipse 300 probe. The
1
1
sample was heated at 55 °C for 15 min before being analysed by B NMR (2500
1
9
scans) and F NMR (200 scans) at 55 °C.
S23

1
1
Figure S15. An example B NMR spectrum showing substrate 1c at equilibrium in THF:H O (10:1) at 55
2
°
C.
1
9
Figure S16. An example F NMR spectrum showing substrate 1c at equilibrium in THF:H O (10:1) at 55
2
°
C.
S24

Table S2. Equilibrium populations of substrates 1a-1i in THF:H O (10:1) at 55 °C at
2
8
mM
a
b
c
Substrate (8 mM)
x1
x2
Kapp
Aryl (a)
0.75
0.81
0.67
0.39
1.00
0.61
0.71
0.83
0.79
0.25
0.19
0.33
0.61
0.00
0.39
0.29
0.17
0.21
5.3E-08
2.2E-08
1.4E-07
1.5E-06
0.00
Furanyl (b)
Vinyl (c)
Cyclopropyl (d)
Alkynyl (e)
Cyclobutyl (f)
Isopropyl (g)
Benzyl (h)
2.5E-07
8.8E-08
1.5E-08
3.0E-08
1
a
,3-Diphenylpropyl (i)
Where x = [1] / ([1 + 2] ) – rounded up to 2 d.p.
1
eq.
eq.
b
Where x = [2] / ([1 + 2] ) – rounded up to 2 d.p.
2
eq.
eq.
c
3
Where Kapp = [2] /(25*[1])
Table S3. Equilibrium populations of substrates 1a-1i in THF:H O (10:1) at 55 °C at
2
1
00 mM
a
b
c
Substrate (100 mM)
x1
x2
Kapp
Aryl (a)
0.97
0.97
0.96
0.89
1.00
0.95
0.97
0.93
0.92
0.03
0.03
0.04
0.11
0.00
0.05
0.03
0.07
0.08
7.1E-11
7.1E-11
1.7E-10
3.8E-09
0.00
Furanyl (b)
Vinyl (c)
Cyclopropyl (d)
Alkynyl (e)
Cyclobutyl (f)
Isopropyl (g)
Benzyl (h)
3.4E-10
7.1E-11
9.4E-10
1.4E-09
1
,3-Diphenylpropyl (i)
Where x = [1] / ([1 + 2] ) – rounded up to 2 d.p
Where x = [2] / ([1 + 2] ) – rounded up to 2 d.p
Where Kapp = [2] /(25*[1])
a
1
eq.
eq.
b
c
2
eq.
eq.
3
S25

Degenerate exchange between 1a and 2a
-5
2
Potassium 4-fluorophenyltrifluoroborate (1a) (2.2 x 10 mol, 4.4 mg) and [ H ] – 4-
4
2
-5
fluorophenylboronic acid ([ H ] -2a) (2.2 x 10 mol, 3.16 mg) were added as solids to
4
dry THF (5 mL) in a PTFE lined test tube. The solution was heated at 55 °C for two
hours under an atmosphere of dry N . A sample was removed with a plastic syringe
2
1
9
and placed immediately into a PTFE lined NMR tube, which was analysed by F
NMR (128 scans, 25 °C)
1
9
Figure S17. F NMR spectrum showing the degenerate exchange between 1a and 2a at 55 °C in THF in
PTFE, as indicated by deuterium labelling of 2a and then the isotope shift (Δδ = 0.4 - 0.5 ppm) observed in
1
9
the F NMR.
pH monitoring
A pre-heated (55 °C) solution of THF:water (10:1, 6.6 mL) was added to a solid
-5
mixture of potassium 4-fluorophenyltrifluoroborate (1a) (5.28 x 10 mol, 10.6 mg)
-4
and Cs CO (3 equiv., 1.58 x 10 mol, 51.6 mg) in a 15 mm diameter PTFE lined test
2
3
tube. This stirring (100 rpm) solution was heated at 55 °C and the pH was read from
an uncalibrated Hanna HI 98103 pH probe every 10 seconds for the first 10 min and
S26

every 30 seconds for the following 2 hours. Samples (0.3 mL) were removed
throughout the reaction using a plastic syringe and placed in pre-cooled (0° C) PTFE
lined NMR tubes. The samples were kept at that temperature until they were placed
1
9
directly into the Eclipse 300 NMR probe. Each sample was analysed by F NMR and
subject to 128 pulses (4 min 17 s) at 25 °C.
1
2
- Trifluoroborate
- Boronic acid
0
0
0
0
0
0
0
0
.008
.007
.006
.005
.004
.003
.002
.001
0
10
9
pH
8
7
6
5
4
3
2
0
200
400
600
Time / sec
800
1000
Figure S18. Equilibration of 1a (8 mM) with 2a in THF:water (10:1) in a PTFE lined test tube at 55 °C
Effect of Cs
2
CO
3
on THF/H O phase splitting
2
Cs
2
CO titration
3
A 2 mL graduated pipette was filled with THF:water (10:1, 2 mL) until the solvent
was above the top graduation mark. The outlet was then occluded with a small piece
of molten plastic and the solvent removed until the meniscus was level with the top
-5
-4
graduation mark. Cs CO was then added in aliquots (1.59 x 10 – 1.38 x 10 mol),
2
3
and the apparatus gently agitated to ensure dissolution of the base and mixing of the
phases before allowing it to settle for about a minute. The level of the phase boundary
was then noted after each addition.
S27

1
1
2.00%
0.00%
8
6
4
2
0
.00%
.00%
.00%
.00%
.00%
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
[
Cs CO ]
/M
2
3 total volume
Figure S19. Volume (%) of minor biphase as a function of [Cs CO ]
in THF:water (10:1) at 25 °C.
2
3 NET
Sonication of a basic biphasic solution of 4-fluorophenylboronic acid (2a)
in THF:water
-5
A NMR tube containing THF (0.8 mL) was charged with Cs CO (2.64 x 10 mol, )
2
3
-6
in H O (0.1 mL). To this was added 4-fluorophenylboronic acid (2a) (8.8 x 10 mol
2
1
9
1
.23 mg) in THF (0.2 mL). This sample was analysed by F NMR at 55 °C before
1
9
and immediately after one minute sonication. A time averaged F NMR signal is
-
indicative of Ar-B(OH) (2a) <-> Ar-B(OH) (5a, n = 3) equilibrium position, where
2
3
S4
δ = -111.5 ppm : ≥ 95 % 2a and δ = -118 ppm : ≥ 95 % 5a.
S28

1
9
Figure S20. F NMR spectra of 2a before (upper spectrum) and after (lower spectrum) one minute
sonication in basic (3 equiv. Cs CO ) and biphasic THF:water (10:1)
2
3
Suzuki Miyaura couplings
Suzuki -Miyaura coupling in different reaction vessels
To a 15 mm wide Schlenk tube with a hemispherical base, under a N atmosphere,
2
-4
was charged Cs CO (3 equiv., 1.32 x 10 mol), degassed water (0.5 mL) and
2
3
degassed THF (3 mL). To this was added 1,3-bis(trifluoromethyl)-5-bromobenzene
-5
(4.4
x
10
mol) in dry degassed THF (0.5 mL), potassium 4-
-5
fluorophenyltrifluoroborate (1a) (4.4 x 10 mol) in dry degassed THF (0.5 mL) and
-7
finally bis(triphenylphosphine)palladium(II) chloride (4.4 x 10 mol) in dry degassed
S29

THF (1 mL), bringing the total THF:water ratio to 10:1 and volume to 5.5 mL and 1a
concentration to 8 mM. The same procedure was repeated in parallel but in a 15 mm
wide Schlenk tube with a conical shaped base.
In both cases ≥ 85 % of the reaction volume was within the cylindrical section of the
tube. The top of the Schlenk tube was left open to the air, to keep the concentration of
O constant. The solution was stirred at a rate in which no vortex was produced and
2
1
9
heated at 55 °C for six hours, before being analysed by F NMR (200 scans, 25 °C).
Figure S21 SM coupling of 1a with ArBr indicating how the ratio of side products to cross-coupled product
varies under identical conditions other than the reaction vessel shape. We note that under fast release
conditions (conical bottomed Schlenk tube) the relative stoichiometries of the 4-fluorophenol and 4,4’-
S10
difluorobiphenyl are not consistent with that expected based on the literature mechanism. We have
noticed this on other occasions when there are high concentrations of fluoride. In contrast, the relative
stoichiometries are correct under slow release conditions in the round bottomed Schlenk tube.
S30