Ja1c08810 Si 001

Supporting information for
Large, tunable and reversible pH changes

by merocyanine photoacids
Laura Wimberger, Shyamal K. K. Prasad, Martin D. Peeks, Joakim Andréasson, Timothy W. Schmidt,
Jonathon E. Beves
School of Chemistry, UNSW Sydney, Sydney, Australia
S1. General information 1

S2. Experimental procedures and analytical data 2
S2.1. Precursor synthesis 2
S2.2. Synthesis of merocyanines/spiropyrans 4
S3. NMR spectra of compounds 13

S3.1. Starting materials and intermediates 13
S3.1. NMR spectra of photoswitches 2a–c, 3a–d 17
1
H NMR (600 MHz) spectra of 2a–c, 3a–d in DMSO-d6 17
13
C{1H} (151 MHz) NMR spectra of 2a–c, 3a–d in DMSO-d6 18
S4. In situ irradiation 1H NMR spectra 29

S5. Concentration dependent NMR of 3c 31
S6. UV-Vis studies 33
S6.1. General procedure 33
S6.2. Setup for visible irradiation experiments (λ = 455 nm) 33
S6.3. Setup for UV irradiation experiments (λ = 265 nm) 34
S6.4. Determination of pKadark and pKahν 35
S6.5. pH-Dependent UV-visible absorption data for photoacids 2a–d, 3a–d 36
pH-Dependent UV-visible absorption data for photoacid 2a 36
pH-Dependent UV-visible absorption data for photoacid 2b 37
pH-Dependent UV-visible absorption data for photoacid 2c 38
pH-Dependent UV-visible absorption data for photoacid 2d 39
pH-Dependent UV-visible absorption data for photoacid 3a 40
pH-Dependent UV-visible absorption data for photoacid 3b 41
pH-Dependent UV-visible absorption data for photoacid 3c 42
pH-Dependent UV-visible absorption data for photoacid 3d 43
S6.6. Kinetic parameters of the SP-MC equilibrium for photoswitches 2a–d, 3a–d 44
Determination of equilibrium constant Kc and rate constants kopen and kclose 44
Temperature based errors (kopen, kclose, Kc) 44
Experimental data and kinetic analysis of the SP-MC equilibrium for photoacids 2a–d, 3a–d 46
pH and temperature dependent kobs for photoacid 2a 46
pH and temperature dependent kobs for photoacid 2b 47
pH and temperature dependent kobs for photoacid 2c 48
pH dependent kinetics for photoacid 2d 49
pH and temperature dependent kobs for photoacid 3a 50
pH and temperature dependent kobs for photoacid 3b 51
pH and temperature dependent kobs for photoacid 3c 52
pH and temperature dependent kobs for photoacid 3d 53
S6.7. Calculation of mole fractions of photoacids 2a–c, 3a–d 54
S6.8. Determination of extinction coefficients of SP and MCH 57
S6.9. Calculation of extinction coefficients of MC 57
Calculation of mole fractions of SP (𝜒𝜒𝑆𝑆𝑆𝑆 ) and MC (𝜒𝜒𝑀𝑀𝑀𝑀 ) from Kc 57

𝜆𝜆
Calculation of extinction coefficients of MC (𝜀𝜀𝑀𝑀𝑀𝑀 ) 57
Calculation of an absorbance spectrum of MC 57
Absorbance spectra of MCH, SP and calculated absorbance spectrum of MC 58
S6.10. Determining the pKa value11 59
pKa calculated from pKadark and Kc 59
Fitting an experimentally determined ratio of MCH/MC/SP to determine the pKa 59
S6.11. Photoswitching of photoacids 2a–c, 3a–d 61

Overview of photoswitching parameters of photoacids 2a–c, 3a–d 63
S7. Transient Absorption data 64

S7.1. Transient absorption spectra of the MCH forms of photoacid 2a–c, 3a–d 64
S7.2. Kinetic analysis of the TA data of the MCH form of photoacids 2a–c, 3a–d 66
S7.3. Transient absorption spectra of the MC forms of photoacid 2a–c, 3a–d 70
S8. Summary of key parameters for photoswitches 2a−c, 3a−d 72

S9. Stability of photoacids 2a−c, 3a−d in methanol 75
S10. Irradiation of saturated solutions 76
S10.1. Procedure for irradiating concentrated solutions 76
S10.1. Procedure for irradiating concentrated solutions – time resolved 76
S10.2. Light induced pH switching of photoacids 3c and 3d 77
S10.3. Time resolved pH changes under light irradiation 79
S11. pH drop model 81

dark
S11.1. pH dependent concentration of MCH calculated from pKa 81
S11.2. Extent of photoswitching 81
S11.3. Equilibria affecting proton release under irradiation 83
S11.4. Predicted pH changes under irradiation of photoacids 3c and 3d 84
S11.5. Application of the pH drop model to literature examples 85
S12. Model describing the time dependent pH recovery in the dark 88

S12.1. Model description 88
Input parameters 88
Kinetic equilibrium system 88
Protonation equilibria 88
S12.2. Results of modelled parameters for compounds 3c and 3d 91
Photoacid 3c: 91
Photoacid 3d: 92
S12.3. Modelled pH recovery of photoacid 3c and 3d 93
S13. Propagation of errors 94

S14. Bibliography 96
S1. General information
All data for this work is deposited on the ChemRxiv server, see Wimberger, L.; Prasad, S.; Andréasson, J.; Schmidt, T.;
Beves, J. ChemRxiv 2021, DOI: 10.33774/chemrxiv-2021-gppx1. That content is a preprint and has not been peer-
reviewed.
Reagents and solvents were purchased from Sigma-Aldrich, VWR International, Chem Supply, BDH Chemicals or
Combi-Blocks and used without further purification unless stated otherwise.
Qualitative TLC was performed on silica-coated aluminium plates (Merck, silica 60 F254) with detection by UV light
(λ = 254 nm). Flash column chromatography (FCC) was carried out with SilicaFlash® P60 (Velocity Scientific
Solutions, 63 μm) with the indicated eluent.
Nuclear magnetic resonance spectra were recorded on a Bruker Avance III 300, Bruker Avance III 400 with a
Prodigy CryoProbe, a Bruker Avance III 500 or a Bruker Avance III 600. Signals are reported relative to the residual 1H
NMR signal of CDCl3 (δ = 2.50 ppm) or DMSO-d5 (δ = 2.50 ppm). 13C NMR spectra were referenced to the 13C-D
triplet of CDCl3 or the 13C-D septet of DMSO-d6 (δ = 39.5). Coincidentally equal coupling constants of magnetically
non-equivalent protons are marked as virtual (virt.). The signal multiplicities are indicated by the following
abbreviations: s-singlet, d-doublet, t-triplet, q-quartet, quin-quintet, sex-sextet, sept-septet, m-multiplet, br-broad. The
coupling constants J are reported in Hertz (Hz) as an average of the experimentally determined ones. Structural
assignments are based on COSY, HMBC, and HSQC experiments.
UV-vis absorption spectra were measured in 41FLUV10 cells from fireflysci made of quartz (pathway 10 mm) on an
Agilent Cary 60 UV-vis spectrometer equipped with a customized Cary Single Cell Peltier Accessory for temperature
control (T = 25 °C) and stirring of the samples. A LED irradiation setup was employed as described previously by our
group.1 A Luxeon Rebel LED mounted ca. 4 cm away from the sample holder and focused with a Carclo 20.0 Fibre
coupling lens served as irradiation source. Photoswitching cycles were controlled with a timer relay module (FRM01).
High resolution mass spectrometry (HRMS) was measured on a hybrid linear quadrupole ion trap mass spectrometer
(Thermo LTQ Orbitrap XL) with an external nanospray ionisation (NSI) source.
1
S2. Experimental procedures and analytical data
S2.1. Precursor synthesis
3-(2,3,3-Trimethyl-3H-indol-1-ium-1-yl)propane-1-sulfonate (S1)
According to a modified literature procedure2: A solution of 1,3-propanesultone (1.65 mL, 2.30 g, 18.8 mmol, 1.00 eq.)
and 2,3,3-trimethyl-3H-indole (3.02 mL, 3.00 g, 18.8 mmol, 1.00 eq.) in toluene (100 mL) was heated to 135 °C for 2.5
days. After letting the solution return to r.t. the purple precipitate was collected by vacuum filtration and washed with
petroleum ether (ca. 15 mL, 40/60). The precipitate was dried in vacuo for ca. 6 h and gave a purple solid as 3-(2,3,3-
trimethyl-3H-indol-1-ium-1-yl)propane-1-sulfonate (3.88 g, 13.8 mmol, 73%) which was used in the consecutive
reaction without further purification.
1
H NMR (400 MHz, DMSO-d6): δ / ppm = 1.53 (s, 6H, Hd), 2.11-2.21 (m, 2H, Hk/l/m), 2.63 (t, 3J = 6.6 Hz, 2H, Hk/l/m),
2.83 (s, 3H, Ha), 4.60-4.70 (m, 2H, Hk/l/m), 7.57-7.66 (m, 2H, Hf/g/h/i), 7.79-7.84 (m, 1H, Hf/g/h/i), 8.01-8.08 (m, 1H,
Hf/g/h/i).
Spectral data matches those reported in the literature.3
2,3,3-Trimethyl-1-(3-(trimethylammonio)propyl)-3H-indol-1-ium dibromide (S2)
According to a modified literature procedure4: In a flame dried flask, (3-bromopropyl)trimethylammonium bromide

(1.80 g, 6.91 mmol, 1.10 eq.) was suspended in acetonitrile (30 mL) and 2,3,3-trimethyl-3H-indole (992 μL, 1.00 g,
6.28 mmol, 1.00 eq.) was added. The solution was heated to 95 °C for seven days, cooled to r.t. and the solvent was
removed in vacuo. The crude solid was dissolved in water (100 mL) and washed with diethyl ether (4 × 50 mL). The
solvent was removed in vacuo and the solid was dried in a desiccator for several days giving a purple, very hygroscopic
solid as product (54.4 w%, 2.13 g, 54%) in a mixture with the (3-bromopropyl)trimethyl-ammonium bromide starting
material. The product was used in the consecutive reaction without further purification.
Similar solubilities of the product and the starting material did not allow purification by precipitation contrary to
literature reports.4
1
H NMR (400 MHz, DMSO-d6): δ [ppm] = 1.56 (s, 6H, Hd), 2.33-2.40 (m, 2H, Hl), 2.94 (s, 3H, Ha), 3.13 (s, 9H,
NMe3), 3.63-3.71 (m, 2H, Hk/m), 4.52 (t, 3J = 8.1 Hz, 1H, Hk/m), 7.60-7.68 (m, 2H, Hf/g/h/i), 7.83-7.89 (m, 1H, Hf/g/h/i),
8.05-8.15 (m, 1H, Hf/g/h/i).
Spectral data differs from that reported in the literature.5
2
Starting material: (3-bromopropyl)trimethyl-ammonium bromide
1
H NMR (400 MHz, DMSO-d6): δ [ppm] = 2.28 (dt, 2J = 16.6 Hz, 3J = 6.5 Hz, 2H, Hb), 3.10 (s, 9H, NMe3), 3.40-3.49
(m, 2H, Ha/c), 3.58 (t, 3J = 6.4 Hz, 2H, Ha/c).
5-Methoxy-2,3,3-trimethyl-3H-indole (S3)
According to a modified literature procedure6: 5-Methoxy-2,3,3-trimethyl-3H-indole was prepared according to a

literature procedure. To a solution of 3-methyl-2-butanon (5.90 mL, 4.75 g, 55.2 mmol, 2.00 eq.) in glacial acetic acid
(92 mL), 4-methoxyphenylhydrazine hydrochloride (4.82 g, 27.6 mmol, 1.00 eq.) was added. The suspension was
heated to 140 °C for 10 min dissolving the starting materials and resulting in a colour change from colourless to purple.
The mixture was then stirred at r.t. for 5.5 h and potassium hydroxide pellets were added stepwise to neutralise the
reaction mixture. The solution was extracted with Et2O (3 × 75 mL), the combined organic phases were washed with
brine (50 mL), dried over MgSO4 and desiccated in vacuo. Separation by column chromatography (6 × 11 cm,
hex/EtOAc = 7/3 → 1/1) resulted in 3.68 g of 5-methoxy-2,3,3-trimethyl-3H-indole (19.5 mmol, 71%). Storing the
product at r.t. seemed to lead to decomposition over time.
TLC: Rf = 0.29 (Hex/EtOAc = 7/3) [UV].
1
H NMR (400 MHz, CDCl3): δ [ppm] = 1.30 (s, 6H, Hd), 2.29 (s, 3H, Ha), 3.83 (s, 3H, Hh), 6.79-6.85 (m, 2H, Hf, Hj),
7.48 (dd, 3J = 8.1 Hz, 4J = 0.8 Hz, 1H, Hi).
3-(5-Methoxy-2,3,3-trimethyl-3H-indol-1-ium-1-yl)propane-1-sulfonate (S4)
According to a modified literature procedure7: To a solution of propane 1,3-propane sultone (696 μL, 968 mg, 1.00 eq.)
in toluene anhydrous (42 mL), 5-methoxy-2,3,3-trimethyl-3H-indole (1.44 mL, 1.50 g, 1.00 eq.) was added under inert
atmosphere and the solution was heated to 135 °C for 5.5 days. After letting it cool to r.t. the formed dark purple
precipitate was collected by vacuum filtration and washed with petroleum ether. The solid was dried in vacuo and
resulted in 2.06 g of 3-(5-methoxy-2,3,3-trimethyl-3H-indol-1-ium-1-yl)propane-1-sulfonate (6.61 mmol, 83%) as dark
purple solid and used without further purification in the consecutive reaction.
1
H NMR (300 MHz, DMSO-d6): δ [ppm] = 1.51 (s, 6 H, Hd), 2.14 (virt. p, 3J ≈ 3J ≈ 7.2 Hz, 2H, Hm), 2.63 (t,
3
J = 6.5 Hz, 2H, Hl), 2.77 (s, 3H, Ha), 3.85 (s, 3H, Hh), 4.60 (t, 3J = 7.9 Hz, 2H, Hn), 7.13 (dd, 3J = 8.8 Hz, 4J = 2.5 Hz,
1H, Hi), 7.45 (d, 4J = 2.5 Hz, 1H, Hf), 7.94 (d, 3J = 8.8 Hz, 1H, Hj).
13
C NMR (101 MHz, CDCl3): δ [ppm] = 13.5 (Cb), 22.1 (Cd), 23.9 (Cm), 46.6 (Cl/n), 47.4 (Cl/n), 53.9 (Cc), 56.1 (Ch),
109.4 (Cf), 114.3 (Ci), 116.4 (Cj), 134.4 (Ck), 144.0 (Ce), 160.6 (Cg), 193.6 (Cb).
3
5-Methoxy-2,3,3-trimethyl-1-(3-(trimethylammonio)propyl)-3H-indol-1-ium dibromide (S5)
According to a modified literature procedure8: To a solution of (3-bromopropyl)trimethyl-ammonium dibromide

(2.28 g, 8.72 mmol, 1.10 eq) in acetonitrile (45 mL) under inert atmosphere, 5-methoxy-2,3,3-trimethyl-3H-indole
(2.28 g, 7.93 mmol, 1.00 eq.) was added. The reaction mixture was heated to 100 °C for six days. After letting the
solution cool down to r.t. the solvent was removed in vacuo. The resulting solid was dissolved in water (200 mL) and
extracted with diethyl ether (4 × 50 mL). After desiccation of the water phase in vacuo it was dissolved in MeOH
(15 mL) and slowly added to diethyl ether (400 mL) at 0 °C. The formed precipitate was collected resulting in a very
dark purple solid as a mixture of 5-methoxy-2,3,3-trimethyl-1-(3-(trimethylammonio)propyl)-3H-indol-1-ium
dibromide (3.05 g, 55.7 w%, 48%) and (3-bromopropyl)trimethyl-ammonium dibromide and was used without further
purification in the consecutive reaction.
Similar solubilities of the product and the starting material ((3-bromopropyl)trimethyl-ammonium dibromide) and
limited solubility in organic solvents made further purification difficult.
1
H NMR (300 MHz, DMSO-d6): δ [ppm] = 1.55 (s, 6H, Hd), 2.39-2.32 (m, 2H, Hm), 2.88 (s, 3H, Ha), 3.13 (s, 9H,
NMe3), 3.62-3.71 (m, 2H, Hn/l), 3.87 (s, 3H, Hh), 4.48 (t, 3J = 8.0 Hz, 2H, Hn/l), 7.16 (dd, 3J = 8.9 Hz, 4J = 2.5 Hz, 1H,
Hi), 7.52 (d, 3J = 2.5 Hz, 1H, Hf), 8.03 (d, 3J = 8.9 Hz, 1H, Hj).
S2.2. Synthesis of merocyanines/spiropyrans

General procedure 1 (GP1): Propylsulfonate Merocyanine/Spiropyran Synthesis
According to a modified literature procedure9: A slight excess of the corresponding aldehyde was added to a solution of
propylsulfonate substituted indolinium precursor in anhydrous ethanol (0.14-0.21 M, absolute, undenatured over 3 Å
molecular sieves). The reaction mixture was refluxed overnight. After letting the solution cool down to r.t. (or cooled
down further to 0 °C) the formed precipitate was collected by vacuum filtration, washed with (ice cold) ethanol and
dried in vacuo.
General procedure 2 (GP2): Propylammonium Merocyanine/Spiropyran Synthesis

The corresponding aldehyde (1.20 eq.) was added to a solution of propylammonium substituted indolinium precursor in
anhydrous ethanol (0.18 M, absolute, undenatured over 3 Å molecular sieves, under inert atmosphere. The reaction
mixture was refluxed overnight. After letting the solution cool down to r.t. the solvent was removed in vacuo. The crude
product was purified by column chromatography (CH2Cl2/MeOH) to give the alkylammonium substituted
merocyanine/spiropyran.
4
Synthesis of SP-2a/MCH-2a
(E)-3-(2-(2-Hydroxystyryl)-3,3-dimethyl-3H-indol-1-ium-1-yl)propane-1-sulfonate (MCH-2a)/ (R/S)-3-(3',3'-
dimethylspiro[chromene-2,2'-indolin]-1'-yl)propane-1-sulfonate (SP-2a)
Following GP1, salicylaldehyde (41.7 μL, 47.7 mg, 391 μmol, 1.44 eq.) was added to a solution of the substituted
indolinium (100 mg, ca. 76.2 w%, 271 μmol, 1.00 eq.) in anhydrous ethanol (136 mM, 2.00 mL). The reaction mixture
was heated to 95 °C for 17 h and let cool down to room temperature. The resulting orange precipitate was collected by
vacuum filtration, washed with ethanol (ca. 2 mL) and dried in vacuo giving an orange powder as merocyanine MCH-
2a/spiropyran SP-2a (80.7 mg, 209 μmol, 77%). In solution (DMSO-d6) 2a is present as a mixture of MCH-2a and SP-
2a in a ratio of 2:1.
Merocyanine form
1
H NMR (600 MHz, DMSO) δ 11.04 (s, 1H, OH), 8.60 (d, J = 16.4 Hz, 1H, Hg), 8.28 (d, J = 7.9 Hz, 1H, He), 8.02 (d, J
= 7.7 Hz, 1H, Hp), 7.88 (d, J = 16.5 Hz, 1H, Hh), 7.86 (d, J = 7.5 Hz, 1H, Hm), 7.66 – 7.61 (m, 1H, Ho), 7.63 – 7.58 (m,
1H, Hn), 7.47 (apparent t, J = 7.7 Hz, 1H, Hc), 7.04 (d, J = 8.4 Hz, 1H, Hb), 6.99 (d, J = 7.3 Hz, 1H, Hd), 4.83 – 4.78 (m,
2H, Hr), 2.66 (t, J = 6.4 Hz, 2H, Ht), 2.18 (p, J = 7.0 Hz, 2H, Hs), 1.77 (s, 6H, Hk+k’).
13
C NMR (151 MHz, DMSO) δ 181.8 (Ci), 159.0 (Ca), 148.7 (Cg), 143.5 (Cl), 140.9 (Cq), 135.8 (Cc), 129.8 (Ce), 129.15
(Cn/o), 129.13 (Cn/o), 123.0 (Cm), 121.3 (Cf), 120.1 (Cd), 116.6 (Cb), 115.1 (Cp), 111.5 (Ch), 51.9 (Cj), 47.3(Ct), 45.5 (Cr),
26.4 (Ck+k’), 24.6 (Cs).
Spiropyran form
1
H NMR (600 MHz, DMSO) δ 7.15 (d, J = 7.4 Hz, 1H, He), 7.10 – 7.05 (m, 3H, Hc+o+m), 6.98 (d, J = 15.7 Hz, 1H, Hg),
6.81 (t, J = 7.4 Hz, 1H, Hd), 6.73 (t, J = 7.2 Hz, 1H, Hn), 6.63 (apparent d, J = 8.0 Hz, 2H, Hb+p), 5.74 (d, J = 10.2 Hz,
1H, Hh), 3.25 – 3.12 (m, 2H, Hr), 2.49 – 2.38 (m, 2H, Ht), 1.92 – 1.78 (m, 2H, Hs), 1.18 (s, 3H, Hk/k’), 1.07 (s, 3H, Hk/k’).
13
C NMR (151 MHz, DMSO) δ 153.8 (Ca), 147.3 (Cq), 136.0 (Cl), 129.7 (Co/m), 129.2 (Cg), 127.4 (Cc), 126.9 (Ce),
121.5 (Co/m), 120.1 (Cd), 119.7 (Ch), 118.5 (Cf), 118.4 (Cn), 114.4 (Cb), 106.3 (Cp), 104.2 (Ci), 51.8 (Cj), 49.2 (Ct), 42.3
(Cr), 25.7 (Ck/k’), 24.9 (Cs), 19.8 (Ck/k’).
5
Synthesis of SP-2b/MCH-2b
(E)-3-(2-(5-(tert-Butyl)-2-hydroxystyryl)-3,3-dimethyl-3H-indol-1-ium-1-yl)pro-pane-1-sulfonate (MCH-2b)/(R/S)-3-(6-
(tert-butyl)-3',3'-dimethylspiro[chromene-2,2'-indolin]-1'-yl)propane-1-sulfonate (SP-2b)
k
k
m v
l u
n
j i h e
d
o f
q N g
c
p
r
a b
s
HO
t
O3S
N + C25H31NO4S
k’
441.59 g/mol
k
m h g
l
n
j f e v
O3S o
i
u
p
q N O a
d
r
b c
s
O3S + H+
Following GP1, 5-(tert-butyl)-2-hydroxybenzaldehyde (209 mg, 1.17 mmol, 1.10 eq.) was added to a solution of the
substituted indolinium (300 mg, 1.07 mmol, 1.00 eq.) in anhydrous ethanol (214 mM, 5.00 mL). The reaction mixture
was heated to 95 °C for 20 h and let cool down to room temperature. As no precipitate was visible the solution was
stored in the freezer overnight. The resulting orange precipitate was collected by vacuum filtration, washed with ice
cold ethanol (ca. 2 mL) and dried in vacuo giving an orange powder as merocyanine MCH-2b/spiropyran SP-2b
(252 mg, 571 μmol, 53%). In solution (DMSO-d6) 2b is present as a mixture of MCH-2b and SP-2b in a ratio of 2:1.
Merocyanine form
1
= 7.4 Hz, 1H, Hp), 7.96 (d, J = 16.3 Hz, 1H, Hh), 7.90 – 7.83 (m, 1H, Hm), 7.66 – 7.58 (m, 2H, Hn+o), 7.51 (dd, J = 8.7,
2.4 Hz, 1H, Hc), 6.97 (d, J = 8.6 Hz, 1H, Hb), 4.92 – 4.77 (m, 2H, Hr), 2.66 – 2.60 (m, 2H, Ht), 2.20 (p, J = 7.6 Hz,
2H, Hs), 1.78 (s, 6H, Hk+k'), 1.33 (s, 9H, Hv).
13
C{1H} NMR (151 MHz, DMSO) δ 182.0 (Ci) 157.2 (Ca), 149.8 (Cg), 143.5 (Cl), 142.7 (Cd), 140.9 (Cq), 133.1 (Cq),
129.1 (Cn), 129.0 (Co), 126.9 (Ce), 123.0 (Cm), 120.7 (Cf), 116.2 (Cb), 115.0 (Cp), 111.4 (Ch), 51.9 (Cj), 47.2 (Ct), 45.4
(Cr), 34.2 (Cn), 31.3 (Cv), 26.5 (Ck+k'), 24.5 (Cs).
Spiropyran form
1
H NMR (600 MHz, DMSO) δ 7.16 (d, J = 2.4 Hz, 1H, He), 7.09 (dd, J = 8.5, 2.5 Hz, 1H, Hc), 7.07 (dd, J = 7.6, 1.3
Hz, 1H, Ho), 7.05 (d, J = 7.1 Hz, 1H, Hm), 6.98 (d, J = 10.1 Hz, 1H, Hd), 6.72 (t, J = 7.3 Hz, 1H, Hn), 6.61 (d, J = 7.7
Hz, 1H, Ho), 6.56 (d, J = 8.5 Hz, 1H, Hb), 5.70 (d, J = 10.2 Hz, 1H, Hh), 3.22 – 3.08 (m, 2H, Hr), 2.47 – 2.32 (m, 2H,
Hs), 1.24 (s, 9H, Hv), 1.17 (s, 3H, Hk/k'), 1.07 (s, 1H, Hk/k').
13
C{1H} NMR (126 MHz, DMSO) δ 151.6 (Ca), 147.5 (Cq), 142.3 (Cd), 136.4 (Cl) 129.6 (Cg), 127.4 (Cm/o), 126.6 (Cc),
123.7 (Ce), 121.5 (Cm/o), 119.6 (Ch), 118.3 (Cn), 117.7 (Cf), 113.0 (Cb), 106.2 (Cp), 104.1 (Ci), 51.8 (Cj), 49.2 (Ct), 42.5
(Cr), 33.8 (Cu), 31.3 (Cv), 25.7 (Ck/k'), 24.8 (Cs), 19.7 (Ck/k').
13
C NMR data assigned using a sample irradiated with 455 nm.
HRMS (ESI): calc. for C25H31NO4S: [M−Na] +: 464.1866 [2M+Na] +: 905.3840

found: [M−Na]+: 464.1864 [2M+Na]+: 905.3842.
6
Synthesis of SP-2c/MCH-2c
(E)-3-(2-(5-(tert-Butyl)-2-hydroxystyryl)-5-methoxy-3,3-dimethyl-3H-indol-1-ium-1-yl)propane-1-sulfonate (MCH-2c)/
(R/S)-3-(6-(tert-butyl)-5'-methoxy-3',3'-dimethylspiro[chromene-2,2'-indo-lin]-1'-yl)propane-1-sulfonate (SP-2c)
k
k
m
O n l u
v
w j i h e
d
o f
q N g
c
O p
r
a b
s
HO
N t C26H33NO5S
O3S 471.61 g/mol
+
k’
k
O3S m h g
O n l
w j f e v
i
o u
p
q N O a
d
r
b c
s
O3S + H+
Following GP1, 5-(tert-butyl)-2-hydroxybenzaldehyde (242 μL, 252 mg, 1.41 mmol, 1.10 eq.) was added to a solution
of the substituted indolinium (400 mg, 1.28 mmol, 1.00 eq.) in anhydrous ethanol (214 mM, 6.00 mL). The reaction
mixture was heated to 100 °C for 26 h, let cool down to room temperature and briefly cooled to 0 °C in an ice bath. The
orange precipitate was collected by vacuum filtration, washed with ice cold pentane (ca. 5 mL), ethanol (ca. 2 mL) and
dried in vacuo giving an orange powder as merocyanine MCH-2c/spiropyran SP-2c (398 mg, 845 μmol, 66%). In
solution (DMSO-d6) 3 is present as a mixture of MCH-2c and SP-2c in a ratio of 100:14.
Merocyanine form:
1
= 8.9 Hz, 1H, Hp), 7.89 (d, J = 16.4 Hz, 1H, Hh), 7.51 (d, J = 2.5 Hz, 1H, Hm), 7.47 (dd, J = 8.7, 2.4 Hz, 1H, Hc), 7.16
(dd, J = 8.9, 2.5 Hz, 1H, Ho), 6.95 (d, J = 8.6 Hz, 1H, Hb), 4.78 (t, J = 7.8 Hz, 2H, Hr), 3.89 (s, 3H, Hw), 2.63 – 2.57 (m,
2H, Ht), 2.17 (p, J = 7.1 Hz, 2H, Hs), 1.77 (s, 6H, Hk+k'), 1.32 (s, 9H, Hv).
13
C{1H} NMR (151 MHz, DMSO) δ 179.7 (Ci), 160.6 (Cn), 156.8 (Ca), 147.9 (Cg), 145.7 (Cl), 142.6 (Cd), 134.2 (Cq),
132.5 (Cc), 126.7 (Ce), 120.8 (Cf), 116.10 (Cp/b), 116.07 (Cp/b), 114.8 (Co), 111.5 (Ch), 108.8 (Cm), 56.1 (Cw), 51.8 (Cj),
47.2 (Ct), 45.5 (Cr), 34.2 (Cu), 31.3 (Cv), 26.5 (Ck+k'), 24.5 (Cs).
Spiropyran form:
13 % of the photoswitch was present in the spiropyran form. The 1H NMR signals were assigned by analogy to signals
from the other SPs. 13C NMR signals were too weak to accurately assign.
1
H NMR (600 MHz, DMSO) δ 7.15 (d, overlapped, 1H, He), 7.09 (dd, J = 8.6, 1.9 Hz, 1H, Hc), 6.95 (d, overlapped,
1H, Hg), 6.72 (d, J = 1.8 Hz, 1H, Hm), 6.63 (dd, J = 8.1, 2.0 Hz, 1H, Ho), 6.54 (d, J = 8.4 Hz, 1H, Hb), 6.51 (d, J = 8.3
Hz, 1H, Hp), 5.68 (d, J = 10.2 Hz, 1H, Hh), 3.69 (s, 3H, Hw), 3.19 – 3.01 (m, 2H, Hr), 2.46 – 2.33 (m, 2H, Ht), 1.86 –
1.79 (m, 2H, Hs), 1.23 (s, 9H, Hv), 1.16 (s, 3H, Hk/k'), 1.07 (s, 3H, Hk/k').
HRMS (ESI): calc. for C26H33NO5S: [M−Na] +: 494.1971

found: [M−Na]+: 494.1971.
7
Synthesis of SP-2d/MCH-2d
(E)-3-(2-(2-hydroxy-5-nitrostyryl)-3,3-dimethyl-3H-indol-1-ium-1-yl)propane-1-sulfonate (MCH-2d)/ (R/S)-3-(3',3'-
dimethyl-6-nitrospiro[chromene-2,2'-indolin]-1'-yl)propane-1-sulfonate (SP-2d)
Following GP1, 2-hydroxy-5-nitrobenzaldehyde (196 mg, 1.17 mmol, 1.10 eq.) was added to a solution of the
substituted indolinium (300 mg, 1.07 mmol, 1.00 eq.) in anhydrous ethanol (178 mM, 6.00 mL). The reaction mixture
was heated to 100 °C for 22 h and let cool down to room temperature. The brown precipitate was collected by vacuum
filtration, washed with ice cold ethanol (ca. 4 mL) and dried in vacuo giving a brown powder as merocyanine MCH-
2d/spiropyran SP-2d (213 mg, 495 μmol, 46%). In solution (DMSO-d6) 3 is present as a mixture of MCH-2d and SP-
2d in a ratio of 100:4.
Merocyanine form:
1
H NMR (600 MHz, DMSO) δ 12.7 (br, 1H, OH), 9.09 (d, J = 2.8 Hz, 1H, He), 8.50 (d, J = 16.5 Hz, 1H, Hg), 8.30 (dd,
J = 9.1, 2.8 Hz, 1H, Hc), 8.11 – 8.09 (m, 1H, Hp), 8.07 (d, J = 16.5 Hz, 1H, Hh), 7.91 – 7.87 (m, 1H, Hm), 7.69 – 7.62
(m, 2H, Hn+o), 7.21 (d, J = 9.2 Hz, 1H, Hb), 4.86 – 4.82 (m, 2H, Hr), 2.65 (t, J = 6.8 Hz, 2H, Ht), 2.26 – 2.17 (m, 2H,
Hs), 1.80 (s, 6H, Hk+k’).
13
C{1H} NMR (151 MHz, DMSO) δ 182.2 (Ci), 164.1 (Ca), 147.3 (Cg), 143.9 (Cl), 140.9 (Cq), 140.3 (Cd), 129.7 (Cn),
129.5 (Cc), 129.3 (Co), 127.4 (Ce), 123.1 (Cm), 121.4 (Cf), 117.3 (Cb), 115.6 (Cp), 114.8 (Ch), 52.4 (Ci), 47.6 (Ct), 46.2
(Cr), 25.9 (Ck+k’), 24.7 (Cs).
8
Synthesis of SP-3a/MCH-3a
(E)-2-(2-Hydroxystyryl)-3,3-dimethyl-1-(3-(trimethylammonio)propyl)-3H-indolium bromide (MCH-3a)/(R/S)-3-(3',3'-
dimethylspiro[chromene-2,2'-indolin]-1'-yl)-N,N,N-trimethylpropan-1-aminium bromide (SP-3a)
Following GP2, salicylaldehyde (127 μL, 146 mg, 1.20 mmol, 1.20 eq) was added to a solution of N,N,N-trimethyl-3-
(2,3,3-trimethyl-3H-1λ4-indol-1-yl)propan-1-aminium dibromide (54.4 w%, 772 mg, 999 μmol, 1.00 eq.) in ethanol
anhydrous (5.6 mL) under inert atmosphere. The reaction mixture was heated to 100 °C for 20 h. After letting the
solution cool down to r.t. the solvent was removed in vacuo. The crude solid was purified by column chromatography
(4.5 × 10 cm, CH2Cl2/MeOH = 95/5 → 90/10, 6 × 6 cm, CH2Cl2/MeOH = 95/5 → 93/7) resulting in an orange solid as
merocyanine MCH-3a/spiropyran SP-3a (248 mg, 475 μmol, 52%). In solution (DMSO-d6) 3a is present only as SP-
3a.
TLC: Rf = 0.01 (CH2Cl2/MeOH = 95/5) [UV].
Spiropyran form:
1
H NMR (600 MHz, DMSO) δ 7.19 (dd, J = 7.5, 1.7 Hz, 1H, He), 7.15 – 7.06 (m, 3H, Hc+o+m ), 7.02 (d, J = 10 Hz, 1H,
Hg), 6.84 (td, J = 7.5, 1.1 Hz, 1H, Hd), 6.79 (td, J = 7.4, 0.9 Hz, 1H, Hn), 6.68 (dt, J = 8.1, 0.7 Hz, 1H, Hb), 6.67 (dt, J =
7.6, 0.7 Hz, 1H, Hp), 5.85 (d, J = 10 Hz, 1H, Hh), 3.39 – 3.26 (m, 2H, Ht/t'), 3.32 – 3.26 (m, 1H, Ht/t'), 3.22 (dt, J = 15,
7.4 Hz, 1H, Hr/r'), 3.13 (ddd, J = 15, 7.3, 5.6 Hz, 1H, Hr/r'), 3.02 (s, 9H, Hu), 2.08 – 1.98 (m, 1H, Hs/s'), 1.96 – 1.86 (m,
1H, Hs/s'), 1.21 (s, 3H, Hk/k'), 1.12 (s, 3H, Hk/k').
13
C{1H} NMR (151 MHz, DMSO) δ 153.5 (Ca), 146.9 (Cq), 136.2 (Cl), 129.9 (Cc), 129.4 (Cg), 127.4 (Co), 127.0 (Ce),
121.7 (Cm), 120.4 (Cd), 119.4 (Ch), 119.0 (Cn), 118.3 (Cf), 114.5 (Cb), 106.5 (Cp), 104.2 (Ci), 63.5 (t, JN-C = 3.2 Hz, Ct),
52.2 (apparent t, J = 3.8 Hz, NMe3), 51.7 (Cj), 40.3 (Cr), 25.8 (Ck/k'), 22.2 (Cs), 19.8 (Ck/k').
HRMS (ESI): calc. for C24H31N2O+: [M] +: 363.2431

found: [M]+: 363.2432.
9
Synthesis of SP-3b/MCH-3b
(E)-2-(5-(tert-Butyl)-2-hydroxystyryl)-3,3-dimethyl-1-(3-(trimethylammonio)pro-pyl)-3H-indol-1-ium dibromide
(MCH-3b)/(R/S)-3-(6-(tert-butyl)-3',3'-dimethylspiro-[chromene-2,2'-indolin]-1'-yl)-N,N,N-trimethylpropan-1-aminium
bromide (SP-3b)
Following GP2, 5-(tert-butyl)-2-hydroxybenzaldehyde (198 μL, 206 mg, 1.16 mmol, 1.20 eq) was added to a solution
of N,N,N-trimethyl-3-(2,3,3-trimethyl-3H-1λ4-indol-1-yl)propan-1-aminium dibromide (54.4 w%, 744 mg, 962 μmol,
1.00 eq.) in ethanol anhydrous (5.4 mL) under inert atmosphere. The reaction mixture was heated to 100 °C for 20 h.
After letting the solution cool down to r.t. the solvent was removed in vacuo. The resulting solid was purified by
column chromatography (4.5 × 7 cm, CH2Cl2/MeOH = 95/5 → 90/10, 6 × 7 cm, CH2Cl2/MeOH = 95/5 → 90/10)
resulting in a light yellow solid as merocyanine MCH-3b/spiropyran SP-3b (207 mg, 395 μmol, 44%). In solution
(DMSO-d6) 3b is present only as SP-3b.
Spiropyran form:
1
H NMR (600 MHz, DMSO) δ 7.20 (d, J = 2.5 Hz, 1H, He), 7.15 – 7.08 (m, 3H, Hc+o+m), 7.02 (d, J = 10 Hz, 1H, Hg),
6.79 (td, J = 7.4, 0.9 Hz, 1H, Hn), 6.65 (d, J = 7.7 Hz, 1H, Hp), 6.61 (d, J = 8.5 Hz, 1H, Hb), 5.81 (d, J = 10 Hz, 1H, Hh),
3.38 – 3.30 (m, 1H, Ht/t'), 3.33 (s, 3H, Hw), 3.27 (td, J = 12.4, 4.8 Hz, 1H, Ht/t'), 3.20 (dt, J = 14.7, 7.4 Hz, 1H, Hr/r'),
3.10 (ddd, J = 14.8, 7.3, 5.6 Hz, 1H, Hr/r'), 3.00 (s, 9H, NMe3), 2.02 (dh, J = 12.2, 7.0 Hz, 1H, Hs/s'), 1.89 (dh, J = 12.5,
5.9 Hz, 1H, Hs/s'), 1.24 (s, 9H, Hv), 1.20 (s, 3H, Hk/k'), 1.11 (s, 3H, Hk/k').
13
C{1H} NMR (151 MHz, DMSO) δ 151.3 (Ca), 147.0 (Cq), 142.5 (Cd), 136.3 (Cl), 129.9 (Cg), 127.4 (Cm/o), 126.7 (Cc),
123.8 (Ce), 121.7 (Cm/o), 119.04 (Ch/n), 118.97 (Ch/n), 117.5 (Cf), 113.9 (Cb), 106.5 (Cp), 104.1 (Ci), 63.5 (apparent t,
J = 3.2 Hz, Ct), 52.2 (apparent t, J = 3.7, NMe3), 51.6 (Cj), 40.3 (Cr), 33.8 (Cu), 31.3 (Cv), 25.9 (Ck/k'), 22.2 (Cs), 19.9
(Ck/k').
HRMS (ESI): calc. for C28H39N2O+: [M] +: 419.3057

found: [M]+: 419.3052.
10
Synthesis of SP-3c/MCH-3c
(E)-2-(5-(tert-Butyl)-2-hydroxystyryl)-5-methoxy-3,3-dimethyl-1-(3-(trimethyl-ammonio)propyl)-3H-indol-1-ium
dibromide (MCH-3c)/(R/S)-3-(6-(tert-butyl)-5'-methoxy-3',3'-dimethylspiro[chromene-2,2'-indolin]-1'-yl)-N,N,N-
trimethylpropan-1-aminium bromide (SP-3c)
Following GP2, 5-tert-butyl-2-hydroxybenzaldehyde (178 μL, 185 mg, 1.04 mmol, 1.20 eq) was added to a solution of
5-methoxy-2,3,3-trimethyl-1-(3-(trimethylammonio)propyl)-3H-indol-1-ium dibromide (55.6 w%, 700 mg, 864 μmol,
1.00 eq.) in ethanol anhydrous (4.9 mL) under inert atmosphere. The reaction mixture was heated to 100 °C for 26 h.
After letting the solution cool down to r.t. the solvent was removed in vacuo. The crude solid was purified by column
chromatography (7 × 13 cm, CH2Cl2/MeOH = 97/3 → 90/10) resulting in an orange solid as merocyanine MCH-
3c/spiropyran SP-3c (300 mg, 541 μmol, 63%). In solution (DMSO-d6) 3c is present only as SP-3c.
Spiropyran form
1
H NMR (600 MHz, DMSO) δ 7.18 (d, J = 2.5 Hz, 1H, He), 7.11 (dd, J = 8.5, 2.5 Hz, 1H, Hc), 7.00 (d, J = 10.2 Hz,
1H, Hg), 6.76 (d, J = 2.6 Hz, 1H, Hm), 6.66 (dd, J = 8.4, 2.6 Hz, 1H, Ho), 6.58 (d, J = 8.7 Hz, 1H, Hb), 6.56 (d, J = 8.6
Hz, 1H, Hp), 5.79 (d, J = 10.2 Hz, 1H, Hh), 3.69 (s, 3H, Hw), 3.33 (td, J = 12.3, 4.7 Hz, 1H, Ht), 3.26 (td, J = 12.4, 4.8
Hz, 1H, Ht'), 3.17 – 3.10 (m, 1H, Hr), 3.06 – 3.01 (m, 1H, Hr'), 3.00 (s, 9H, NMe3), 2.00 (tq, J = 12.4, 6.7 Hz, 1H, Hs),
1.88 (tq, J = 12.4, 6.5, 5.5 Hz, 1H, Hs'), 1.23 (s, 9H, Hv), 1.17 (s, 3H, Hk/k'), 1.11 (s, 3H, Hk/k').
13
C{1H} NMR (151 MHz, DMSO) δ 153.6 (Cn), 151.5 (Ca), 142.5 (Cd), 141.2 (Cq), 138.0 (Cl), 129.9 (Cg), 126.7 (Cn),
123.9 (Ce), 119.2 (Ch), 117.7 (Cf), 113.9 (Cb), 111.4 (Co), 109.6 (Cm), 106.8 (Cp), 104.7 (Ci), 63.7 (apparent t, J = 2.5
Hz, Ct), 55.7 (Cw), 52.3 (apparent t, J = 3.8 Hz, NMe3), 51.9 (Cj), 40.8 (Cr), 33.8 (Cu), 31.4 (Cv), 25.8 (Ck/k'), 22.3 (Cs),
19.8 (Ck/k').
HRMS (ESI): calc. for C29H41N2O2+: [M] +: 449.3163

found: [M]+: 449.3161.
11
Synthesis of SP-3d/MCH-3d
(E)-2-(2-Hydroxystyryl)-5-methoxy-3,3-dimethyl-1-(3-(trimethylammonio)propyl)-3H-indol-1-ium dibromide (MCH-
3d)/(R/S)-3-(5'-methoxy-3',3'-dimethylspiro-[chromene-2,2'-indolin]-1'-yl)-N,N,N-trimethylpropan-1-aminium bromide
(SP-3d)
Following GP2, salicylaldehyde (94.4 μL, 109 mg, 889 μmol, 1.20 eq) was added to a solution of 5-methoxy-2,3,3-
trimethyl-1-(3-(trimethylammonio)propyl)-3H-indol-1-ium dibromide (55.6 w%, 600 mg, 741 μmol, 1.00 eq.) in
ethanol anhydrous (4.2 mL) under inert atmosphere. The reaction mixture was heated to 100 °C for 18 h. After letting
the solution cool down to r.t. the solvent was removed in vacuo. The crude solid was purified by column
chromatography (7 × 15 cm, CH2Cl2/MeOH = 95/5 → 93/7 → 90/10) resulting in an orange solid as merocyanine
MCH-3d/spiropyran SP-3d (93.3 mg, 168 μmol, 23%). An additional 254 mg (458 μmol, 62%) were obtained in
slightly less pure form. In solution (DMSO-d6) 3d is present only as SP-3d.
Spiropyran form
1
H NMR (600 MHz, DMSO) δ 7.17 (dd, J = 7.5, 1.7 Hz, 1H, He), 7.10 (ddd, J = 8.1, 7.4, 1.7 Hz, 1H, Hc), 7.00 (dd, J =
10.4, 0.7 Hz, 1H, Hg), 6.83 (td, J = 7.4, 1.1 Hz, 1H, Hd), 6.78 (d, J = 2.6 Hz, 1H, Hm), 6.70 – 6.63 (m, 2H, Ho+b), 6.57 (d,
J = 8.4 Hz, 1H, Hp), 5.82 (d, J = 10.2 Hz, 1H, Hh), 3.70 (s, 3H, Hw), 3.30 – 3.23 (m, 1H, Hr+r'), 3.15 (dt, J = 14.7, 7.4
Hz, 1H, Ht/t'), 3.09 – 3.02 (m, 1H, Ht/t'), 3.01 (s, 9H, NMe3), 2.01 (tq, J = 12.3, 6.5 Hz, 1H, Hs/s'), 1.89 (tq, J = 12.3, 5.8
Hz, 1H, Hs/s'), 1.19 (s, 3H, Hk/k'), 1.12 (s, 3H, Hk/k').
13
C{1H} NMR (151 MHz, DMSO) δ 153.7 (Ca/n), 153.5 (Ca/n), 141.1 (Cq), 137.8 (Cl), 129.9 (Cc), 129.3 (Cg), 127.0 (Ce),
120.3 (Cd), 119.5 (Ch), 118.4 (Cf), 114.4 (Cb), 111.4 (Co), 109.6 (Cm), 106.7 (Cp), 104.7 (Ci), 63.6 (apparent t, J = 2.5
Hz, Ct), 52.2 (apparent t, J = 3.6, NMe3), 51.9 (Cj), 40.7 (Cr), 25.7 (Ck/k'), 22.2 (Cs), 19.7 (Ck/k').
HRMS (ESI): calc. for C25H33N2O2+: [M] +: 393.2537

found: [M]+: 393.2539.
12
S3. NMR spectra of compounds
S3.1. Starting materials and intermediates
3-(2,3,3-Trimethyl-3H-indol-1-ium-1-yl)propane-1-sulfonate (S1)
1
H NMR (400 MHz, DMSO-d6):
2,3,3-Trimethyl-1-(3-(trimethylammonio)propyl)-3H-indol-1-ium dibromide (S2)
1
13
Starting material: (3-bromopropyl)trimethyl-ammonium bromide
1
5-Methoxy-2,3,3-trimethyl-3H-indole (S3)
1
H NMR (400 MHz, CDCl3):
14
3-(5-Methoxy-2,3,3-trimethyl-3H-indol-1-ium-1-yl)propane-1-sulfonate (S4)
1
13
C NMR (101 MHz, DMSO-d6):
15
5-Methoxy-2,3,3-trimethyl-1-(3-(trimethylammonio)propyl)-3H-indol-1-ium dibromide (S5)
1
16
S3.1. NMR spectra of photoswitches 2a–c, 3a–d
1
H NMR (600 MHz) spectra of 2a–c, 3a–d in DMSO-d6
17
13
C{1H} (151 MHz) NMR spectra of 2a–c, 3a–d in DMSO-d6
18
MCH-2a/SP-2a
1
H NMR (600 MHz, DMSO-d6): Red assignments are for MC/MCH and blue are for SP.
19
13
20
Superimposed 1H-13C HSQC (red/blue) and 1H-13C HMBC (black) (600 MHz, DMSO-d6). Peak labels are from the
HSQC with MC/MCH (red) and SP (blue). The 1H-13C HMBC couplings are very similar across all compounds and
allow reliable assignment of all 13C signals.
21
MCH-2b/SP-2b
1
13
22
MCH-2c/SP-2c
1
13
23
MCH-2d/SP-2d
1
13
24
SP-3a
1
13
25
SP-3b
1
13
26
SP-3c
1
13
27
SP-3d
1
13
28
S4. In situ irradiation 1H NMR spectra
For MCH/SP products which had low ratios of SP present in solution (DMSO-d6) in-situ NMR irradiation allowed
full assignment of the SP signals. A fibre optic cable (Thorlabs, FT1500UMT) attached to a LUXEON Rebel
colour line LED light source (LXML-PR02-A900, Royal Blue) mounted to a heatsink was fed into a screw cap NMR
tube and inserted into the NMR spectrometer. The end of the fiber optic cable (ca. 3-4 cm) was roughened with
sandpaper to ensure light emission across the entire sample volume. This setup allowed light irradiation during NMR
acquisition. At the PSS the samples consisted only of the SP form.
SP-2a
1
H NMR (500 MHz, DMSO-d6) before and after 455 nm irradiation:
29
SP-2b
1
13
C{1H} NMR (151 MHz top, 125 MHz bottom, DMSO-d6) with and without irradiation (separate samples):
30
S5. Concentration dependent NMR of 3c
In order to study the concentration dependent behavior of compound 3c we chose to study a pH value at which both SP
and MC(H) could be observed. Samples consisted of a mixture of H2O/D2O (6/4) with a 20 mM concentration of
phosphate buffer (Table S1).
Table S1. Concentration of 3c and pH/pD of NMR samples.
Sample # c / mM pH/pD
1 0.1 5.5
2 1.0 5.8
3 2.5 6.2
4 5.0 6.2
5 10 6.2
Figure S1. MCH-3c and SP-3c structures and NMR stack of samples #1-5 with increasing concentrations from top to bottom: a) 0.1
mM, b) 1 mM, c) 2.5 mM, d) 5 mM, d) 10 mM in H2O/D2O (6/4, 20 mM phosphate buffer) at pH ~ 6. Signals are assigned for the
spiropyran form using COSY, HSQC and by comparison with data in DMSO-d6.
At pH ~ 6 the solution consisted of a mixture of SP-3c (blue) and MCH-3c (red) (Figure S1). Signals were assigned in
analogy to previous signal assignment in DMSO-d6. The characteristic signals of the double bond are those of SP (h)
and MCH (g) as the other signals are overlapped. Both species seem to aggregate above 1 mM concentrations with the
signals corresponding to SP showing a more significant shift than MCH signals.
31
1
H-13C-HSQC (600 MHz, D2O/H2O) with excitation sculpting for water suppression of 3c (5 mM in H2O/D2O = 6/4, 20
mM phosphate buffer at pH ~ 6). Values given in pink are for the 1H and 13C assignment of the corresponding signals in
DMSO-d6 showing that although the 1H signals do move significantly with concentration, the 13C signals are less
sensitive.
13
C{1H} (top) and DEPT135 (bottom) NMR spectra (151 MHz, D2O/H2O) of 5 mM 3c.
32
S6. UV-Vis studies
S6.1. General procedure
Buffer solutions ranging from pH 2.5-10 were prepared by adding KOH (10 M) to H3PO3 (100 mM) solutions and from
pH 0.5-2.5 by adding standard HCl. The pH of the prepared solutions was measured with a SPER benchtop meter with a
ThermoFisher Scientific OrionTM PerpHecTTM ROSSTM combination pH microelectrode.
MilliQ water was used for all solutions and degassed for 30 min prior to use (apart from the prepared buffer solutions).
Stock solutions of the spiropyrans/merocyanines (c ∼ 5 mM) were prepared with anhydrous methanol filtered through
syringe filters (0.45 μm) and stored in 10 mL and 25 mL volumetric flasks wrapped in aluminum foil at 4 °C.
Stock solutions were diluted with milliQ water to give 10% MeOH v/v (c ∼500 μM).
UV-vis samples were prepared as follows. In a 1 cm cuvette 600 μL of the corresponding phosphate buffer were
mixed with 1800 μL of milliQ water, put in the spectrometer and let equilibrate thermally for 5 min. The acquisition
was started and 450 μL of a premixed solution of milliQ water with 10% MeOH v/v dilution was added in situ, mixed
by gentle pipetting and continuous stirring of the solution. Scans were collected until no further noticeable change in the
spectrum occurred (5-45 min, pH and sample dependent). The last spectrum of the scanning collection was used as the
equilibrated spectrum in the dark.
For measurements of the cis-MCH and spiropyran forms, spectra were recorded under continuous irradiation
(λ = 455 nm). For most samples the equilibration was almost immediate and no change in the spectrum was observed
over time. The last spectrum of the scanning collection under irradiation was used as the equilibrated spectrum in the
light.
For compound 2d:

A stock solution (c ∼250 μM) was prepared with 20 mL of phosphate buffer (pH 2.0, 100 mM) and 80 mL milliQ water
and let equilibrate for a day. The stock solution was filtered prior to measurements. UV-vis samples were prepared by
adding 2 mL of milliQ water and 700 μL of the respective phosphate buffer to a 1 cm cuvette. The acquisition was
started and 300 μL of the filtered stock solution were added.
Variable temperature measurements at high pH

The general procedure was followed as described above. To ensure the premixed sample added to the buffer solution in
the cuvette was made up of predominantly the MCH form 50 µL of pH 4 buffer were added to the second dilution from
the stock solution and let equilibrate for > 30 min.
S6.2. Setup for visible irradiation experiments (λ = 455 nm)

UV-Vis absorbance spectroscopy experiments were performed on an Agilent Cary 60 Bio UV-visible
spectrophotometer equipped with a customized Cary Single Cell Peltier Accessory, maintaining the samples at 298 K.
The cell holder was modified to allow for irradiation perpendicular to the direction of measurement, as previously
described.10 A Luxeon Rebel LED was mounted on a heat sink positioned 4 cm away from the cell and driven using a
1000 mA LuxDrive PowerPuck. A Carclo 20.0 mm Fibre Coupling Lens was used to focus the beam. Samples were
stirred to ensure homogeneity and a timer relay module (FRM01) was used to control the irradiation cycles.
Table S2. Technical information for LUXEON Rebel color line LED light source (https://www.lumileds.com/wp-
content/uploads/files/DS68.pdf).
Descriptor Dominant wavelength / nm Radiometric power / mW

Test
LED maximum
current
part number emission Minimum Maximum Minimum Typical
/ mA
wavelength / nm
LXML-PR02-
A900 455 440 460 900 1030 700
(Royal Blue)
33
Figure S2. Emission spectrum of blue LED light source (λmax = 455 nm) for UV-vis irradiation experiments.
S6.3. Setup for UV irradiation experiments (λ = 265 nm)

Measurements of the time-dependent absorption changes at high pH were undertaken with a home-built setup, as
described below.
The sample was placed into a temperature-controlled holder (Quantum Northwest TC-1), with the temperature set as
mentioned in the text.
The probe light was centred at 560 nm. It was derived from the output of a broadband lamp (Energetiq LDLS) after
being passed through a monochromator (Spectral Products CM110).
The probe was modulated at 1 kHz using an optical chopper (Thorlabs MC1000A) and then split into two arms, one for
reference and the other for measurement. The reference was sent directly to a photodiode (Thorlabs PDA10A2), while
the measurement arm was passed through the sample before irradiating a photodiode (Thorlabs DET10A2). The signal
from each photodiode was amplified (Stanford Research Systems SR570) and measured by a lock-in amplifier
(Stanford Research Systems SR830) with the light intensity (Icalc) given by the ratio of the measurement arm to the
reference arm. For each sample, the absorption was calculated by comparing Icalc measured with the sample with respect
to the measured Icalc with only the buffer solution.
The sample was photo-switched by irradiating with a pump light; the irradiation length was varied based on the sample,
with the duty cycle (fraction of time the laser is on) of ~50%.
The pump light was 265 nm, 2 mW of LED light (Thorlabs, M265L3).
The best time-resolution achievable was ~2 ms; this was determined by the frequency of the chopper (1000 kHz). Each
scan consists of continually sampling the probe intensity while the pump is switched on (time depending on the sample)
and then off. The data points are timestamped with reference to the last event (either on/off). This is then repeated ~10-
100 times depending on the sample to give the final absorption vs time data presented.
34
S6.4. Determination of pKadark and pKahν
The methodology developed by Berton et al.11 was applied to spiropyran compounds 2a-d, 3a-d. Compound 2a served
as reference compound to compare to the reported data which is in excellent agreement.
To determine pKadark and pKahν values the equilibrium absorbance values (Aeq) were plotted over the range of pH. For
pKadark wavelengths characteristic for the MCH and MC forms were chosen and for pKahν wavelengths characteristic for
the SP and cis-MCH forms were chosen. The resulting sigmoidal curves were fit to following equation with AH being
the absorbance at high pH and AOH the absorbance at low pH, and a fitting factor P.11
𝐴𝐴𝐻𝐻 − 𝐴𝐴𝑂𝑂𝑂𝑂
𝐴𝐴𝑒𝑒𝑒𝑒 = 𝐴𝐴𝑂𝑂𝑂𝑂 + 𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑/ℎν
𝑝𝑝𝑝𝑝 − 𝑝𝑝𝑝𝑝𝑎𝑎 (Eq. S1)
1 + 𝑒𝑒𝑒𝑒𝑒𝑒 � �
𝑃𝑃
Fitting was done in Origin was done feeding in following initial values:
AOH - absorbance at lowest recorded pH
AH - absorbance at highest recorded pH
pKadark/hν approximate inflection point from plot
P = 1 (arbitrary)
Boundaries of the fitting parameters were set to:
0 ≤ pKaGS/hν ≤ 14
0 ≤ AOH ≤ 1, 0 ≤ AH ≤ 1.
pKadark and pKahν were obtained from the fit to Eq. S1 and are reported as an average of fits to two characteristic
absorbance bands of the respective species.
Note: A suitable pH range was chosen for the fit that resembled the expected sigmoidal curve. For some examples a few
values at very low pH were excluded from the fit as the recorded spectra at this pH were not entirely equilibrated. At
very low pH formation of cis-MCH is likely to slow down the equilibration process.
35
S6.5. pH-Dependent UV-visible absorption data for photoacids 2a–d, 3a–d
pH-Dependent UV-visible absorption data for photoacid 2a

a) b)
c) d)
Figure S3. UV-vis titrations of photoacid 2a. a) Equilibrated absorbance spectra at different pH values in the dark (orange: pH = 3.2,
yellow/purple: pH = 9.8). b) Change in absorbance maxima of 2a-MCH (orange, λ = 423 nm) and 2a-MC (purple, λ = 531 nm) over
the pH range with corresponding fits to Eq. S1 shown as solid lines. Experimental conditions: [2a] = 20.0 μM, [phosphate
buffer] = 20 mM, pH 3.2-9.8, T = 25 °C. c) Equilibrated absorbance spectra at different pH values (blue: pH =1.0, yellow: pH = 5.2)
under continuous blue light irradiation (λ = 455 nm). d) Change in absorbance maxima of 2a-SP (yellow, λ = 251 nm) and 2a-cis-
MCH (blue, λ = 377 nm) over the pH range with corresponding fits to Eq. S1 shown as solid lines. Experimental conditions: hν (λ =
455 nm), [2a] = 20.0 μM, [phosphate buffer] = 20 mM, pH 1.0-5.2, T = 25 °C.
36
pH-Dependent UV-visible absorption data for photoacid 2b
a) b)
c) d)
Figure S4. UV-vis titrations of photoacid 2b. a) Equilibrated absorbance spectra at different pH values in the dark (orange: pH = 3.7,
yellow/purple: pH = 9.8). b) Change in absorbance maxima of 2b-MCH (orange, λ = 436 nm) and 2b-MC (purple, λ = 546 nm) over
the pH range with corresponding fits to Eq. S1 shown as solid lines. Experimental conditions: [2b] = 20.5 μM, [phosphate
buffer] = 20 mM, pH 3.2-9.8, T = 25 °C. c) Equilibrated absorbance spectra at different pH values (blue: pH = 0.9, yellow: pH = 7.2)
under continuous blue light irradiation (λ = 455 nm). d) Change in absorbance maxima of 2b-SP (yellow, λ = 248 nm) and 2b-cis-
455 nm), [2b] = 20.5 μM, [phosphate buffer] = 20 mM, pH 0.9-7.2, T = 25 °C.
37
pH-Dependent UV-visible absorption data for photoacid 2c
a) b)
c) d)
Figure S5. UV-vis titrations of photoacid 2c. a) Equilibrated absorbance spectra at different pH values in the dark (orange: pH = 3.2,
yellow/purple: pH = 9.8). b) Change in absorbance maxima of 2c-MCH (orange, λ = 446 nm) and 2c-MC (purple, λ = 548 nm) over
the pH range with corresponding fits to Eq. S1 shown as solid lines. Experimental conditions: [2c] = 18.5 μM, [phosphate
buffer] = 20 mM, pH 2.3-9.8, T = 25 °C. c) Equilibrated absorbance spectra at different pH values (blue: pH = 0.9, yellow: pH = 7.2)
under continuous blue light irradiation (λ = 455 nm). d) Change in absorbance maxima of 2c-SP (yellow, λ = 249 nm) and 2c-cis-
455 nm), [2c] = 18.5 μM, [phosphate buffer] = 20 mM, pH 0.9-7.2, T = 25 °C.
38
pH-Dependent UV-visible absorption data for photoacid 2d
a) b)
c) d)
Figure S6. UV-vis titrations of photoacid 2d. a) Equilibrated absorbance spectra at different pH values in the dark (blue: pH = –1.1,
orange: pH = 4.0, purple: pH = 7.4). b) Change in absorbance maximum of 2d-MC (purple, λ = 507 nm) over the pH range with
corresponding fits to Eq. S1 shown as solid lines. Experimental conditions: [2d] = < 25 μM, [phosphate buffer] = 25 mM, pH –1.1-
7.4, T = 25 °C. c) Equilibrated absorbance spectra at different pH values (blue: pH = –1.1, yellow: pH = 4.0) under continuous blue
light irradiation (λ = 455 nm). d) Change in absorbance maxima of 2d-SP (yellow, λ = 271 nm) and 2d-cis-MCH (blue, λ = 310 nm)
over the pH range with corresponding fits to Eq. S1 shown as solid lines. Experimental conditions: hν (λ = 455 nm), [2d] = <25 μM,
[phosphate buffer] = 25 mM, pH –1.1-4.0, T = 25 °C.
39
pH-Dependent UV-visible absorption data for photoacid 3a
a) b)
c) d)
Figure S7. UV-vis titrations of photoacid 3a. a) Equilibrated absorbance spectra at different pH values in the dark (orange: pH = 2.3,
yellow/purple: pH = 9.8). b) Change in absorbance maxima of 3a-MCH (orange, λ = 430 nm) and 3a-MC (purple, λ = 543 nm) over
the pH range with corresponding fits to Eq. S1 shown as solid lines. Experimental conditions: [3a] = 16.7 μM, [phosphate
buffer] = 20 mM, pH 2.3-9.8, T = 25 °C. c) Equilibrated absorbance spectra at different pH values (blue: pH = –1.1, yellow: pH = 6.6)
under continuous blue light irradiation (λ = 455 nm). d) Change in absorbance maxima of 3a-SP (yellow, λ = 243 nm) and 3a-cis-
455 nm), [3a] = 16.7 μM, [phosphate buffer] = 20 mM, pH –1.1-6.6, T = 25 °C.
40
pH-Dependent UV-visible absorption data for photoacid 3b
a) b)
c) d)
Figure S8. UV-vis titrations of photoacid 3b. a) Equilibrated absorbance spectra at different pH values in the dark (orange: pH = 0.9,
yellow/purple = 9.9). b) Change in absorbance maxima of 3b-MCH (orange, λ = 441 nm) and 3b-MC (purple, λ = 560 nm) over the
pH range with corresponding fits to Eq. S1 shown as solid lines. Experimental conditions: [3b] = 16.2 μM, [phosphate
buffer] = 20 mM, pH 0.9-9.9, T = 25 °C. c) Equilibrated absorbance spectra at different pH values (blue: pH = −1.1, yellow: pH = 6.6)
under continuous blue light irradiation (λ = 455 nm). d) Change in absorbance maxima of 3b-SP (yellow, λ = 222 nm) and 3b-cis-
455 nm), [3b] = 16.2 μM, [phosphate buffer] = 20 mM, pH −1.1-6.6, T = 25 °C.
41
pH-Dependent UV-visible absorption data for photoacid 3c
a) b)
c) d)
Figure S9. UV-vis titrations of photoacid 3c. a) Equilibrated absorbance spectra at different pH values in the dark (orange: pH = 1.9,
yellow/purple: pH = 9.9). b) Change in absorbance maxima of 3c-MCH (orange, λ = 451 nm) and 3c-MC (purple, λ = 551 nm) over
the pH range with corresponding fits to Eq. S1 shown as solid lines. Experimental conditions: [3c] = 16.9 μM, [phosphate
under continuous blue light irradiation (λ = 455 nm). d) Change in absorbance maxima of 3c-SP (yellow, λ = 249 nm) and 3c-cis-
455 nm), [3c] = 16.9 μM, [phosphate buffer] = 20 mM, pH –1.1-6.2, T = 25 °C.
42
pH-Dependent UV-visible absorption data for photoacid 3d
a) b)
c) d)
Figure S10. UV-vis titrations of photoacid 3d. a) Equilibrated absorbance spectra at different pH values in the dark (orange: pH =
3.1, yellow/purple: pH = 9.9). b) Change in absorbance maxima of 3d-MCH (orange, λ = 442 nm) and 3d-MC (purple, λ = 538 nm)
over the pH range with corresponding fits to Eq. S1 shown as solid lines. Experimental conditions: [3d] = 16.3 μM, [phosphate
under continuous blue light irradiation (λ = 455 nm). d) Change in absorbance maxima of 3d-SP (yellow, λ = 246 nm) and 3d-cis-
455 nm), [3d] = 16.3 μM, [phosphate buffer] = 20 mM, pH –1.1-6.1, T = 25 °C.
43
S6.6. Kinetic parameters of the SP-MC equilibrium for photoswitches 2a–d, 3a–d
Determination of equilibrium constant Kc and rate constants kopen and kclose

We determined the equilibrium constant Kc for the MC to SP equilibrium in the dark by following the equilibration
process of samples at different pH values by UV-vis absorbance measurements and fitting for the observed rate
constant, kobs. The absorbance at wavelengths characteristic for the individual species (SP, MC, MCH) were plotted
over time and fit to a single exponential function (Eq. S2) to give the observed rate constant kobs at the respective pH
value. The initial absorption is A0, and the final absorption at equilibrium is Aeq.
𝐴𝐴𝑡𝑡 = 𝐴𝐴𝑒𝑒𝑒𝑒 + �𝐴𝐴0 − 𝐴𝐴𝑒𝑒𝑒𝑒 �𝑒𝑒 −𝑘𝑘𝑜𝑜𝑜𝑜𝑜𝑜,𝑒𝑒𝑒𝑒𝑡𝑡 (Eq. S2)
At low pH values (pH << pKadark) the ring opening reaction of SP to MCH is almost exclusively observed. Since MCH
is thermodynamically stable the ring closing reaction can be considered negligible and kobs is effectively the rate
constant for ring opening reaction kopen (Eq. S3).12,11
pH << pKadark: 𝑘𝑘𝑜𝑜𝑜𝑜𝑜𝑜 = 𝑘𝑘𝑜𝑜𝑝𝑝𝑝𝑝𝑝𝑝 (Eq. S3)
At high pH values (pH >> pKadark) the concentration of MCH is negligible as it is predominantly deprotonated to form
MC. Both the ring closing from MC to SP and the ring opening in the opposite direction are observed, which are first
order in each direction. At high pH the observed rate constant is therefore equal to kopen + kclose (Eq. S4), as is the case
for any reversible first order reaction.
pH >> pKadark: 𝑘𝑘𝑜𝑜𝑜𝑜𝑜𝑜 = 𝑘𝑘𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜 + 𝑘𝑘𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 (Eq. S4)
The equilibrium constant Kc is calculated from the ratio of kclose over kopen (Eq. S5).
𝑘𝑘𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐
𝐾𝐾𝑐𝑐 = (Eq. S5)
𝑘𝑘𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜
We observed differences in the obtained kobs based on the experimental setup we were using and concluded this was due
to the extremely high temperature sensitivity of the system. Even slight variations of the determined rate constants will
have a significant impact on the determination of Kc as kopen and kclose do not vary by more than one to two orders in
magnitude. Based on this observation we decided to incorporate an error caused by a ±1.5 °C temperature fluctuation.
Temperature based errors (kopen, kclose, Kc)

Photoacids 2a–c (with Kc < 10)
The equilibration process of SP and MC at high pH was monitored by UV-vis absorption over a range in temperatures
(15–35 °C, 2.5 °C increments). The change in absorption (at a wavelength where SP does not absorb) was fit to a single
exponential function to give the observed rate constant kobs. As previously reported11, the kobs value was used to
determine Kc, kopen and kclose using equations 6-8.
[𝑆𝑆𝑆𝑆] �𝐴𝐴0 − 𝐴𝐴𝑒𝑒𝑒𝑒 �
𝐾𝐾𝑐𝑐 = =
[𝑀𝑀𝑀𝑀] 𝐴𝐴𝑒𝑒𝑒𝑒
(Eq. S6)
(assumes that at the measured wavelength the absorption is only due to MC and at t = 0
the sample is 100% MC)
𝐾𝐾𝑐𝑐
𝑘𝑘𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 = 𝑘𝑘 (Eq. S7)
1 + 𝐾𝐾𝑐𝑐 𝑜𝑜𝑜𝑜𝑜𝑜
1
𝑘𝑘𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜 = 𝑘𝑘 (Eq. S8)
1 + 𝐾𝐾𝑐𝑐 𝑜𝑜𝑜𝑜𝑜𝑜
The obtained rate constants kopen and kclose were analysed in an Eyring plot to give the respective thermodynamic
parameters (ΔH‡open/close, ΔS‡open/close, ΔG‡open/close, ΔG) for the ring opening and ring closing reaction (Eq. S9-14).
Reported errors were obtained by propagating the errors from the linear fit. The final values for kopen and kclose were
obtained by averaging three values (T = 22.5, 25.0, 27.5 °C) extracted from the linear fit in the Eyring plot to account
44
for possible errors in temperature. Reported errors correspond to the standard deviation of these three values. The final
equilibrium constant Kc was determined with Eq. S5.
𝑘𝑘𝑜𝑜𝑜𝑜𝑜𝑜 𝛥𝛥𝛥𝛥 ‡ 1 𝛥𝛥𝛥𝛥 ‡ 𝑘𝑘𝑏𝑏

ln � � = �− �� + � � + ln � � (Eq. S9)
𝑇𝑇 𝑅𝑅 𝑇𝑇 𝑅𝑅 ℎ
𝛥𝛥𝛥𝛥 ‡ = 𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠 ∙ 𝑅𝑅 (Eq. S10)
𝑘𝑘𝑏𝑏
𝛥𝛥𝛥𝛥 ‡ = 𝑅𝑅 �𝑦𝑦𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 − ln � �� (Eq. S11)
ℎ
Using the standard constants:
𝐽𝐽
𝑅𝑅 = 8.3145
𝑚𝑚𝑚𝑚𝑚𝑚 ∙ 𝐾𝐾
ℎ = 6.62607015 ∙ 10−34 𝐽𝐽𝐽𝐽
𝐽𝐽
𝑘𝑘𝑏𝑏 = 1.380649 ∙ 10−23
𝐾𝐾
𝛥𝛥𝛥𝛥 ‡ = 𝛥𝛥𝛥𝛥 ‡ − 𝑇𝑇𝛥𝛥𝛥𝛥 ‡ (Eq. S12)
Photoacid
𝛥𝛥𝛥𝛥 = 𝛥𝛥𝛥𝛥 ‡ 𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜 − 𝛥𝛥𝛥𝛥 ‡ 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 (Eq. S13)
2a–c
Photoacid
𝛥𝛥𝛥𝛥 = −𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝐾𝐾𝑐𝑐 (Eq. S14)
3a–d
For photoacids 3a–d (with Kc > 10)
For these photoacids the ring closing reaction is sufficiently fast to prevent accurate measurement of kobs at high pH values
using standard UV-vis spectroscopy measurements. Instead, samples at high pH were irradiated with UV-light (λ = 265
nm) to drive the equilibrium from SP towards MC. The equilibration process was then monitored in the dark at λ = 560 nm
over time (seconds to minutes) to determine kobs. Due to the small change in absorbance many measurements of the
pump/probe cycle were averaged to give the final data (10–300 repetitions).
Since the rate constant for ring closure kclose for photoacids 3a–d is two orders of magnitude higher than kopen we
approximate kobs to be equal to kclose at high pH. Using this assumption, we determined the temperature dependence of kclose
for photoacids 3a–d. The final value of kclose with a temperature dependent error was obtained by averaging three values (T
= 22.5, 25.0, 27.5 °C) extracted from the linear fit in the Eyring plot to account for possible errors in temperature. Reported
errors correspond to the standard deviation of these three values. The final equilibrium constant Kc was determined with Eq.
S5. We approximate the error of kclose to be dominant for the calculation of Kc and the temperature dependence of kopen was
neglected. The energy difference between SP and MC (∆G) was calculated directly from Kc using Eq. S14.
Figure S11. Energy diagram of the ring opening/closing reaction of MC and SP.
45
Experimental data and kinetic analysis of the SP-MC equilibrium for photoacids 2a–d, 3a–d
pH and temperature dependent kobs for photoacid 2a

a) b)
c) d)
e) f)
Figure S12. Kinetic parameters of photoacid 2a. Unless otherwise specified an aqueous sample of photoacid 2a was added to
respective buffer solution in situ to monitor equilibration processes. a) Kinetic profiles for the equilibration of MCH (λ = 423 nm) at
different pH (dark blue = low pH, light blue = high pH) with corresponding exponential fits (Eq. S2) depicted as solid lines.
Experimental conditions: [2a] = 20.0 μM, [phosphate buffer] = 20 mM, pH 3.20-9.80, T = 25 °C. b) Kinetic profiles for the
equilibration of MC (λ = 531 nm) at different pH (dark blue = low pH, light blue = high pH) with corresponding exponential fits (Eq.
S2) depicted as solid lines. Experimental conditions: [2a] = 20.0 μM, [phosphate buffer] = 20 mM, pH 6.15-9.80, T = 25 °C. c)
Observed rate constants kobs obtained from the exponential fit in a) over a pH range (orange) and from the exponential fit in b) over a
pH range (purple). Corresponding errors from exponential fits are < 2% and therefore not shown. Data point at ~5.7 can be neglected
as no change in absorbance is observed at this pH resulting in an inaccurate exponential fit. d) Kinetic profiles for the equilibration of
MC (λ = 531 nm) at different temperatures (15-35 °C, 2.5 °C increments) over time (grey) with corresponding fits to Eq. S2 (red).
Experimental conditions: [2a] = 20.3 μM, [phosphate buffer] = 20 mM, pH ~ 9, T = 15-35 °C. e) Observed rate constants kobs
obtained from the exponential fit in d) over a temperature range (15-35 °C) with corresponding errors from exponential fit. f) Eyring
plot for ring opening and ring closing reaction: kopen and kclose were calculated from Eq. S7, 8 using obtained temperature dependent
kobs shown in e).
46
pH and temperature dependent kobs for photoacid 2b
a) b)
c) d)
e) f)
Figure S13. Kinetic parameters of photoacid 2b. Unless otherwise specified an aqueous sample of the photoacid 2b was added to
Experimental conditions: [2b] = 20.5 μM, [phosphate buffer] = 20 mM, pH 3.16-9.81, T = 25 °C. b) Kinetic profiles for the
S2) depicted as solid lines. Experimental conditions: [2b] = 20.5 μM, [phosphate buffer] = 20 mM, pH 6.12-9.81, T = 25 °C. c)
Observed rate constants kobs obtained from the exponential fit in a) over a pH range (orange) and from the exponential fit in b) over a
pH range (purple) with corresponding error bars from exponential fits. Data point at ~5.1 can be neglected as no change in
absorbance is observed at this pH resulting in an inaccurate exponential fit. d) Kinetic profiles for the equilibration of MC (λ = 546
nm) at different temperatures (15-35 °C, 2.5 °C increments) over time (grey) with corresponding fits to Eq. S2 (red). Experimental
conditions: [2b] = 20.6 μM, [phosphate buffer] = 20 mM, pH ~ 9, T = 15-35 °C. e) Observed rate constants kobs obtained from the
exponential fit in d) over a temperature range (15-35 °C) with corresponding errors from exponential fit. f) Eyring plot for ring
opening and ring closing reaction: kopen and kclose were calculated from Eq. S7, 8 using obtained temperature dependent kobs shown in
e).
47
pH and temperature dependent kobs for photoacid 2c
a) b)
c) d)
e) f)
Figure S14. Kinetic parameters of photoacid 2c. Unless otherwise specified an aqueous sample of the photoacid 2c was added to
Experimental conditions: [2c] = 18.5 μM, [phosphate buffer] = 20 mM, pH 2.29-9.80, T = 25 °C. b) Kinetic profiles for the
S2) depicted as solid lines. Experimental conditions: [2c] = 18.5 μM, [phosphate buffer] = 20 mM, pH 8.13-9.80, T = 25 °C. c)
Observed rate constants kobs obtained from the exponential fit in a) (orange) and rate constants obtained from the exponential fit in b)
(purple). d) Kinetic profiles for the equilibration of MC (λ = 548 nm) at different temperatures (15-35 °C, 2.5 °C increments) over
time (grey) with corresponding fits to Eq. S2 (red). Experimental conditions: [2c] = 19.4 μM, [phosphate buffer] = 20 mM, pH ~ 9, T
= 15-35 °C. e) Observed rate constants kobs obtained from the exponential fit in d) over a temperature range (15-35 °C) with
corresponding errors from exponential fit. f) Eyring plot for ring opening and ring closing reaction: kopen and kclose were calculated
from Eq. S7, 8 using obtained temperature dependent kobs shown in e).
48
pH dependent kinetics for photoacid 2d
The thermal equilibrium establishment between SP and MC for compound 2d is significantly slower compared to all
other compounds studied here. In addition, hydrolytic degradation of the MC isomer occurs on a similar timescale to the
ring opening and ring closing, implying that an expanded kinetic model had to be used for analysis of the data. The
kinetic model is displayed in Scheme 1 below, and the experimental results are shown in Fig. S15.
Scheme S1. Kinetic model of the thermal equilibration at high pH for compound 2d.
At pH values well below the pKa of 2d, but still above pKahν (where virtually all MC is protonated to form MCH but
negligible cis-MCH is formed) the kinetic model simplifies to SP → MCH with the observed rate constant kobs = ko (See
Eq. S3). The rate constant of ring opening, ko, was determined by monitoring the absorption of MCH after irradiation
(hν = 455 nm) of the solution at low pH and fitting the data to Eq. S2 (Figure S15, b).
At pH values well above the pKa (where virtually no MCH is formed) the corresponding kinetic trace (monitoring the
absorbance of MC) is expected to display a “rise-and-decay” behavior, where the initial rise reflects the thermal
equilibrium between SP and MC, and the subsequent decay reflects the hydrolysis of MC.12 Indeed, this behavior is
clearly observed in Fig. S15 d) where the absorption band of MC was monitored after irradiation (hν = 455 nm). The
change in the absorbance of MC was fit to a biexponential expression (Eq. S15), allowing the extraction of the two rate
constants krise and kdecay.
𝐴𝐴 = 𝐴𝐴0 + 𝑎𝑎𝑒𝑒 −𝑘𝑘𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟 (𝑡𝑡−𝑡𝑡0) + 𝑏𝑏𝑒𝑒 −𝑘𝑘𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑 (𝑡𝑡−𝑡𝑡0) (Eq. S15)
Given that krise is significantly higher than kdecay, the kinetic analysis simplifies to equations S16 and S17, where khydr is
the hydrolysis rate constant.13
𝑘𝑘𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟 = 𝑘𝑘𝑜𝑜 + 𝑘𝑘𝑐𝑐 (Eq. S16)
𝑘𝑘𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟
𝑘𝑘ℎ𝑦𝑦𝑦𝑦𝑦𝑦 = 𝑘𝑘𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑 (Eq. S17)
𝑘𝑘𝑜𝑜
a) b)
c) d)
Figure S15. Kinetic parameters of photoacid 2d. Unless otherwise specified an aqueous sample of the photoacid 2d was added to
respective buffer solution in situ to monitor equilibration processes. a) Absorbance spectra recorded of the MCH recovery (orange) in
the dark after complete bleaching (yellow) at low pH. Experimental conditions: [2d] = < 25 μM, [phosphate buffer] = 25 mM, pH 2.5,
T = 25 °C. b) Kinetic profile for the equilibration of MCH (λ = 407 nm) of a) with corresponding exponential fit (Eq. S2) depicted as
red line. c) Absorbance spectra recorded of the MC recovery (purple) in the dark after bleaching (yellow) at high pH. Experimental
conditions: [2d] = < 25 μM, [phosphate buffer] = 25 mM, pH 7.7, T = 25 °C. d) Kinetic profile for the formation of MC (λ = 506 nm)
after irradiation (shown in c)) and following hydrolytic degradation. Biexponential fit (Eq. S15) is depicted as red line.
49
pH and temperature dependent kobs for photoacid 3a
a)
c) b)
d) e)
Figure S16. Kinetic parameters of photoacid 3a. a) Kinetic profiles for the equilibration of MCH (λ = 430 nm) at different pH (dark
blue = low pH, light blue = high pH) with corresponding exponential fits (Eq. S2) depicted as solid lines. An aqueous sample of
photoacid 3a was added to respective buffer solution in situ to monitor equilibration processes. Experimental conditions: [3a] =
16.7 μM, [phosphate buffer] = 20 mM, pH 2.3-6.1, T = 25 °C. b) Kinetic profile of MC (λ = 560 nm) under irradiation (λ = 265 nm,
blue box) and in the dark. The exponential fit (Eq. S2) is depicted as a red line. Experimental conditions: [3a] = 16.7 μM, [phosphate
buffer] = 20 mM, pH ~ 9.5, T = 15-35 °C. c) Observed rate constants kobs obtained from the exponential fit in a) over a pH range
(orange, error bars from exponential fits) and additional rate constant at high pH obtained by UV-irradiation experiments in b)
(purple, error bar from standard deviation by averaging three kclose values T = 25 ± 1.5 °C) d) Observed rate constants kobs (at high pH
kobs ~ kclose) obtained from the exponential fit in b) over a temperature range (15-35 °C) with corresponding errors from exponential
fit. e) Eyring plot for ring closing reaction.
50
pH and temperature dependent kobs for photoacid 3b
a)
c) b)
d) e)
Figure S17. Kinetic parameters of photoacid 3b. a) Kinetic profiles for the equilibration of MCH (λ = 441 nm) at different pH (dark
blue = low pH, light blue = high pH) with corresponding exponential fits (Eq. S2) depicted as solid lines. An aqueous sample of
photoacid 3b was added to respective buffer solution in situ to monitor equilibration processes. Experimental conditions:
[3b] = 16.2 μM, [phosphate buffer] = 20 mM, pH 0.89-7.16, T = 25 °C. b) Kinetic profile of MC (λ = 560 nm) under irradiation (λ =
265 nm, blue box) and in the dark. The exponential fit (Eq. S2) is depicted as a red line. Experimental conditions: [3b] = 16.2 μM,
[phosphate buffer] = 20 mM, pH ~ 9.5, T = 15-35 °C. c) Observed rate constants kobs obtained from the exponential fit in a) over a pH
range (orange, error bars from exponential fits) and additional rate constant at high pH obtained by UV-irradiation experiments in b)
(purple, error bar from standard deviation by averaging three kclose values T = 25 ± 1.5 °C) d) Observed rate constants kobs (at high pH
kobs ~ kclose) obtained from the exponential fit in b) over a temperature range (15-35 °C) with corresponding errors from exponential
fit. e) Eyring plot for ring closing reaction.
51
pH and temperature dependent kobs for photoacid 3c
a) b)
d) c)
e) f)
Figure S18. Kinetic parameters of photoacid 3c. Unless otherwise specified an aqueous sample of photoacid 3c was added to
Experimental conditions: [3c] = 16.9 μM, [phosphate buffer] = 20 mM, pH 0.89-9.85, T = 25 °C. b) Kinetic profiles for the
equilibration of MCH (λ = 551 nm) at different pH (dark blue = low pH, light blue = high pH) with corresponding exponential fits
(Eq. S2) depicted as solid lines. Experimental conditions: [3c] = 16.9 μM, [phosphate buffer] = 20 mM, pH 5.26-9.85, T = 25 °C. c)
Kinetic profile of MC (λ = 560 nm) under irradiation (λ = 265 nm, blue box) and in the dark. The exponential fit (Eq. S2) is depicted
as a red line. Experimental conditions: [3c] = 18.3 μM, [phosphate buffer] = 20 mM, pH ~ 9.5, T = 15-35 °C. d) Observed rate
constants kobs obtained from the exponential fit in a) (orange, error bars from exponential fits) and b) (purple, error bars from
exponential fits) over a pH range and additional rate constant at high pH obtained by UV-irradiation experiments in c) (grey, error
bar from standard deviation by averaging three kclose values T = 25 ± 1.5 °C) e) Observed rate constants kobs (at high pH kobs ~ kclose)
obtained from the exponential fit in c) over a temperature range (15-35 °C) with corresponding errors from exponential fit. f) Eyring
plot for ring closing reaction.
52
pH and temperature dependent kobs for photoacid 3d
a) b)
d) c)
e) f)
Figure S19. Kinetic parameters of photoacid 3d. Unless otherwise specified an aqueous sample of photoacid 3d was added to
Experimental conditions: [3d] = 16.3 μM, [phosphate buffer] = 20 mM, pH 2.7-9.9, T = 25 °C. b) Kinetic profiles for the equilibration
of MC (λ = 538 nm) at different pH (dark blue = low pH, light blue = high pH) with corresponding exponential fits (Eq. S2) depicted
as solid lines. Experimental conditions: [3d] = 16.3 μM, [phosphate buffer] = 20 mM, pH 7.67-9.85, T = 25 °C. c) Kinetic profile of
MC (λ = 560 nm) under irradiation (λ = 265 nm, blue box) and in the dark. The exponential fit (Eq. S2) is depicted as a red line.
Experimental conditions: [3d] = 16.7 μM, [phosphate buffer] = 20 mM, pH ~ 9.5, T = 15-35 °C. d) Observed rate constants kobs
obtained from the exponential fit in a) (orange, error bars from exponential fits) and b) (purple, error bars from exponential fits) over
a pH range and additional rate constant at high pH obtained by UV-irradiation experiments in c) (grey, error bar from standard
deviation by averaging three kclose values T = 25 ± 1.5 °C) e) Observed rate constants kobs (at high pH kobs ~ kclose) obtained from the
exponential fit in c) over a temperature range (15-35 °C) with corresponding errors from exponential fit. f) Eyring plot for ring
closing reaction.
53
S6.7. Calculation of mole fractions of photoacids 2a–c, 3a–d
Calculation of mole fractions in the dark based on Kc and pKa
In the dark mole fractions of MCH, MC and SP can be calculated from the respective pKa and Kc values:
Hendersson Hasselbach equation:
[𝐴𝐴− ]
𝑝𝑝𝑝𝑝 = 𝑝𝑝𝐾𝐾𝑎𝑎 + log � � (Eq. S18)
[𝐻𝐻𝐻𝐻]
Applied to merocyanines:
[𝑀𝑀𝑀𝑀]
= 10𝑝𝑝𝑝𝑝−𝑝𝑝𝐾𝐾𝑎𝑎
[𝑀𝑀𝑀𝑀𝑀𝑀]
[𝑀𝑀𝑀𝑀] = 10𝑝𝑝𝑝𝑝−𝑝𝑝𝐾𝐾𝑎𝑎 [𝑀𝑀𝑀𝑀𝑀𝑀] (Eq. S19)
With equilibrium constant Kc (Eq. S5):

𝑘𝑘𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 [𝑆𝑆𝑆𝑆]
𝐾𝐾𝑐𝑐 = =
𝑘𝑘𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜 [𝑀𝑀𝑀𝑀]
[𝑆𝑆𝑆𝑆] = 𝐾𝐾𝑐𝑐 [𝑀𝑀𝑀𝑀] (Eq. S20)
Total concentration C:
𝐶𝐶 = [𝑆𝑆𝑆𝑆] + [𝑀𝑀𝑀𝑀] + [𝑀𝑀𝑀𝑀𝑀𝑀]
[𝑆𝑆𝑆𝑆] = 𝐶𝐶 − [𝑀𝑀𝑀𝑀] − [𝑀𝑀𝑀𝑀𝑀𝑀] (Eq. S21)
Eq. S20 in Eq. S21 gives:

𝐾𝐾𝑐𝑐 [𝑀𝑀𝑀𝑀] = 𝐶𝐶 − [𝑀𝑀𝑀𝑀] − [𝑀𝑀𝑀𝑀𝑀𝑀]
𝐾𝐾𝑐𝑐 [𝑀𝑀𝑀𝑀] + [𝑀𝑀𝑀𝑀] = 𝐶𝐶 − [𝑀𝑀𝑀𝑀𝑀𝑀]
𝐶𝐶 − [𝑀𝑀𝑀𝑀𝑀𝑀]
[𝑀𝑀𝑀𝑀] = (Eq. S22)
𝐾𝐾𝑐𝑐 + 1
Eq. S22 in Eq. S19:

𝐶𝐶 − [𝑀𝑀𝑀𝑀𝑀𝑀]
= 10𝑝𝑝𝑝𝑝−𝑝𝑝𝐾𝐾𝑎𝑎 [𝑀𝑀𝑀𝑀𝑀𝑀]
𝐾𝐾𝑐𝑐 + 1
𝐶𝐶 = (𝐾𝐾𝑐𝑐 + 1) 10𝑝𝑝𝑝𝑝−𝑝𝑝𝐾𝐾𝑎𝑎 [𝑀𝑀𝑀𝑀𝑀𝑀] + [𝑀𝑀𝐶𝐶𝐶𝐶]
𝐶𝐶
[𝑀𝑀𝑀𝑀𝑀𝑀] = (Eq. S23)
(𝐾𝐾𝑐𝑐 + 1) ∙ 10𝑝𝑝𝑝𝑝−𝑝𝑝𝐾𝐾𝑎𝑎 + 1
Mole fractions:
[𝑀𝑀𝑀𝑀𝑀𝑀] 1
χ𝑀𝑀𝑀𝑀𝑀𝑀 = = (Eq. S24)
𝐶𝐶 (𝐾𝐾𝑐𝑐 + 1) ∙ 10𝑝𝑝𝑝𝑝−𝑝𝑝𝐾𝐾𝑎𝑎 + 1
[𝑀𝑀𝑀𝑀] 1
χ𝑀𝑀𝑀𝑀 = = (Eq. S25)
𝐶𝐶 1
(𝐾𝐾𝑐𝑐 + 1) + 𝑝𝑝𝑝𝑝−𝑝𝑝𝐾𝐾𝑎𝑎
10
[𝑆𝑆𝑆𝑆] 𝐾𝐾𝑐𝑐
χ𝑆𝑆𝑆𝑆 = = (Eq. S26)
𝐶𝐶 1
(𝐾𝐾𝑐𝑐 + 1) + 𝑝𝑝𝑝𝑝−𝑝𝑝𝐾𝐾𝑎𝑎
10
54
Calculation of mole fractions under irradiation based on pKahν
We assume that under irradiation only cis-MCH and SP are in equilibrium and cis-MCH is treated as pseudo “SPH+”.
Under irradiation mole fractions of cis-MCH and SP can be calculated from the respective pKahν values:
With Hendersson Hasselbach equation (Eq. S27):

[𝑆𝑆𝑆𝑆]ℎ𝜈𝜈
= 10𝑝𝑝𝐻𝐻−𝑝𝑝𝐾𝐾𝑎𝑎
[cis-MCH]
ℎ𝜈𝜈
[𝑆𝑆𝑆𝑆]ℎ𝜈𝜈 = [cis-MCH]10𝑝𝑝𝑝𝑝−𝑝𝑝𝐾𝐾𝑎𝑎 (Eq. S27)
The total concentration C is:

𝐶𝐶 = [𝑆𝑆𝑆𝑆]ℎ𝜈𝜈 + [cis-MCH]
[cis-MCH] = 𝐶𝐶 − [𝑆𝑆𝑆𝑆]ℎ𝜈𝜈 (Eq. S28)
Eq. S27 in Eq. S28 gives

𝐶𝐶
[cis-MCH] = ℎ𝜈𝜈 (Eq. S29)
10𝑝𝑝𝑝𝑝−𝑝𝑝𝐾𝐾𝑎𝑎 + 1
[𝑆𝑆𝑆𝑆]ℎ𝜈𝜈 = 𝐶𝐶 − [cis-MCH] (Eq. S30)
Mole fractions of cis-MCH (χcis-MCH ) and SP (χ𝑆𝑆𝑆𝑆ℎ𝜈𝜈 ):

[cis-MCH] 1
χ𝑐𝑐𝑐𝑐𝑐𝑐−𝑀𝑀𝑀𝑀𝑀𝑀 = = (Eq. S31)
𝐶𝐶 ℎ𝜈𝜈
10𝑝𝑝𝑝𝑝−𝑝𝑝𝐾𝐾𝑎𝑎 + 1
χ𝑆𝑆𝑆𝑆ℎ𝜈𝜈 = 1 − χcis-MCH (Eq. S32)
55
Overview over calculated mole fractions of photoacids 2a–c, 3a–d in the dark and under irradiation
2a 2b
2c
Figure S20. Overview over calculated mole fractions of different species in the dark and under irradiation (λ = 455 nm) of
photoacids 2a–c.
3a 3b
3c 3d
Figure S21. Overview over calculated mole fractions of different species in the dark and under irradiation (λ = 455 nm) of
photoacids 3a–d.
56
S6.8. Determination of extinction coefficients of SP and MCH
For each photoswitch the absorption spectrum of “pure” SP and “pure” MCH were collected. Appropriate pH values
were chosen from the titration curves. To collect a spectrum with ~ 100% MCH the highest pH value was chosen that
had a maximum absorption ~ 450 nm corresponding to the absorbance maximum for MCH. To collect a spectrum
consisting of ~ 100% SP a pH value was chosen at which the absorbance could be entirely bleached in the visible
region. Too low pH values needed to be avoided due to formation of cis-MCH under irradiation and too high pH values
did not result in complete bleach in the visible region of the MC band with the LED light in the UV-vis instrument.
Mole fractions of the respective MCH and SP forms were calculated from Eq. S24 and Eq. S32.
Table S3. Experimental conditions for collection of “pure” spectra of MCH and SP and their respective calculated mole fractions
under the experimental conditions.
MCH-2a MCH-2b MCH-2c MCH-3a MCH-3b MCH-3c MCH-3d
pH 3.2 3.2 3.7 3.2 3.7 3.8 3.7
calc. X / % 99.9 99.9 99.9 96.8 93.6 98.4 98.7
SP-2a SP-2b SP-2c SP-3a SP-3b SP-3c SP-3d
pH under hν 5.2 7.2 7.2 4.7 5.2 4.3 4.2
calc. X / % 99.8 99.9 99.9 99.9 99.9 99.5 99.6
S6.9. Calculation of extinction coefficients of MC

𝜆𝜆
Extinction coefficients of MC were calculated from absorbance spectrum (𝐴𝐴𝑒𝑒𝑒𝑒 ) of an equilibrated sample at
concentration C consisting only of SP and MC. This is the case at high pH (~ 9.5) in the dark where complete
deprotonation of MCH can be assumed. The respective mole fractions were determined from previously determined Kc
(Eq. S5) using Eq. S33 and 34. The extinction coefficients of MC can then be calculated by subtracting the absorbance
due to SP from the equilibrated absorbance spectrum at high pH containing only SP and MC and dividing by the
concentration of MC (Eq. S35).
Calculation of mole fractions of SP (𝝌𝝌𝑺𝑺𝑺𝑺 ) and MC (𝝌𝝌𝑴𝑴𝑴𝑴 ) from Kc

𝜒𝜒𝑆𝑆𝑆𝑆 + 𝜒𝜒𝑀𝑀𝑀𝑀 = 1
𝐾𝐾𝑐𝑐
𝜒𝜒𝑆𝑆𝑆𝑆 = (Eq. S33)
1 + 𝐾𝐾𝑐𝑐
1
𝜒𝜒𝑀𝑀𝑀𝑀 = (Eq. S34)
Calculation of extinction coefficients of MC (𝜺𝜺𝝀𝝀𝑴𝑴𝑴𝑴 )

𝜆𝜆 𝜆𝜆
𝐴𝐴𝑒𝑒𝑒𝑒 − 𝜀𝜀𝑆𝑆𝑆𝑆 𝜒𝜒𝑆𝑆𝑆𝑆 𝐶𝐶
𝜆𝜆
𝜀𝜀𝑀𝑀𝑀𝑀 = (Eq. S35)
𝜒𝜒𝑀𝑀𝑀𝑀 𝐶𝐶
Calculation of an absorbance spectrum of MC

𝜆𝜆
An absorbance spectrum of MC (𝐴𝐴𝑀𝑀𝑀𝑀 ) can be calculated from the equilibrated absorbance spectrum at high pH
containing only SP and MC, subtracting the absorbance due to SP at the same concentration, and dividing the remaining
absorbance by the mole fraction of MC (Eq. S36).
𝜆𝜆 𝜆𝜆
𝐴𝐴𝑒𝑒𝑒𝑒 − 𝐴𝐴𝑆𝑆𝑆𝑆 𝜒𝜒𝑆𝑆𝑆𝑆
𝜆𝜆
𝐴𝐴𝑀𝑀𝑀𝑀 = (Eq. S36)
𝜒𝜒𝑀𝑀𝑀𝑀
𝜆𝜆 𝜆𝜆
where 𝐴𝐴𝑆𝑆𝑆𝑆 = 𝜀𝜀𝑆𝑆𝑆𝑆 𝐶𝐶
57
Absorbance spectra of MCH, SP and calculated absorbance spectrum of MC
2a 2b
2c
Figure S22. Absorbance spectra of SP (yellow), MCH (orange), an equilibrated absorbance spectrum at high pH containing SP and
MC (grey) used to calculate spectrum of MC (purple) with Eq. 36. Experimental conditions:
Photoacid ASP under hν (λ = 455 nm) AMCH (dark) ASP+MC (dark) AMC
[2a] = 20.0 μM pH 5.2 pH 3.2 pH 9.4
[phosphate buffer] = 20 mM Calculated
[2b] = 20.5 μM pH 7.2 pH 3.2 pH 9.3
T = 25 °C (Eq. 36)
[2c] = 18.5 μM pH 7.2 pH 3.7 pH 9.4
3a 3b
3c 3d
Figure S23. Absorbance spectra of SP (yellow), MCH (orange), an equilibrated absorbance spectrum at high pH containing SP and
MC (grey) used to calculate spectrum of MC (purple) with Eq. S36. Experimental conditions:
Photoacid ASP under hν (λ = 455 nm) AMCH (dark) ASP+MC (dark) AMC
[phosphate buffer] = 20 mM [3a] = 16.7 μM pH 4.7 pH 3.2 pH 9.3 Calculated
T = 25 °C [3b] = 16.2 μM pH 5.2 pH 3.7 pH 9.4 (Eq. S36)
[3c] = 16.9 μM pH 4.3 pH 3.8 pH 9.4
[3d] = 16.3 μM pH 4.2 pH 3.7 pH 9.4
58
S6.10. Determining the pKa value11
pKa calculated from pKadark and Kc

pKa values of photoacids 2a–c, 3a–d were calculated from the reported11 equation 34 using experimentally determined
parameters 𝑝𝑝𝑝𝑝𝑎𝑎𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑 (from fit to Eq. S1) and 𝐾𝐾𝑐𝑐 (Eq. S5) in Eq. S38.
𝐾𝐾𝑎𝑎𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑 = 𝐾𝐾𝑎𝑎 (1 + 𝐾𝐾𝑐𝑐 ) (Eq. S37)

−𝑝𝑝𝑝𝑝𝑎𝑎𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑
10
𝑝𝑝𝐾𝐾𝑎𝑎 = − log � � = 𝑝𝑝𝑝𝑝𝑎𝑎𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑 + log(1 + 𝐾𝐾𝑐𝑐 ) (Eq. S38)
(1 + 𝐾𝐾𝑐𝑐 )
Fitting an experimentally determined ratio of MCH/MC/SP to determine the pKa

Determining the pH-dependent mole fractions for SP, MCH and MC from absorbance spectra
The mole fractions of SP (χ𝑆𝑆𝑆𝑆 ), MCH (χ𝑀𝑀𝑀𝑀𝑀𝑀 ) and MC (χ𝑀𝑀𝑀𝑀 ) were determined from equilibrated absorbance spectra at
different pH values.
1
𝜆𝜆
χ𝑀𝑀𝑀𝑀 is calculated from the equilibrium absorbance (𝐴𝐴𝑒𝑒𝑒𝑒 ) at a wavelength (λ1) where only MC absorbs ~550 nm (Eq.
S40).
2
𝜆𝜆
χ𝑀𝑀𝑀𝑀𝑀𝑀 is calculated from the equilibrium absorbance (𝐴𝐴𝑒𝑒𝑒𝑒 ) at a wavelength (λ2) where only MCH and MC absorb ~450
nm (Eq. S41).
Since we assume only MCH, MC and SP are present in the dark over this pH range χ𝑆𝑆𝑆𝑆 can be calculated from Eq. S39.
χ𝑆𝑆𝑆𝑆 = 1 − χ𝑀𝑀𝑀𝑀 − χ𝑀𝑀𝑀𝑀𝑀𝑀 (Eq. S39)

𝜆𝜆1
𝐴𝐴𝑒𝑒𝑒𝑒
[𝑀𝑀𝑀𝑀] = 𝜆𝜆1
𝜀𝜀𝑀𝑀𝑀𝑀
𝜆𝜆1
𝐴𝐴𝑒𝑒𝑒𝑒
χ𝑀𝑀𝑀𝑀 = 𝜆𝜆1 (Eq. S40)
𝜀𝜀𝑀𝑀𝑀𝑀 𝐶𝐶
𝜆𝜆 2 𝜆𝜆2 𝜆𝜆2
𝐴𝐴𝑒𝑒𝑒𝑒 = �χ𝑀𝑀𝑀𝑀 𝜀𝜀𝑀𝑀𝑀𝑀 + χ𝑀𝑀𝑀𝑀𝑀𝑀 𝜀𝜀𝑀𝑀𝑀𝑀𝑀𝑀 �𝐶𝐶
𝜆𝜆 2
𝐴𝐴𝑒𝑒𝑒𝑒 𝜆𝜆2 𝜆𝜆2
= χ𝑀𝑀𝑀𝑀 𝜀𝜀𝑀𝑀𝑀𝑀 + χ𝑀𝑀𝑀𝑀𝑀𝑀 𝜀𝜀𝑀𝑀𝑀𝑀𝑀𝑀
𝐶𝐶
𝜆𝜆2
𝐴𝐴𝑒𝑒𝑒𝑒 𝜆𝜆2
−χ𝑀𝑀𝐶𝐶 𝜀𝜀𝑀𝑀𝑀𝑀
χ𝑀𝑀𝑀𝑀𝑀𝑀 = 𝐶𝐶
𝜆𝜆2
𝜀𝜀𝑀𝑀𝑀𝑀𝑀𝑀
𝜆𝜆 2
𝐴𝐴𝑀𝑀𝑀𝑀𝑀𝑀 𝜆𝜆2
1
χ𝑀𝑀𝑀𝑀𝑀𝑀 =� −χ𝑀𝑀𝑀𝑀 𝜀𝜀𝑀𝑀𝑀𝑀 � 𝜆𝜆2
𝐶𝐶 𝜀𝜀𝑀𝑀𝑀𝑀𝑀𝑀 (Eq. S41)
Calculating the ratio of 𝝋𝝋 to determine the pKa value

Using the mole fractions at different pH, we calculated the ratio 𝜑𝜑 (Eq. S42) of the three species and plotted it over a
range in pH. The curve of 𝜑𝜑 over pH was fit to Eq. S42 to give parameters Kc and Ka.
[𝑆𝑆𝑆𝑆] 𝐾𝐾𝑐𝑐
𝜑𝜑 = = (Eq. S42)
[𝑀𝑀𝑀𝑀𝑀𝑀] + [𝑀𝑀𝑀𝑀] [𝐻𝐻 + ]
�1 + �
𝐾𝐾𝑎𝑎
59
𝝋𝝋 over pH to determine Kc and pKa (from fit to Eq. S42) for photoacids 2a–c, 3a–d
2a 2b
2c
Figure S24. φ as a function of pH (grey squares) and corresponding fit to Eq. S42 as solid black line. The ratio of φ was determined
from equilibrated absorbance spectra collected over a range in pH. Experimental conditions:
Photoacid 2a: [2a] = 20.0 μM, [phosphate buffer] = 20 mM, pH 3.2-9.8, T = 25 °C.
Photoacid 2b: [2b] = 20.5 μM, [phosphate buffer] = 20 mM, pH 3.2-9.8, T = 25 °C.
Photoacid 2c: [2c] = 18.5 μM, [phosphate buffer] = 20 mM, pH 2.3-9.8, T = 25 °C.
3a 3b
3c 3d
Figure S25. φ as a function of pH (grey squares) and corresponding fit to Eq. S42 as solid black line. The ratio of φ was determined
from equilibrated absorbance spectra collected over a range in pH. Experimental conditions:
Photoacid 3a: [3a] = 16.7 μM, [phosphate buffer] = 20 mM, pH 2.3-9.8, T = 25 °C.
Photoacid 3b: [3b] = 16.2 μM, [phosphate buffer] = 20 mM, pH 0.9-9.9, T = 25 °C.
Photoacid 3c: [3c] = 16.9 μM, [phosphate buffer] = 20 mM, pH 0.9-9.9, T = 25 °C.
Photoacid 3d: [3d] = 16.3 μM, [phosphate buffer] = 20 mM, pH 2.7-9.9, T = 25 °C.
60
S6.11. Photoswitching of photoacids 2a–c, 3a–d
2a
2b
2c
Figure S26. Photoswitching of 2a–c. Thermal recovery in the dark after initial equilibration in the dark followed by ca. 30 s of
irradiation, scans collected every 12 s (left). Absorbance at respective MCH maximum over time during three photoirradiation cycles
(time of irradiation indicated as blue boxes) followed by thermal recovery in the dark (right). Experimental conditions:
Photoirradiation cycles
Photoacid pH
thν / min tdark / min
[phosphate buffer] = 20 mM [2a] = 20.0 μM 5.2 0.50 28
T = 25 °C [2b] = 20.5 μM 5.2 0.50 15
hν (λ = 455 nm) [2c] = 18.5 μM 6.2 0.33 13
61
3a
3b
3c
3d
Figure S27. Photoswitching of 3a-d. Thermal recovery in the dark after initial equilibration in the dark followed by ca. 30 s of
irradiation, scans collected every 12 s (left). Absorbance at respective MCH maximum over time during three photoirradiation cycles
(time of irradiation indicated as blue boxes) followed by thermal recovery in the dark. Experimental conditions:
Photoirradiation cycles
Photoacid pH
thν / min tdark / min
[3a] = 16.7 μM 3.2 0.33 45
[phosphate buffer] = 20 mM [3b] = 16.2 μM 3.1 0.50 24
T = 25 °C
[3c] = 23.7 μM 4.3 0.33 25
hν (λ = 455 nm)
[3d] = 16.3 μM 3.7 0.33 50
62
Overview of photoswitching parameters of photoacids 2a–c, 3a–d
An overview over relevant photoswitching parameters obtained from experiments listed in Figures S23, S24 are shown
below in Table S4. Observed rate constants were obtained from a single exponential fit for the thermal recovery in the
dark after irradiation (λ = 455 nm).
Table S4. Overview over photoswitching parameters. The highest pH with predominantly MCH present was chosen for
photoswitching experiments.
Proportion of Proportion of
kobs Recovery per
pH t½ /sb MCH in the MCH under hν
/10−2 s−1 a cycle /% d
dark /% c /% c
2a ∼ 5.2 0.36 194 90.7 0.5 99.2
2b ∼ 5.2 1.37 50.8 88.9 0.1 99.3
2c ∼ 6.2 1.51 45.9 79.2 0.4 99.7
3a ∼ 3.2 0.18 390 97.7 − 99.9
3b ∼ 3.1 0.56 125 >99f 0.1 99.8
3c ∼ 4.3 0.65 106 98.4 0.1 99.5
3d ∼ 3.7 0.17 404 97.3 0.6 99.8
a) observed rate constants obtained from single exponential fit to the characteristic absorption band of MCH in the dark over time
after irradiation at λ = 455 nm.
ln 2
b) approximate half life for first order process: 𝑡𝑡½ = at 298K.
𝑘𝑘𝑜𝑜𝑏𝑏𝑏𝑏
𝐴𝐴𝑀𝑀𝑀𝑀𝑀𝑀 ∙𝜀𝜀𝑀𝑀𝑀𝑀𝑀𝑀
c) % 𝑀𝑀𝑀𝑀𝑀𝑀 = , determined at a wavelength at which absorption only is due to MCH
𝑐𝑐𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡
1
𝐴𝐴𝑀𝑀𝑀𝑀𝑀𝑀,𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓 𝑛𝑛
d) % 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟 = � � , n: number of switching cycles.
𝐴𝐴𝑀𝑀𝑀𝑀𝑀𝑀,𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖
63
S7. Transient Absorption data
S7.1. Transient absorption spectra of the MCH forms of photoacid 2a–c, 3a–d
Transient absorption spectra were recorded at pH values at which predominantly the MCH form is present in the dark
and at which no cis-MCH formation was observed under irradiation (λ = 455 nm).
Transient absorption spectroscopy was performed using a home-built set-up. The probe pulse was generated by
focusing the 785 nm pulse (1000 Hz, Clark-MXR CPA 2210, 150 fs, Ti:sapph. regen. amp.) onto a translated CaF2
crystal. The supercontinuum was dispersed by a home-built prism spectrometer. The pump pulse was generated with
the 355 nm output of a 800 ps pulse-width Nd:YAG laser (InnoLas picolo), with the timing controlled electronically.
The supercontinuum spectrum (probe) and pump pulse energy were measured on a shot-to-shot basis. The probe
spectrum was converted to differential transmission by comparison of the shots collected respectively with and without
the pump pulse. Data points more than 2.5 standard deviations from the mean were removed. The remaining shots were
averaged. The final time-wavelength dataset was generated from multiple data points at each pump-probe delay.
2a
2b
2c
Figure S28. Evolution of the transient absorption spectrum of photoacids 2a–c following excitation at 355 nm (left) and respective
kinetic traces at selected wavelengths (right). Experimental conditions:
Photoacid 2a: [2a] = 20.3 μM, [phosphate buffer] = 20 mM, pH ∼ 5, T = 25 °C.
Photoacid 2b: [2b] = 20.6 μM, [phosphate buffer] = 20 mM, pH ∼ 5, T = 25 °C.
Photoacid 2c: [2c] = 19.4 μM, [phosphate buffer] = 20 mM, pH ∼ 6, T = 25 °C.
64
3a
3b
3c
3d
Figure S29. Evolution of the transient absorption spectrum of photoacids 3a–d following excitation at 355 nm (left) and respective
Photoacid 3d: [3d] = 16.7 μM, [phosphate buffer] = 20 mM, pH ∼ 4, T = 25 °C.
65
S7.2. Kinetic analysis of the TA data of the MCH form of photoacids 2a–c, 3a–d
The transient absorption data was fit to a function of the form
𝑡𝑡
Δ𝑂𝑂𝑂𝑂(𝜆𝜆, 𝑡𝑡) = 𝐼𝐼𝐼𝐼𝐼𝐼(𝑡𝑡) × �𝐷𝐷𝐷𝐷𝐷𝐷 1(𝜆𝜆) × exp �− � + 𝐷𝐷𝐷𝐷𝐷𝐷 2(𝜆𝜆)� (Eq. S43)
𝜏𝜏
where the 𝐷𝐷𝐷𝐷𝐷𝐷 𝑛𝑛(𝜆𝜆) are the decay-associated spectra (DAS), and the IRF is the instrument response function. On
application of a sequential model, where the decay of a photo-excited species yields the persistent species, we expect
the following time-dependences of species 1 and 2,
𝑡𝑡
[1]𝑡𝑡 = [1]0 exp �− � (Eq. S44)
𝜏𝜏
𝑡𝑡
[2]𝑡𝑡 = [1]0 �1 − exp �− �� (Eq. S45)
𝜏𝜏
The decaying spectrum, DAS 1, is thus a contribution from species 1 and 2,

𝐷𝐷𝐴𝐴𝑆𝑆 1(𝜆𝜆) = 𝑆𝑆𝑆𝑆𝑆𝑆 1(𝜆𝜆) − 𝑆𝑆𝑆𝑆𝑆𝑆 2(𝜆𝜆) (Eq. S46)
but the persistent spectrum is due to species 2 only:
𝑆𝑆𝑆𝑆𝑆𝑆 2(𝜆𝜆) = 𝐷𝐷𝐷𝐷𝐷𝐷 2(𝜆𝜆) (Eq. S47)
The SAS is the species-associated spectrum, where

𝑆𝑆𝑆𝑆𝑆𝑆 𝑛𝑛(𝜆𝜆) = [𝑛𝑛]𝑡𝑡 Δ𝜖𝜖𝑛𝑛 (𝜆𝜆)𝑙𝑙 (Eq. S48)
and Δ𝜖𝜖𝑛𝑛 (𝜆𝜆) is the change in absorption due to the generation of species n, and accounts for the loss of the initial species.
It follows that,
𝑆𝑆𝑆𝑆𝑆𝑆 1(𝜆𝜆) = 𝐷𝐷𝐷𝐷𝐷𝐷 1(𝜆𝜆) + 𝐷𝐷𝐷𝐷𝐷𝐷 2(𝜆𝜆) (Eq. S49)
The photo-excited species has previously been identified for photoacid 2a as the TCT cis-MC form with a predicted
lifetime > 3.5 ns.14 SAS 2 is the difference between the absorption spectrum of the persistent species and the original
chromophore, and thus exhibits a persistent ground state bleach.
The magnitude of this bleach signal is related to the concentration of switched molecules by
𝑆𝑆𝑆𝑆𝑆𝑆 2 ≈ −Δ[SP](𝑥𝑥𝜖𝜖MC + (1 − 𝑥𝑥)𝜖𝜖MCH )𝑙𝑙 (Eq. S50)
Where l is the pathlength and x is the fraction of MC. The quantum yield of photo-switching, Φ, was then calculated by
dividing Δ[𝑆𝑆𝑆𝑆] by the concentration of absorbed photons, as determined from measurements of the absorbed laser
energy, E, and the spot-size (area) a.
Δ[𝑆𝑆𝑆𝑆]ℎ𝜈𝜈𝑁𝑁𝐴𝐴 𝑎𝑎𝑎𝑎
Φ= (Eq. S51)
𝐸𝐸
Lifetimes of SAS1a
2a 2b 2c 3a 3b 3c 3d
Lifetime (τ)
140 140 280 30 30 80 550
of SAS 1 / ns
a
see τ in Eq. S44.
66
2a
2b
2c
Figure S30. a) Evolution of the transient absorption spectrum of photoacids 2a–c following excitation at 355 nm. b) Residual of the
data shown in a). c) Species associated spectra (SAS). d) Concentration of the SAS in c) with the yield of SAS 2 from SAS 1
indicated in %. e) Decay associated spectra (DAS). f) Kinetic trace of the DAS in e).
Experimental conditions:
67
3a
3b
3c
68
3d
Figure S31. a) Evolution of the transient absorption spectrum of photoacids 3a–d following excitation at 355 nm. b) Residual of the
data shown in a). c) Species associated spectra (SAS). d) Concentration of the SAS in c) with the yield of SAS 2 from SAS 1
indicated in %. e) Decay associated spectra (DAS). f) Kinetic trace of the DAS in e).
Experimental conditions:
69
S7.3. Transient absorption spectra of the MC forms of photoacid 2a–c, 3a–d
Transient absorption spectra were recorded at pH values at which predominantly the MC form is present in the dark.
The measurements were undertaken as in Section 6.1, with the following differences. The pump pulse at 530 nm was
generated by a 2-stage optical parametric amplifier (TOPAS-C, light conversion). The pump pulses were delayed with
respect to the probe using a motorized delay stage (Thorlabs, LTS300). An optical chopper modulated the pump laser,
blocking every second pulse. The surfaces plotted below have not been chirp-corrected.
2a
2b
2c
Figure S32. Evolution of the transient absorption spectrum of photoacids 2a–c following excitation at 530 nm (left) and respective
70
3a
3b
3c
3d
Figure S33. Evolution of the transient absorption spectrum of photoacids 3a–d following excitation at 530 nm (left) and respective
71
S8. Summary of key parameters for photoswitches 2a−c, 3a−d
Table S5. Overview of key parameters for photoacids 2a−c, 3a−d.
εSP / 103 × L εMCH / 103 × L εMC / 103 × kopen kcloseg ΦMCH→SP ΦMC→SP
# λ /nm pKadark pKahν c Πf Kcj pKam
mol−1cm−1 mol−1cm−1 L mol−1cm−1 /10−2 s−1 /10−2 s−1 (lit.) (lit.)
251 13.9 5.3 13 ± 2 0.11
6.27 ± 0.02c 5.2 ± 1 7.07 ± 0.10 < ~1%
2a 423 − 29 3.0 ± 0.7 2.42 ± 0.04 3.85 ± 0.08 0.54 ± 0.1g 2.8 ± 0.5 (0.37 in
(6.28 ± 0.10)d (5.6 ± 0.02)k (7.10 ± 0.01)n (< ~1%)r
531 − − 20 ± 4.6 DMSO)q
297 5.2 2.2 1.9 ± 0.8
6.20 ± n.a.c 8.6 ± 2 7.18 ± 0.11
2b 436 − 26 1.7 ± 0.4 2.81 ± 0.04 3.39 ± 0.04 1.7 ± 0.3g 15 ± 3 0.04 < ~1%
(6.26 ± 0.11)d (8.2 ± 0.04)k (7.24 ± 0.01)n
546 − − 18. ± 4.4
249 16.8 4.8 12 ± 0.7

7.11 ± 0.03c 1.2 ± 0.3 7.46 ± 0.07
2c 446 − 30 3.3 ± 0.5 3.72 ± 0.01 3.38 ± 0.03 1.9 ± 0.3g 2.3 ± 0.5 0.16 < ~1%
(7.07 ± 0.07)d (1.25 ± 0.01)k (7.42 ± 0.02)n
548 − − 21 ± 3
4.85 ± 0.04o
2ds − − − 4.66 ± 0.04e 1.88 ± 0.03 2.78 ± 0.05 3.8·10−3 h 2.1·10−3 i 0.55 ± n.a.l − −
(4.72)p
243 16.9 5.1 −a

4.66 ± 0.04c 123 ± 23 6.76 ± 0.09
3a 430 − 30 2.4 ± 0.4 1.03 ± 0.08 3.63 ± 0.09 0.18h 22 ± 4 0.06 < ~1%
(4.68 ± 0.09)d (99 ± 1)k (6.77 ± 0.03)n
543 − − 43 ± 8
296 5.8 2.7 −a
4.68 ± 0.08c 183 ± 40 6.95 ± 0.12
3b 441 − 25 9.7 ± 2b 1.20 ± 0.02 3.48 ± 0.08 0.57h 104 ± 22 0.03 < ~1%
(4.89 ± 0.15)d (202 ± 9)k (7.16 ± 0.12)n
560 − − 47 ± 10b
249 25.1 4.8 −a
5.56 ± 0.01c 41 ± 10 7.18 ± 0.10
3c 451 − 30 1 ± 0.2 1.98 ± 0.01 3.58 ± 0.01 0.62h 25 ± 6 0.11 < ~1%
(5.58 ± 0.10)d (32 ± 0.2)k (7.20 ± 0.01)n
551 − − 54 ± 12
246 16.1 4.9 6.7 ± 1.8
5.55 ± n.a.c 26 ± 5 6.97 ± 0.08
3d 442 − 29 6.5 ± 1.2 1.78 ± 0.03 3.76 ± 0.03 0.17h 4.4 ± 0.9 0.03 < ~1%
(5.56 ± 0.08)d (21 ± 0.1)k (6.99 ± 0.01)n
538 − − 46 ± 8.9
a) Large spectral overlap with MCH and SP leads to significant errors for calculation of εMC.
b) Due to minimal absorption in the visible range of the reference spectrum at high pH used to calculate εMC the extinction coefficient of SP was set to 0 for values < 100.
c) From fit to Eq. S1, average of two fits, error from standard deviation (n.a. = error negligible).
d) 𝑝𝑝𝐾𝐾𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑
𝑎𝑎 = −𝑙𝑙𝑙𝑙𝑙𝑙(𝐾𝐾𝑎𝑎 (1 + 𝐾𝐾𝑐𝑐 )) with Ka determined from fit to Eq. S42, Kc from Eq. S5, uncertainty determined by error propagation.
e) Calculated from 𝑝𝑝𝐾𝐾𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑 𝑎𝑎 = −𝑙𝑙𝑙𝑙𝑙𝑙(𝐾𝐾𝑎𝑎 (1 + 𝐾𝐾𝑐𝑐 )) with Ka determined from fit to Eq S1 and Kc from Eq. S5, uncertainty determined by error propagation.
f) Π = 𝑝𝑝𝐾𝐾𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑
𝑎𝑎 𝑎𝑎 , uncertainty from error propagation.
− 𝑝𝑝𝐾𝐾ℎ𝜈𝜈
g) k averaged over three temperatures (T = 25 ± 1.5 °C) extracted from linear fit in Eyring plot, errors calculated by standard deviation.
h) Error from single exponential fit are negligible, temperature dependence was not determined.
72
i) From fit to Eq. S15 and calculated with Eq. S16, errors from fit are negligible.
j) Calculated from Eq. S5, uncertainty from error propagation; values in parenthesis obtained from fit to Eq. S42 with corresponding fitting errors.
k) Determined from fit to Eq. S42.
l) Calculated from Eq. S5, errors from previous fit were negligible.
m) Calculated from Eq. S38 with 𝑝𝑝𝑝𝑝𝑎𝑎𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑 experimentally determined from fit to Eq. S1 and Kc from Eq. S5.
n) Ka determined from fit to Eq. S42.
o) From fit to Eq. S1, error from fit.
p) As reported in Ref. 7.
q) As reported in Ref. 15.
r) As reported in Ref. 14.
−1
s) Rate of hydrolysis khydr is 1.6·10−5 s , determined from fit to Eq. S15 and calculated from Eq. S17.
73
Table S6. Overview of thermodynamic parameters of photoacids 2a−c, 3a−d determined by Eyring plots.
ΔH‡closed ΔS‡closee ΔG‡close a,f ΔH‡opend ΔS‡opene ΔG‡open a,f

# ΔG a,b
/ kJ mol-1 / J K-1 mol-1 / kJ mol-1 / kJ mol-1 / J K-1 mol-1 / kJ mol-1
4.1 ± 0.7
2a 91.5 ± 1.5 32.2 ± 4.9 81.9 ± 2.1 93.6 ± 1.9 25.5 ± 6.3 86.0 ±2.7
(4.1 ± 3.4)c
5.3 ± 0.7
2b 88.1 ± 1.0 34.5 ± 3.4 77.8 ± 1.4 96.8 ± 2.2 45.7 ± 7.4 83.1 ± 3.1
(5.3 ± 3.4)c
0.53 ± 0.7
2c 98.3 ± 1.3 53.4 ± 4.5 82.4 ± 1.9 90.2 ± 2.5 24.2 ± 8.3 82.9 ± 3.5
(0.52 ± 4.0)c
3a 12 ± 0.5 91.0 ± 2.8 47.4 ± 9.4 76.8 ± 4.0 – – 88.8 ± 4.0
3b 13 ± 0.5 105 ± 8.3 108 ± 28 73.0 ± 12 – – 85.9 ± 12
3c 9.2 ± 0.6 110 ± 4.8 113 ± 16 76.5 ± 5 – – 85.7 ± 4.8
3d 8.0 ± 0.5 98.5 ± 4.8 59.4 ± 16 80.8 ± 6.7 – – 88.8 ± 6.8

a) Determined for 298.15 K.
b) Calculated from Eq. S14 (ΔG = –RTlnKc), uncertainty determined by error propagation.
c) Calculated from Eq. S13 (𝛥𝛥𝛥𝛥 = 𝛥𝛥𝛥𝛥 ‡ 𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜 − 𝛥𝛥𝐺𝐺‡ 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 ), uncertainty determined by error propagation.
d) Linear fit in Eyring plot used for calculation with Eq. S10, uncertainty from propagating the error of the linear fit.
e) Linear fit in Eyring plot used for calculation with Eq. S11, uncertainty from propagating the error of the linear fit.
f) Calculated from Eq. S12.
Figure S34. Energy diagram of the ring opening/closing reaction of MC and SP.
74
S9. Stability of photoacids 2a−c, 3a−d in methanol
a) b)
c)
Figure S35. Equilibrated spectra of a) 2a-MCH, b) 2b-MCH and c) 2c-MCH in the dark at pH ~ 3 obtained shortly after preparing
stock solutions and after 3-4 months of storing them at 4 °C wrapped in aluminum. Decomposition is <6%.
a) b)
c) d)
Figure S36. Equilibrated spectra of a) 3a-MCH, b) 3b-MCH, c) 3c-MCH and d) 3d-MCH in the dark at pH ~ 3 obtained shortly
after preparing stock solutions and after 3-4 months of storing them at 4 °C wrapped in aluminum. Decomposition is <4%.
75
S10. Irradiation of saturated solutions
S10.1. Procedure for irradiating concentrated solutions
Saturated solutions of samples were prepared by adding the compound to 3 mL of a 20 mM KCl solution, sonicated in
an ultrasound bath for 10 min and filtered through syringe filters. The vial was placed in a Penn PhD photoreactior m2.
A HI2221 Calibration Check pH/ORP Meter by HANNA instruments in conjunction with a ThermoFisher Scientific
OrionTM PerpHecTTM ROSSTM combination pH microelectrode was used to monitor the pH during the irradiation
cycles. The sample was irradiated with λ = 450 nm light (100% intensity), 300 rpm and allowed to recover in the dark
under stirring. For consecutive switching cycles this program was immediately restarted after finishing for the required
number of cycles. pH values were recorded at the end of each irradiation cycle and at the end of the recovery in the
dark. pH readings were stable and did not fluctuate when they were recorded. A plastic shield was used as safety
measure to protect from light exposure. Starting pH values in the dark were adjusted by adding minimal amounts of
HClaq (12 M or 0.10 M), or KOHaq (10 M). The concentration of the solutions at different starting pH values was
determined by measuring the UV-vis absorption spectrum of diluted samples.
Figure S37. Setup for irradiation experiments of saturated solutions of the photoswitches: pH-meter with pH probe inserted into a
sample placed in Penn PhD photoreactor m2 (left) and sample under irradiation (right).
S10.1. Procedure for irradiating concentrated solutions – time resolved

Samples were prepared by adding the compound to 1.5 mL of a 20 mM KCl solution, sonicated in an ultrasound bath for
10 min and filtered through a syringe filter. An equivalent of concentrated HCl was added. A 10% dilution of this
sample was prepared by adding 136 μL of the concentrated solution to 1224 μL of 20 mM KCl solution. Both solutions
were stirred at r.t. for ca. 1 h to ensure complete equilibration of the system.
The vial was placed in a Penn PhD photoreactior m2. A Metrohm 913 pH meter in conjunction with the Metrohm LL
Biotrode 3mm WOC and a Pt1000 Temp Sensor Steel was used to monitor the pH and the temperature during the
irradiation, and thermal recovery in the dark. The sample was irradiated with λ = 450 nm light (100% power intensity,
25% power intensity), 300 rpm and allowed to recover in the dark while stirring. A plastic shield was used as safety
measure to protect from light exposure. Starting pH values in the dark were adjusted by adding equivalents (~ 0.5-1 μL)
of HClaq (12 M or 1.2 M) or KOHaq (10 M or 1.0 M). The concentration of the solutions at different starting pH values
was determined by measuring the UV-vis absorption spectrum of diluted samples.
76
S10.2. Light induced pH switching of photoacids 3c and 3d
Results of switching concentrated solutions of photoacid 3d are shown in Table S7 and Fig. S38. pH drops under
irradiation (ΔpHhν) and recovery in the dark (increase in pH) of two separate samples were measured over time.
Table S7. Light induced pH changes at different starting pH values of photoacid 3d.
Under hν # Irradiation
Initial pH value Conc. / mM ΔpHhν
(λ = 450 nm) cycles
1.60 1.23 < 8.6 0.37 1
5.45 2.49 < 7.9 2.96 1
5.79 2.69 < 7.9 3.10 1
Sample 1 6.00 2.86 7.9 3.14 5
6.42 4.02 7.8 2.40 5
7.54 5.96 8.6 1.58 5
12.04 11.75 < 8.6 0.29 1
1.21 1.03 8.0 0.18 1
5.20 2.57 7.8 2.63 1
Sample 2 6.51 3.26 8.0 3.24 16
6.98 4.18 7.9 2.79 5
8.78 8.70 7.8 0.08 6
Figure S38. pH changes under irradiation with λ = 450 nm (pH drop over 2 min) and recovery in the dark (pH increase over ~ 2–20
min) of concentrated solutions of photoacid 3d (left: sample 1, right: sample 2). Respective concentrations are indicated in the figure.
77
Results of switching concentrated solutions of photoacid 3c are shown in Fig. S39 and Table S8. pH drops under
irradiation (ΔpHhν) and recovery in the dark (increase in pH) of two separate samples were measured over time.
Table S8. Light induced pH changes at different starting pH values of photoacid 3c.
Under hν # Irradiation
Initial pH value Conc. / mM ΔpHhν
(λ = 450 nm) cycles
2.80 2.31 < 17.3 0.49 1
5.22 2.56 < 17.3 2.66 1
5.38 2.72 17.3 2.66 5
5.66 3.07 > 17.3 2.59 1
5.79 3.23 > 18.9 2.56 1
Sample 1
5.93 3.45 > 18.9 2.48 1
6.00 3.37 19.5 2.63 5
6.04 3.86 > 17.3 2.18 1
7.23 6.25 18.9 0.97 5
12.85 12.77 > 18.9 0.08 1
2.08 1.70 12.1 0.38 1
4.79 2.08 12.1 2.71 1
5.07 2.36 12.1 2.71 1
Sample 2
6.10 3.57 12.1 2.53 1
6.44 4.32 11.3 2.12 1
11.96 11.73 12.0 0.23 1
Figure S39. pH changes under irradiation with λ = 450 nm light (pH drop over 1–2 min) and recovery in the dark (pH increase over
~ 1–10 min) of concentrated solutions of photoacid 3c (left: sample 1, right: sample 2). Respective concentrations are indicated in the
figure.
78
S10.3. Time resolved pH changes under light irradiation
Photoacid 3d:
a) b)
c) d)
e) f)
g) h)
Figure S40. Light irradiation (λ = 450 nm) of solutions of photoacid 3d followed by thermal equilibration in the dark. pH profiles are
shown on the left and respective temperature profiles are shown on the right. The initial pH value in the dark (pHdark) is indicated in
the legend.
c of 3d / mM Power / %
a), b) 8.5 25
c), d) 8.5 100 [KCl]aq = 20 mM
e), f) 0.9 25 hν (λ = 450 nm)
g), h) 0.9 100
79
Photoacid 3c:
a) b)
c) d)
e) f)
g) h)
Figure S41. Light irradiation (λ = 450 nm) of solutions of photoacid 3c followed by thermal equilibration in the dark. pH profiles are
shown on the left and respective temperature profiles are shown on the right. The initial pH value in the dark (pHdark) is indicated in
the legend.
c of 3c / mM Power / %
a), b) 8 25
c), d) 8 100 [KCl]aq = 20 mM
e), f) 0.8 25 hν (λ = 450 nm)
g), h) 0.8 100
80
S11. pH drop model
To determine the possible pH drop of an aqueous solution of a merocyanine photoacid under irradiation we developed a
model which requires the concentration, pKadark and pKahν values. A basic estimation of the ratio of photoswitching in
concentrated solutions requires extinction coefficients of the individual forms and details of the light source.
S11.1. pH dependent concentration of MCH calculated from pKadark

Using the Hendersson Hasselbach equation (Eq. S52) the mole fraction of MCH (Eq. S53) can be determined from
𝑝𝑝𝐾𝐾𝑎𝑎 𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑 by taking into account that the deprotonated form consists of both MC and SP:
[𝑀𝑀𝑀𝑀 + 𝑆𝑆𝑆𝑆]
𝑝𝑝𝑝𝑝 = 𝑝𝑝𝐾𝐾𝑎𝑎 𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑 + log � � (Eq. S52)
[𝑀𝑀𝑀𝑀𝑀𝑀]
[𝑀𝑀𝑀𝑀 + 𝑆𝑆𝑆𝑆] 𝑐𝑐 − [𝑀𝑀𝑀𝑀𝑀𝑀]
[𝑀𝑀𝑀𝑀𝑀𝑀] = 𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑 = 𝑑𝑑𝑎𝑎𝑟𝑟𝑟𝑟
10𝑝𝑝𝑝𝑝−𝑝𝑝𝐾𝐾𝑎𝑎 10𝑝𝑝𝑝𝑝−𝑝𝑝𝐾𝐾𝑎𝑎
1
χ𝑀𝑀𝑀𝑀𝑀𝑀 = 𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑 (Eq. S53)
1 + 10𝑝𝑝𝑝𝑝−𝑝𝑝𝐾𝐾𝑎𝑎
S11.2. Extent of photoswitching

We estimated the extent of photoswitching of concentrated solutions by considering the power of the light source and
the concentration of MCH.
Approximations of the model

- Calculated for one wavelength not the entire emission spectrum of the light source
- Only isomerization of MCH to SP is taken into account and the role of MC is omitted. This approximation is
reasonable as the quantum yield of MC to form SP is <1%.
- All reacted MCH forms SP and released protons in solution.
- Protonation is extremely fast, so that (de)protonation of MC and water are immediate.
Model description
We define a new equilibrium constant under irradiation Khν consisting of the thermal rate constants kopen and kclose and a
rate constant under irradiation khv.
𝑘𝑘𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 + 𝑘𝑘ℎ𝜈𝜈
𝐾𝐾ℎ𝜈𝜈 = (Eq. S54)
We use Khν to estimate the equilibrium position under irradiation by determining when the net rate approaches zero.
𝑛𝑛𝑛𝑛𝑛𝑛 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟 = 𝑘𝑘𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 + 𝑘𝑘ℎ𝜈𝜈 − 𝑘𝑘𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜 (Eq. S55)
The photonflux is calculated as follows:

𝑘𝑘𝑘𝑘𝑚𝑚2
𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝 [𝑊𝑊 = ]
𝑠𝑠 3
−1
𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓[𝑠𝑠 ] = 1000 (Eq. S56)
𝑚𝑚
𝑚𝑚2 𝑘𝑘𝑘𝑘 𝐶𝐶 � 𝑠𝑠 �
ℎ� �∙
𝑠𝑠 𝜆𝜆 [𝑚𝑚]
Determining rate of photon absorption rateabs of the sample:

𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓 [𝑠𝑠 −1 ]
𝑚𝑚𝑚𝑚𝑚𝑚 𝑁𝑁 [𝑚𝑚𝑚𝑚𝑚𝑚 −1 ] (Eq. S57)
𝑟𝑟𝑟𝑟𝑟𝑟𝑒𝑒𝑎𝑎𝑎𝑎𝑎𝑎 � � = 𝐴𝐴
𝐿𝐿 ∙ 𝑠𝑠 𝑉𝑉 [𝐿𝐿]
81
The rate constant under irradiation is given by khv:
𝑘𝑘ℎ𝜈𝜈 = [𝑀𝑀𝑀𝑀𝑀𝑀] ∙ ∙ (1 − 10−𝐴𝐴𝐴𝐴𝐴𝐴𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 ) ∙ 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟_𝑎𝑎𝑎𝑎𝑎𝑎 (Eq. S58)
𝐴𝐴𝐴𝐴𝐴𝐴𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡
[𝑀𝑀𝑀𝑀𝑀𝑀] = light absorbed by MCH
𝐴𝐴𝐴𝐴𝐴𝐴𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡
1 − 10−𝐴𝐴𝑏𝑏𝑏𝑏𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 = transmitted light
We incrementally changed the concentration of SP under irradiation ([𝑆𝑆𝑆𝑆]′, by setting a value for y) and calculating
respective initial concentration of MCH ([𝑀𝑀𝑀𝑀𝑀𝑀]′) and MC ([𝑀𝑀𝑀𝑀]′).
[𝑆𝑆𝑆𝑆]′ = 𝑦𝑦 ∙ 𝑐𝑐 (Eq. S59)
[𝑀𝑀𝑀𝑀]′ = 𝑥𝑥 ∙ 𝑐𝑐 ∙ (1 − 𝑦𝑦)
[𝑀𝑀𝑀𝑀𝑀𝑀]′ = (1 − 𝑥𝑥) ∙ 𝑐𝑐 ∙ (1 − 𝑦𝑦)
To consider the equilibrium between MCH and MC and water the equilibrated concentrations under irradiation are
calculated as follows:
[𝑀𝑀𝑀𝑀𝑀𝑀] = [𝑀𝑀𝑀𝑀𝑀𝑀]′ + 𝛼𝛼
[𝑀𝑀𝑀𝑀] = [𝑀𝑀𝑀𝑀]′ − 𝛼𝛼
[𝐻𝐻 + ] = [𝐻𝐻 + ]′ − 𝛼𝛼 − 𝑧𝑧
[𝑂𝑂𝑂𝑂 − ] = [𝑂𝑂𝑂𝑂 − ]′ − 𝑧𝑧
Parameters α and z are calculated from two protonation equilibria described by Ka (Eq. S60) and Kw (Eq. S61) using a
circular reference in excel. Iterations were set to a high value and maximum change to a very low value.
([𝐻𝐻 + ] − 𝛼𝛼 − 𝑧𝑧) ([𝑀𝑀𝑀𝑀] − 𝛼𝛼)
𝐾𝐾𝑎𝑎 = (Eq. S60)
([𝑀𝑀𝑀𝑀𝑀𝑀] + 𝛼𝛼)
𝐾𝐾𝑤𝑤 = ([𝐻𝐻 + ] − 𝛼𝛼 − 𝑧𝑧) ([𝑂𝑂𝑂𝑂 − ] − 𝑧𝑧) (Eq. S61)
0 = 𝛼𝛼 − 𝛼𝛼([𝐻𝐻 − 𝑧𝑧 + [𝑀𝑀𝑀𝑀] + 𝐾𝐾𝑎𝑎 ) + ([𝐻𝐻 + ] − 𝑧𝑧)[𝑀𝑀𝑀𝑀] − 𝐾𝐾𝑎𝑎 [𝑀𝑀𝑀𝑀𝑀𝑀] (𝐼𝐼)
2 +]
𝑏𝑏𝛼𝛼 = −([𝐻𝐻 + ] − 𝑧𝑧 + [𝑀𝑀𝑀𝑀] + 𝐾𝐾𝑎𝑎 )

𝑐𝑐𝛼𝛼 = ([𝐻𝐻 + ] − 𝑧𝑧)[𝑀𝑀𝑀𝑀] − 𝐾𝐾𝑎𝑎 [𝑀𝑀𝑀𝑀𝑀𝑀]
0 = 𝑧𝑧 2 − 𝑧𝑧 ([𝑂𝑂𝑂𝑂 − ] + [𝐻𝐻 + ] − 𝛼𝛼) + ([𝐻𝐻 + ] − 𝛼𝛼)[𝑂𝑂𝑂𝑂 − ] − 𝐾𝐾𝑤𝑤 (𝐼𝐼𝐼𝐼)
𝑏𝑏𝑧𝑧 = −([𝑂𝑂𝑂𝑂 − ] + [𝐻𝐻 + ] − 𝛼𝛼)
𝑐𝑐𝑧𝑧 = ([𝐻𝐻 + ] − 𝛼𝛼)[𝑂𝑂𝑂𝑂 − ] − 𝐾𝐾𝑤𝑤
By arbitrarily changing the concentration of SP (Eq. S59) under irradiation we determined the equilibrium under
irradiation at which the net rate = 0 (Eq. S55).
From this model we predicted that ~100% of the MCH would be isomerized to SP at equilibrium under the irradiation
conditions we used. We therefore simplified the model to assume 100% isomerization of MCH under our experimental
conditions.
82
S11.3. Equilibria affecting proton release under irradiation
Scheme S2. Equilibria affecting the concentration of released protons under irradiation.
Protons that are released by MCH upon irradiation are in equilibrium with the autoionization of water described by Kw:
𝐾𝐾𝑊𝑊 = [𝑂𝑂𝑂𝑂 − ][𝐻𝐻3 𝑂𝑂+ ] = 10−14
Population of the cis-MCH form at low pH values also affects the concentration of released protons under irradiation
([𝐻𝐻 + ]𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟 ) by acting as a proton sink. Since the cis-MCH is only formed at very low pH values and the pH change
under irradiation becomes increasingly small at low pH values we approximate the concentration of cis-MCH under
irradiation by calculating [cis-MCH] at the starting pH value in the dark.
[𝐻𝐻 + ]𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟 = [MCH]𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 − [cis-MCH]
[MCH]𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 = χ𝑀𝑀𝑀𝑀𝑀𝑀 ∙ 𝐶𝐶
χ𝑀𝑀𝑀𝑀𝑀𝑀 from Eq. S53
[cis-MCH] = χ𝑐𝑐𝑐𝑐𝑐𝑐−𝑀𝑀𝑀𝑀𝑀𝑀 ∙ 𝐶𝐶
Concentrations after/during irradiation and changes in equilibrium position:

At initial equilibrium [𝐻𝐻 + ]𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 [𝑂𝑂𝑂𝑂 − ]𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖
After/during hν [𝐻𝐻 + ]𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 + [𝐻𝐻 + ]𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟 [𝑂𝑂𝑂𝑂 − ]𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖
Change in concentration due to −𝑥𝑥𝐻𝐻2𝑂𝑂 −𝑥𝑥𝐻𝐻2𝑂𝑂
autoionization of water
Equilibrium after/during hν [𝐻𝐻 + ]𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 + [𝐻𝐻 + ]𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟 − 𝑥𝑥𝐻𝐻2𝑂𝑂 [𝑂𝑂𝑂𝑂 − ]𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 − 𝑥𝑥𝐻𝐻2𝑂𝑂
With [𝐻𝐻 + ] = [H + ]𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 + [𝐻𝐻 + ]𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟

𝐾𝐾𝑊𝑊 = �[𝐻𝐻 + ] − 𝑥𝑥𝐻𝐻2𝑂𝑂 ��[𝑂𝑂𝑂𝑂 − ]𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 − 𝑥𝑥𝐻𝐻2𝑂𝑂 �
This gives us quadratic equation:

0 = 𝑥𝑥𝐻𝐻2𝑂𝑂 2 − 𝑥𝑥𝐻𝐻2𝑂𝑂 ([𝑂𝑂𝑂𝑂 − ]𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 + [𝐻𝐻 + ]) + [𝐻𝐻 + ][𝑂𝑂𝑂𝑂 − ]𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 − 𝐾𝐾𝑤𝑤
Solving the quadratic equation gives the following proton concentration at equilibrium:
[𝐻𝐻 + ]𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓 = [H + ]𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 + [𝐻𝐻 + ]𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟 − 𝑥𝑥𝐻𝐻2𝑂𝑂
The pH under irradiation is then calculated:

𝑝𝑝𝑝𝑝 = −𝑙𝑙𝑙𝑙𝑙𝑙�[𝐻𝐻 + ]𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓 �
83
S11.4. Predicted pH changes under irradiation of photoacids 3c and 3d
Predicted changes in pH (ΔpHhν) upon irradiation (λ = 450 nm) of photoacids 3c and 3d based on the initial pH values
in the dark before irradiation (pHdark) are shown below and compared to experimentally determined changes in pH
(Figure S42). The power of the photoreactor was varied between 100% and 25% in order to observe the difference in
achieved pH changes. For most samples using 100% power achieved a slightly higher ΔpHhν indicating that the bleach
with 25% power may not be complete. Increasing the concentration of the photoacid by one order of magnitude resulted
in an increase of the predicted pH change of one pH unit. This was confirmed with experimental results of compound
3d.
For compound 3c concentrations above 1 mM result in aggregation which was confirmed by NMR (see Section S5),
which also leads to divergence from the pH drop model at higher concentrations. For a concentration of 0.8 mM,
however, the model is in similar agreement to the data as it is for compound 3d.
The divergence of the model at higher pH values is attributed to incomplete bleaching due to increasing concentration
of MC and the very low quantum yield of photoisomerising MC (see Section S7). At high pH the autoionisation of
water also becomes important. An overview of the pH resolved speciation for the individual compounds is in Section
S6.7.
Figure S42. Predicted pH changes upon irradiation (ΔpHhν, λ = 450 nm) at different starting pH values in the dark (pHdark) compared
to the experimental results for compounds 3c (left) and 3d (right) for different power settings of the photoreactor.
84
S11.5. Application of the pH drop model to literature examples
Variation of side chain length of photoacid 2a
Figure S43. Structures of literature reported photoacids 4, 2a and 5.11
Table S9. pKa values, solubility and Kc values of literature reported photoacids 4, 2a and 5.11
Photoacids pKadark pKahν Solubility / mM pKa Kc

4 5.74 2.14 0.56 7.23 29
2a 6.19 2.47 0.19 7.2 8.6
5 6.56 3.07 2.66 7.25 3.7
a) b)
2a
c) d)
4 5
Figure S44. a) Application of the pH drop model to literature reported photoacids 4, 2a and 5 using pKadark, pKahν and the solubility
to estimate the change in pH upon irradiation. b)-d) Calculated mole fractions of 4, 2a and 5 using pKa, Kc and pKahν.
85
Figure S45. Comparison of predicted pH changes upon irradiation of photoacids 3c-d from this study with literature reported
photoacids 2a, 4 and 5.
Side chain length variation and substitution with electron withdrawing groups on the phenol moiety
Figure S46. Structures of previously reported photoacids 6-11.12
Table S10. pKa values and Kc values of literature reported photoacids 6-11.12
Photoacids pKadark, a pKahν, b Solubility / mM pKac Kcd

6 3.3 0.4 − 3.7 1.45
7 3.9 0.4 − 4.2 1.15
8 4.3 1.6 − 4.4 0.27
9 4.1 1.4 − 4.2 0.34
10 4.0 0.4 − 4.5 2.38
11 3.5 0.4 − 4.4 6.25
a) Calculated from 𝑝𝑝𝐾𝐾𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑
𝑎𝑎 = −𝑙𝑙𝑙𝑙𝑙𝑙(𝐾𝐾𝑎𝑎 (1 + 𝐾𝐾𝑐𝑐 )).
b) reported as pKaI.
c) reported as pKaII.
d) equilibrium was defined the other way around, therefore Kc = (Kc-reported)–1.
86
a) b)
c) d)
6 7
e) f)
8 9
g) h)
10 11
Figure S47. Application of the pH drop model to literature reported photoacids 6-11 using pKadark, pKahν and an arbitrary
concentration of 1mM in a) and 10 mM in b) to estimate the change in pH upon irradiation. c)-h) Calculated mole fractions of 6-11
using pKa, Kc and pKahν.
87
S12. Model describing the time dependent pH recovery in the dark
S12.1. Model description
The time taken for the pH of a photoirradiated solution of a photoacid to return to its pre-irradiation value can vary
significantly, and depends strongly on the pre- and post-irradiation pH values. Here we describe an approach which
allows us to model the pH recovery in the dark, using a combined equilibrium and kinetic model. Most of the input
parameters for this model come from the time-resolved experimental pH recovery studies described in section S10.3.
The model was fit to experimental data and the experimentally derived parameters (pKahv, pKa, ko, and kc) were varied
within their expected experimental error ranges. The input parameter totalconc (defined as the total concentration of
photoacid-related species) was also allowed to vary within about 20%, to account for experimental error. The errors on
ko and kc are most significant since these parameters are very sensitive to temperature. Each dataset was also fit to a
non-experimental parameter “bleach”, which is related to the extent of photobleaching achieved under irradiation,
which likely also captures other sources of error (e.g. concentration, temperature).
The list of kinetic parameters returned by the model can be found in Tables S12-15.
Input parameters:
𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐: 𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡
The starting concentrations of cis-MCH and SP are calculated from the pH value reached under light irradiation and the
factor “bleach” which was fit to the experimental data.
From Eq. S31:
[cis-MCH] = 𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏ℎ ∙ (𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 ∙ χcis-MCH )
From Eq. S32:
[𝑆𝑆𝑆𝑆] = 𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏ℎ ∙ �𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 ∙ χ𝑆𝑆𝑆𝑆ℎ𝜈𝜈 �
The factor “bleach” allows for calculation of the remaining MCH concentration under irradiation:
[𝑀𝑀𝑀𝑀𝐻𝐻] = (1 − 𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏ℎ) ∙ 𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡
Kinetic equilibrium system:
The following equilibrium (Scheme S3) is modelled with a kinetic approach. In our development of the model we also
incorporated a cis-trans isomerization step between MCH and cis-MCH, but this did not appreciably affect the fit, and
its inclusion is not yet supported by experimental data or estimates of the rate constants. Therefore this cis-MCH to
MCH process was not included in our model.
Scheme S3. Equilibrium modelled with a kinetic approach for the thermal pH recovery in the dark.
Protonation equilibria:
The following protonation equilibria are assumed to be many orders of magnitude faster than other kinetic processes in
the equilibrium system. The changes in concentration due to protonation events were therefore calculated for every time
step based on the respective equilibrium constants.
Scheme S4. Protonation equilibria modelled with their respective equilibrium constants for the thermal pH recovery in the dark.
88
We used the approach described by Maeder and coworkers to simulate the combined equilibrium-kinetic system.16
Maeder’s algorithm involves including the solution of the multi-component equilibrium problem using a Newton-
Raphson algorithm into the system of differential equations describing the kinetic system. We re-implemented this
algorithm in Python, and the code is available on github (github.com/martinp23/photoacid-fitting).
Experimental data were fit to the model by minimising the normalized mean squared errors between the experimental
datasets and the simulated data, while varying the values of the parameters within bounds as described above (where the
bounds on “bleach” were typically 0.7–1.0). The system is very sensitive to the choice of initial values of the
parameters, and subtle variations reliably led to (usually subtly) different outputs from the L-BFGS minimisation
routine (from scipy.optimize.minimize). We therefore used a differential evolution algorithm (again, from
scipy.optimize) to explore the parameter space, with a maximum of 100 iterations. More iterations appeared to have
negligible effect on the goodness-of-fit, both quantitatively and when evaluated graphically. The optimum parameter
values from differential evolution were subject to L-BFGS optimisation (“polishing”). The differential evolution
algorithm is parallelizable in the scipy implementation. In our hands, with 12 CPU cores, 4-5 experimental pH traces,
treated as a single data set, could be fit in parallel within about 15 minutes. The fits shown in Figure S49 are from fitting
several pH experiments with the same compound, concentration, and light intensity simultaneously.
Based on our experimental results the temperature fluctuated < 1 °C for 3c and 3d when using 25% power of the
photoreactor. When using 100% power it fluctuated < 3 °C for 3c and < 2 °C for 3d. We calculated values at 22.5 °C
for rate constants ko and kc based on our earlier described variable temperature experiments (see Section S6.6, additional
data in Fig. S48) and determined an error based on a 2.5°C temperature fluctuation (see Table S11). We used these
parameters and similar errors (though the error bound on ko was broadened slightly) for fitting the model to the
experimental data. It is important to clarify that the model and fitting code do not take account of the temperature
fluctuations as a function of time during the experiment, and so the final fit values of kc and ko are compromise values
which offer the best fit throughout the experimental temperature range.
89
Table S11. Rate constants ko and kc of compounds 3c and 3d at 22.5 °C with a temperature-based error (± 2.5 °C).
# T / °C ko / s–1 kc / s–1
3c 22.5 ± 2.5 0.0032 ± 0.002 0.18 ± 0.07
3d 22.5 ± 2.5 9.0 x 10–4 ± 6 x 10–4 0.032 ± 0.01
a) b)
c) d)
Figure S48. Variable temperature experiments to determine a temperature-based error (± 2.5 °C) on ko for compounds 3c and 3d. a)
An aqueous sample of photoacid 3c was added to a buffer solution in situ at low pH to monitor equilibration processes of MCH. The
MCH absorption band (λ = 451 nm) was monitored at different temperatures (20-25 °C) with corresponding exponential fits (Eq. S2)
depicted as black lines. Experimental conditions: [3c] = 16.9 μM, [phosphate buffer] = 20 mM, pH 4.3, T = 20-25 °C. b) Eyring plot
for observed rate constants kobs (= ko) obtained from the exponential fit in a). c) An aqueous sample of photoacid 3d was added to a
buffer solution in situ at low pH to monitor equilibration processes of MCH. The MCH absorption band (λ = 442 nm) was monitored
at different temperatures (20-25 °C) with corresponding exponential fits (Eq. S2) depicted as black lines. Experimental conditions:
[3d] = 16.3 μM, [phosphate buffer] = 20 mM, pH 3.7, T = 20-25 °C. b) Eyring plot for observed rate constants kobs (= ko) obtained
from the exponential fit in c).
As shown in Fig. S49, the kinetic/equilibrium model with the addition of a fitted “bleach” parameter is able to
adequately reproduce the experimental pH vs time traces for photoacids 3c and 3d in the dark. The final pH of the
system is typically reproduced to within 0.1 pH units. The model is invalidated by large temperature changes. The
model can be trivially extended to include the cis-trans isomerism of MCH as a kinetic step, but we found that this
addition gave only a modest improvement to the goodness-of-fit. We opted not to include it because of uncertainty
around its true mechanistic role in the dark for this system, and because no suitable estimate of kinetic parameters
exists.
The initial parameters and final fitted parameters are given in the tables below.
90
S12.2. Results of modelled parameters for compounds 3c and 3d
Photoacid 3c:
Table S12. Results from the model for time resolved pH measurements of compound 3c (8 mM) using 25% LED power.
Normalized mean
Parameter Initial range Optimised result Experimental data
squared error ×1000
pKalight 1.97 – 1.99 1.97 1.98 ± 0.01
kclose 0.11 – 0.20 0.11 0.18 ± 0.07
pKa(MCH) 7.08 – 7.28 7.28 7.18 ± 0.10
kopen 1.00 x 10–3 – 1.00 x 10–2 0.0056 0.0032 ± 0.002
3.76
totalconc 6.00 x 10–3 – 1.00 x 10–2 0.0087 0.008
bleach0 (pHinitial = 4.48) 0.30 – 1.00 1.00 –
Table S13. Results from the model for time resolved pH measurements of compound 3c (0.8 mM) using 25% power.
Normalized mean
squared error × 1000
pKalight 1.97 – 1.99 1.97 1.98 ± 0.01
kclose 0.11 – 0.20 0.11 0.18 ± 0.07
pKa(MCH) 7.08 – 7.28 7.12 7.18 ± 0.10
kopen 1.00 x 10–3 – 1.00 x 10–2 0.0082 0.0032 ± 0.002
totalconc 6.50 x 10–4 – 9.50 x 10–4 0.0008 0.0008
7.43
91
Photoacid 3d:
Table S14. Results from the model for time resolved pH measurements of compound 3d (8.5 mM) using 25% LED power.
Normalized mean
pKalight 1.75 – 1.81 1.76 1.78 ± 0.03
kclose 0.01 – 0.053 0.017 0.032 ± 0.01
pKa(MCH) 6.89 – 7.05 6.99 6.97 ± 0.08
kopen 0.0001 – 0.002 0.0020 0.0009 ± 0.0006
totalconc 0.0075 – 0.0095 0.0075 0.0009
2.49
Table S15. Results from the model for time resolved pH measurements of compound 3d (0.9 mM) using 25% power.
Normalized mean
squared error x 1000
pKalight 1.75 – 1.81 1.76 1.78 ± 0.03
kclose 0.01 – 0.053 0.022 0.032 ± 0.01
pKa(MCH) 6.89 – 7.05 6.89 6.97 ± 0.08
kopen 0.0005 – 0.005 0.0011 0.0009 ± 0.0006
totalconc 0.0008 – 0.001 0.0008 0.0009
5.48
Table S16. Results from the model for time resolved pH measurements of compound 3d (0.9 mM) using 100% power.
Normalized mean
pKalight 1.75 – 1.81 1.76 1.78 ± 0.03
kclose 0.01 – 0.0530 0.02 0.032 ± 0.01
pKa(MCH) 6.89 – 7.05 6.90 6.97 ± 0.08
kopen 0.0001 – 0.002 0.0020 0.0009 ± 0.0006
totalconc 0.0008 – 0.001 0.0008 0.0009
5.32
92
S12.3. Modelled pH recovery of photoacid 3c and 3d
The modelled pH recovery in the dark of photoacids 3c and 3d is shown in Fig. S49 and compared to the
experimentally determined pH recovery.
a) b)
c) d)
e)
Figure S49. pH recovery in the dark after light irradiation (λ = 450 nm) in comparison to the model.
# c / mM Power / %
a) 3c 8 25
b) 3c 0.8 25
[KCl]aq = 20 mM
c) 3d 8.5 25
hν (λ = 450 nm)
d) 3d 0.9 25
e) 3d 0.9 100
93
S13. Propagation of errors
pKadark calculated from Kc (calc. from Eq. S5) and Ka (from fit to φ Eq. S42)
𝐾𝐾𝑎𝑎𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑 = 𝐾𝐾𝑎𝑎 (1 + 𝐾𝐾𝑐𝑐 )
𝑝𝑝𝑝𝑝𝑎𝑎𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑 = − log(𝐾𝐾𝑎𝑎𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑 ) = − log�𝐾𝐾𝑎𝑎 (1 + 𝐾𝐾𝑐𝑐 )�
∂𝑝𝑝𝑝𝑝𝑎𝑎𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑 ∂ 1 1
= (− log(𝐾𝐾𝑎𝑎 + 𝐾𝐾𝑎𝑎 𝐾𝐾𝑐𝑐 )) = − (1 + 𝐾𝐾𝑐𝑐 ) = −
∂𝐾𝐾𝑎𝑎 ∂𝑝𝑝𝑝𝑝𝑎𝑎𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑 (𝐾𝐾𝑎𝑎 + 𝐾𝐾𝑎𝑎 𝐾𝐾𝑐𝑐 ) ln(10) 𝐾𝐾𝑎𝑎 ln(10)
∂𝑝𝑝𝑝𝑝𝑎𝑎𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑 1 1
=− 𝐾𝐾 = −
∂𝐾𝐾𝑐𝑐 (𝐾𝐾𝑎𝑎 + 𝐾𝐾𝑎𝑎 𝐾𝐾𝑐𝑐 ) ln(10) 𝑎𝑎 (1 + 𝐾𝐾𝑐𝑐 ) ln(10)
2 2 2 2
∂𝑝𝑝𝑝𝑝𝑎𝑎𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑 ∂𝑝𝑝𝑝𝑝𝑎𝑎𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑 1 1
𝛿𝛿𝑝𝑝𝑝𝑝𝑎𝑎𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑 = �� (𝛿𝛿𝐾𝐾𝑎𝑎 )2 + � � (𝛿𝛿𝐾𝐾𝑐𝑐 )2 = ��− � (𝛿𝛿𝐾𝐾𝑎𝑎 )2 + �− � (𝛿𝛿𝐾𝐾𝑐𝑐 )2
∂𝐾𝐾𝑎𝑎 ∂𝐾𝐾𝑐𝑐 𝐾𝐾𝑎𝑎 ln(10) (1 + 𝐾𝐾𝑐𝑐 ) ln(10)
𝜫𝜫
𝛱𝛱 = 𝑝𝑝𝑝𝑝𝑎𝑎𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑 − 𝑝𝑝𝐾𝐾ℎ𝜈𝜈
𝑎𝑎
𝛿𝛿𝛿𝛿 = �(𝛿𝛿𝑝𝑝𝑝𝑝𝑎𝑎𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑 )2 + (𝛿𝛿𝑝𝑝𝑝𝑝𝑎𝑎ℎ𝜈𝜈 )2
Eyring plot analysis

𝑘𝑘𝑜𝑜𝑜𝑜𝑜𝑜 𝛥𝛥𝛥𝛥 ‡ 1 𝛥𝛥𝛥𝛥 ‡ 𝑘𝑘𝑏𝑏
ln � � = �− �� + � � + ln � �
𝑇𝑇 𝑅𝑅 𝑇𝑇 𝑅𝑅 ℎ
With linear equation:
𝑦𝑦 = 𝑚𝑚𝑚𝑚 + 𝑏𝑏
𝜟𝜟𝜟𝜟‡ :
𝛥𝛥𝛥𝛥 ‡ = 𝑚𝑚 ∙ 𝑅𝑅
2
∂𝛥𝛥𝛥𝛥 ‡
𝛿𝛿𝛥𝛥𝛥𝛥 = ��
‡
� (𝛿𝛿𝛿𝛿)2 = �(R)2 (𝛿𝛿𝛿𝛿)2
∂𝑚𝑚
𝑘𝑘𝑏𝑏
𝛥𝛥𝛥𝛥 ‡ = 𝑅𝑅 �𝑏𝑏 − ln � ��
ℎ
2
∂𝛥𝛥𝛥𝛥 ‡
𝛿𝛿𝛿𝛿𝛿𝛿 ‡ = �� (𝛿𝛿𝛿𝛿)2 = �(R)2 (𝛿𝛿𝛿𝛿)2
∂𝑏𝑏
𝛥𝛥𝛥𝛥 ‡ = 𝛥𝛥𝛥𝛥 ‡ − 𝑇𝑇𝛥𝛥𝛥𝛥 ‡
2 2
∂𝛥𝛥𝛥𝛥 ‡ ∂𝛥𝛥𝐺𝐺 ‡
𝛿𝛿𝛥𝛥𝛥𝛥 = ��
‡
� (𝛿𝛿𝛥𝛥𝛥𝛥 ‡ )2 + � � (𝛿𝛿𝛥𝛥𝛥𝛥 ‡ )2 = �(𝛿𝛿𝛥𝛥𝛥𝛥 ‡ )2 + (−T)2 (𝛿𝛿𝛥𝛥𝛥𝛥 ‡ )2
∂𝛥𝛥𝛥𝛥 ‡ ∂𝛥𝛥𝛥𝛥 ‡
𝜟𝜟𝜟𝜟:
𝛥𝛥𝛥𝛥 = −𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝐾𝐾𝑐𝑐
∂𝛥𝛥𝛥𝛥 2 ∂ 2
1 2
𝛿𝛿𝛿𝛿𝛿𝛿 = �� (𝛿𝛿𝐾𝐾𝑐𝑐 )2 = �� (−𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅(𝐾𝐾𝑐𝑐 )� (𝛿𝛿𝐾𝐾𝑐𝑐 )2 = ��(−𝑅𝑅𝑅𝑅 � (𝛿𝛿𝐾𝐾𝑐𝑐 )2
∂𝐾𝐾𝑐𝑐 ∂𝐾𝐾𝑐𝑐 𝐾𝐾𝑐𝑐
𝑲𝑲𝒄𝒄 :
𝑘𝑘𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐
𝐾𝐾𝑐𝑐 =
2 2
𝛿𝛿𝛿𝛿𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜 𝛿𝛿𝑘𝑘𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐
𝛿𝛿𝐾𝐾𝑐𝑐 = |𝐾𝐾𝑐𝑐 |�� +� �
𝑘𝑘𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜 𝑘𝑘𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐
pKa calculated from experimentally determined pKadark (Eq. S1) and Kc (calc. from Eq. S5):
𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑
10−𝑝𝑝𝑝𝑝𝑎𝑎
p𝐾𝐾𝑎𝑎 = −log � � = 𝑝𝑝𝑝𝑝𝑎𝑎𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑 + log(1 + 𝐾𝐾𝑐𝑐 )
94
∂p𝐾𝐾𝑎𝑎 2 ∂p𝐾𝐾𝑎𝑎 2 1 2
𝛿𝛿p𝐾𝐾𝑎𝑎 = �� (𝛿𝛿𝑝𝑝𝑝𝑝 𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑 )2 + �
𝑎𝑎 � (𝛿𝛿𝐾𝐾𝑐𝑐 ) 2 = �(𝛿𝛿𝑝𝑝𝑝𝑝 𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑 )2 + �
𝑎𝑎 � (𝛿𝛿𝐾𝐾𝑐𝑐 )2
∂𝑝𝑝𝑝𝑝𝑎𝑎𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑 ∂𝐾𝐾𝑐𝑐 (1 + 𝐾𝐾𝑐𝑐 ) ∙ ln(10)
pKadark calculated from experimentally determined pKa and Kc (calc. from Eq. S5) for compound 2d:
𝑝𝑝𝑝𝑝𝑎𝑎𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑 = p𝐾𝐾𝑎𝑎 − log(1 + 𝐾𝐾𝑐𝑐 )
2 2 2
∂𝑝𝑝𝑝𝑝𝑎𝑎𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑 ∂𝑝𝑝𝑝𝑝𝑎𝑎𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑 1
𝛿𝛿p𝐾𝐾𝑎𝑎 = �� (𝛿𝛿p𝐾𝐾𝑎𝑎 )2 + � � (𝛿𝛿𝐾𝐾𝑐𝑐 )2 = �(𝛿𝛿p𝐾𝐾𝑎𝑎 )2 + � � (𝛿𝛿𝐾𝐾𝑐𝑐 )2
∂p𝐾𝐾𝑎𝑎 ∂𝐾𝐾𝑐𝑐 (1 + 𝐾𝐾𝑐𝑐 ) ∙ ln(10)
Calculating extinction coefficients and an absorption spectrum for MC
𝝌𝝌𝑺𝑺𝑺𝑺 :
𝐾𝐾𝑐𝑐
𝜒𝜒𝑆𝑆𝑆𝑆 =
∂χ𝑆𝑆𝑆𝑆 2 1 2
1
𝛿𝛿χ𝑆𝑆𝑆𝑆 = �� (𝛿𝛿𝐾𝐾𝐶𝐶 )2 = �� (𝛿𝛿𝐾𝐾𝐶𝐶 )2 = � 𝛿𝛿𝐾𝐾 �
∂𝐾𝐾𝑐𝑐 (1 + 𝐾𝐾𝑐𝑐 )2 (1 + 𝐾𝐾𝑐𝑐 )2 𝐶𝐶
𝝌𝝌𝑴𝑴𝑴𝑴 :
1
𝜒𝜒𝑀𝑀𝑀𝑀 =
∂χ𝑀𝑀𝑀𝑀 2 1 2
1
𝛿𝛿χ𝑀𝑀𝑀𝑀 = �� 2 �
� (𝛿𝛿𝐾𝐾𝐶𝐶 ) = �− � (𝛿𝛿𝐾𝐾𝐶𝐶 )2 = �− 𝛿𝛿𝐾𝐾 �
∂𝐾𝐾𝑐𝑐 (1 + 𝐾𝐾𝑐𝑐 )2 (1 + 𝐾𝐾𝑐𝑐 )2 𝐶𝐶
𝜺𝜺𝑴𝑴𝑴𝑴,𝝀𝝀𝒎𝒎𝒎𝒎𝒎𝒎 :
𝐾𝐾𝑐𝑐
𝐴𝐴𝑒𝑒𝑒𝑒 − 𝜀𝜀𝑆𝑆𝑆𝑆,𝜆𝜆𝑚𝑚𝑚𝑚𝑚𝑚 𝜒𝜒𝑆𝑆𝑆𝑆 𝐶𝐶 𝐴𝐴𝑒𝑒𝑒𝑒 − 𝜀𝜀𝑆𝑆𝑆𝑆,𝜆𝜆𝑚𝑚𝑚𝑚𝑚𝑚 1 + 𝐾𝐾𝑐𝑐 𝐶𝐶 𝐴𝐴𝑒𝑒𝑒𝑒 𝐴𝐴𝑒𝑒𝑒𝑒 𝐾𝐾𝑐𝑐
𝜀𝜀𝑀𝑀𝑀𝑀,𝜆𝜆𝑚𝑚𝑚𝑚𝑚𝑚 = = = + − 𝜀𝜀𝑆𝑆𝑆𝑆,𝜆𝜆𝑚𝑚𝑚𝑚𝑚𝑚 𝐾𝐾𝑐𝑐
𝜒𝜒𝑀𝑀𝑀𝑀 𝐶𝐶 1 𝐶𝐶 𝐶𝐶
𝐶𝐶
Errors of 𝐴𝐴𝑒𝑒𝑒𝑒 , 𝜀𝜀𝑆𝑆𝑃𝑃,𝜆𝜆𝑚𝑚𝑚𝑚𝑚𝑚 , 𝐶𝐶 assumed to be negligible compared to the error from Kc:
2 2
∂𝜀𝜀𝑀𝑀𝑀𝑀,𝜆𝜆𝑚𝑚𝑚𝑚𝑚𝑚 𝐴𝐴𝑒𝑒𝑒𝑒
𝛿𝛿𝜀𝜀𝑀𝑀𝑀𝑀,𝜆𝜆𝑚𝑚𝑚𝑚𝑚𝑚 = �� (𝛿𝛿𝐾𝐾𝑐𝑐 )2 = �� − 𝜀𝜀𝑆𝑆𝑆𝑆,𝜆𝜆𝑚𝑚𝑚𝑚𝑚𝑚 � (𝛿𝛿𝐾𝐾𝑐𝑐 )2
∂𝐾𝐾𝑐𝑐 𝐶𝐶
95
S14. Bibliography
(1) Mallo, N.; Foley, E. D.; Iranmanesh, H.; Kennedy, A. D. W.; Luis, E. T.; Ho, J.; Harper, J. B.; Beves, J. E.
Structure–function relationships of donor–acceptor Stenhouse adduct photochromic switches. Chem. Sci. 2018,
9, 8242-8252.
(2) Lynch, D. E.; Kirkham, A. N.; Chowdhury, M. Z. H.; Wane, E. S.; Heptinstall, J. Water soluble squaraine dyes for
use as colorimetric stains in gel electrophoresis. Dyes Pigm. 2012, 94, 393-402.
(3) Bokan, M.; Bondar, K.; Marks, V.; Gellerman, G.; Patsenker, L. D. Switchable phenolo-cyanine reporters
containing reactive alkylcarboxylic groups for fluorescence-based targeted drug delivery monitoring. Dyes
Pigm. 2018, 159, 18-27.
(4) Nanjunda, R.; Owens, E. A.; Mickelson, L.; Dost, T. L.; Stroeva, E. M.; Huynh, H. T.; Germann, M. W.; Henary,
M. M.; Wilson, W. D. Selective G-Quadruplex DNA Recognition by a New Class of Designed Cyanines.
Molecules 2013, 18, 13588-13607.
(5) Choi, H. S.; Nasr, K.; Alyabyev, S.; Feith, D.; Lee, J. H.; Kim, S. H.; Ashitate, Y.; Hyun, H.; Patonay, G.;
Strekowski, L.; Henary, M.; Frangioni, J. V. Synthesis and In Vivo Fate of Zwitterionic Near-Infrared
Fluorophores. Angew. Chem. Int. Ed. 2011, 50, 6258-6263.
(6) Balmond, E. I.; Tautges, B. K.; Faulkner, A. L.; Or, V. W.; Hodur, B. M.; Shaw, J. T.; Louie, A. Y. Comparative
Evaluation of Substituent Effect on the Photochromic Properties of Spiropyrans and Spirooxazines. J. Org.
Chem. 2016, 81, 8744-8758.
(7) Liu, J.; Tang, W.; Sheng, L.; Du, Z.; Zhang, T.; Su, X.; Zhang, S. X.-A. Effects of Substituents on Metastable-State
Photoacids: Design, Synthesis, and Evaluation of their Photochemical Properties. Chem.– Asian J. 2019, 14,
438-445.
(8) Hyun, H.; Owens, E. A.; Wada, H.; Levitz, A.; Park, G.; Park, M. H.; Frangioni, J. V.; Henary, M.; Choi, H. S.
Cartilage-Specific Near-Infrared Fluorophores for Biomedical Imaging. Angew. Chem. Int. Ed. 2015, 54, 8648-
8652.
(9) Shi, Z.; Peng, P.; Strohecker, D.; Liao, Y. Long-Lived Photoacid Based upon a Photochromic Reaction. J. Am.
Chem. Soc. 2011, 133, 14699-14703.
(10) Mallo, N.; Brown, P. T.; Iranmanesh, H.; MacDonald, T. S. C.; Teusner, M. J.; Harper, J. B.; Ball, G. E.; Beves, J.
E. Photochromic switching behaviour of donor–acceptor Stenhouse adducts in organic solvents. Chem.
Commun. 2016, 52, 13576-13579.
(11) Berton, C.; Busiello, D. M.; Zamuner, S.; Solari, E.; Scopelliti, R.; Fadaei-Tirani, F.; Severin, K.; Pezzato, C.
Thermodynamics and kinetics of protonated merocyanine photoacids in water. Chem. Sci. 2020, 11, 8457-
8468.
(12) Hammarson, M.; Nilsson, J. R.; Li, S.; Beke-Somfai, T.; Andréasson, J. Characterization of the Thermal and
Photoinduced Reactions of Photochromic Spiropyrans in Aqueous Solution. J. Phys. Chem. B 2013, 117,
13561-13571.
(13) Andraos, J. A Streamlined Approach to Solving Simple and Complex Kinetic Systems Analytically. J. Chem.
Educ. 1999, 76, 1578.
(14) Aldaz, C. R.; Wiley, T. E.; Miller, N. A.; Abeyrathna, N.; Liao, Y.; Zimmerman, P. M.; Sension, R. J.
Experimental and Theoretical Characterization of Ultrafast Water-Soluble Photochromic Photoacids. J. Phys.
Chem. B 2021, 125, 4120-4131.
(15) Johns, V. K.; Wang, Z.; Li, X.; Liao, Y. Physicochemical Study of a Metastable-State Photoacid. J. Phys. Chem. A
2013, 117, 13101-13104.
96
(16) (a) Maeder, M.; Neuhold, Y.-M.; Puxty, G.; King, P. Analysis of reactions in aqueous solution at non-constant pH:
no more buffers? Phys. Chem. Chem. Phys. 2003, 5, 2836-2841; (b) In Practical Data Analysis in Chemistry;
(Eds. Maeder, M., Neuhold, Y.-M.); Elsevier: 2007; Vol. 26, p 29-100
97

Ja1c08810 Si 001

Uploaded by

Copyright:

Available Formats

Ja1c08810 Si 001

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Ja1c08810 Si 001

Uploaded by

Copyright:

Available Formats

Supporting information for

Large, tunable and reversible pH changes

S1. General information 1

S3. NMR spectra of compounds 13

S4. In situ irradiation 1H NMR spectra 29

Calculation of mole fractions of SP (𝜒𝜒𝑆𝑆𝑆𝑆 ) and MC (𝜒𝜒𝑀𝑀𝑀𝑀 ) from Kc 57

S6.11. Photoswitching of photoacids 2a–c, 3a–d 61

S7. Transient Absorption data 64

S8. Summary of key parameters for photoswitches 2a−c, 3a−d 72

S11. pH drop model 81

S12. Model describing the time dependent pH recovery in the dark 88

S13. Propagation of errors 94

2,3,3-Trimethyl-1-(3-(trimethylammonio)propyl)-3H-indol-1-ium dibromide (S2)

According to a modified literature procedure4: In a flame dried flask, (3-bromopropyl)trimethylammonium bromide

Spectral data differs from that reported in the literature.5

According to a modified literature procedure6: 5-Methoxy-2,3,3-trimethyl-3H-indole was prepared according to a

TLC: Rf = 0.29 (Hex/EtOAc = 7/3) [UV].

According to a modified literature procedure8: To a solution of (3-bromopropyl)trimethyl-ammonium dibromide

S2.2. Synthesis of merocyanines/spiropyrans

General procedure 2 (GP2): Propylammonium Merocyanine/Spiropyran Synthesis

Spectral data matches those reported in the literature.9

HRMS (ESI): calc. for C25H31NO4S: [M−Na] +: 464.1866 [2M+Na] +: 905.3840

HRMS (ESI): calc. for C26H33NO5S: [M−Na] +: 494.1971

Spectral data matches those reported in the literature.7

TLC: Rf = 0.01 (CH2Cl2/MeOH = 95/5) [UV].

HRMS (ESI): calc. for C24H31N2O+: [M] +: 363.2431

TLC: Rf = 0.01 (CH2Cl2/MeOH = 95/5) [UV].

HRMS (ESI): calc. for C28H39N2O+: [M] +: 419.3057

TLC: Rf = 0.01 (CH2Cl2/MeOH = 95/5) [UV].

HRMS (ESI): calc. for C29H41N2O2+: [M] +: 449.3163

TLC: Rf = 0.01 (CH2Cl2/MeOH = 95/5) [UV].

HRMS (ESI): calc. for C25H33N2O2+: [M] +: 393.2537

2,3,3-Trimethyl-1-(3-(trimethylammonio)propyl)-3H-indol-1-ium dibromide (S2)

For compound 2d:

Variable temperature measurements at high pH

S6.2. Setup for visible irradiation experiments (λ = 455 nm)

Descriptor Dominant wavelength / nm Radiometric power / mW

S6.3. Setup for UV irradiation experiments (λ = 265 nm)

pH-Dependent UV-visible absorption data for photoacid 2a

Determination of equilibrium constant Kc and rate constants kopen and kclose

𝐴𝐴𝑡𝑡 = 𝐴𝐴𝑒𝑒𝑒𝑒 + �𝐴𝐴0 − 𝐴𝐴𝑒𝑒𝑒𝑒 �𝑒𝑒 −𝑘𝑘𝑜𝑜𝑜𝑜𝑜𝑜,𝑒𝑒𝑒𝑒𝑡𝑡 (Eq. S2)

pH << pKadark: 𝑘𝑘𝑜𝑜𝑜𝑜𝑜𝑜 = 𝑘𝑘𝑜𝑜𝑝𝑝𝑝𝑝𝑝𝑝 (Eq. S3)

pH >> pKadark: 𝑘𝑘𝑜𝑜𝑜𝑜𝑜𝑜 = 𝑘𝑘𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜 + 𝑘𝑘𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 (Eq. S4)

Temperature based errors (kopen, kclose, Kc)

𝑘𝑘𝑜𝑜𝑜𝑜𝑜𝑜 𝛥𝛥𝛥𝛥 ‡ 1 𝛥𝛥𝛥𝛥 ‡ 𝑘𝑘𝑏𝑏

pH and temperature dependent kobs for photoacid 2a

With equilibrium constant Kc (Eq. S5):

Eq. S20 in Eq. S21 gives:

Eq. S22 in Eq. S19:

With Hendersson Hasselbach equation (Eq. S27):

The total concentration C is:

Eq. S27 in Eq. S28 gives

Mole fractions of cis-MCH (χcis-MCH ) and SP (χ𝑆𝑆𝑆𝑆ℎ𝜈𝜈 ):

MCH-2a MCH-2b MCH-2c MCH-3a MCH-3b MCH-3c MCH-3d

pH 3.2 3.2 3.7 3.2 3.7 3.8 3.7

calc. X / % 99.9 99.9 99.9 96.8 93.6 98.4 98.7

SP-2a SP-2b SP-2c SP-3a SP-3b SP-3c SP-3d

pH under hν 5.2 7.2 7.2 4.7 5.2 4.3 4.2

calc. X / % 99.8 99.9 99.9 99.9 99.9 99.5 99.6