Multicolor Luminescence Switching, and Controllable

Supplemental Information

Multicolor Luminescence Switching, and Controllable Thermally Activated Delayed Fluorescence Turn on/Turn off in Carbazole-Quinoxaline-Carbazole Triads Ramin Pashazadeh,† Piotr Pander,‡ Algirdas Lazauskas,§ Fernando B. Dias,*,‡ and Juozas V. Grazulevicius*,† †

Department of Polymer Chemistry and Technology, Kaunas University of Technology, Kaunas,

Lithuania ‡

Department of Physics, University of Durham, South Road DH1 3LE, Durham, UK

§

Institute of Materials Science, Kaunas University of Technology, Kaunas, Lithuania

Corresponding Authors * [email protected], * [email protected]

Content: 1.

General

2.

Synthetic procedure

3.

Photopysical properties

4.

Electrochemical characterization

5.

Mechanochromic properties

6.

DSC and TGA analysis

7.

Device fabrication

8.

References

9.

1

H and 13C NMR of compounds

S1

1. General 1

H and 13C NMR spectroscopy was carried out on a Bruker Avance 400 NMR spectrometer at 400 MHz and 100 MHz, respectively; δ in ppm. The residue signals of the solvents were used as internal standards. Attenuated total reflection infrared (ATR IR) spectra were recorded using a Bruker VERTEX 70 spectrometer. MS data was recorded on UPLC-MS Acquity Waters SQ Detector 2. Absorption spectra of 10-4 M solutions or films were measured with Perkin Elmer Lambda 35 spectrometer or UV-3600 double beam spectrophotometer (Shimadzu). Photoluminescence (PL) spectra of 10-5 M solutions, films, and powders were recorded using Edinburgh Instruments’ FLS980 Fluorescence Spectrometer or FluoroMax-3 fluorescence spectrometer (Jobin Yvon). Phosphorescence, prompt fluorescence (PF), and delayed fluorescence (DF) spectra and fluorescence decay curves were recorded using nanosecond gated luminescence and lifetime measurements (from 400 ps to 1 s) using either third harmonics of a high energy pulsed Nd:YAG laser emitting at 355 nm (EKSPLA) or a N2 laser emitting at 337 nm. Emission was focused onto a spectrograph and detected on a sensitive gated iCCD camera (Stanford Computer Optics) having sub-nanosecond resolution. PF/DF time resolved measurements were performed by exponentially increasing gate and integration times. Temperature-dependent experiments were conducted using a continuous flow liquid nitrogen cryostat (Janis Research) under nitrogen atmosphere, while measurements at room temperature were recorded in vacuum in the same cryostat. Power dependence data was fitted using two expressions: a∙xb (I) for power laws and a∙x (II) for linear relation. In first attempt we try to use a linear expression (II) for fitting. If the data can be fitted using a linear expression then it is described as a linear fit. If the data cannot be fitted with an expression (II) then we attempt to use power law expression (I) for fitting. In this case the power law exponent is fitted and its value is presented. Thermogravimetric analysis (TGA) was performed on a Metter TGA/SDTA851e/LF/1100 apparatus at a heating rate of 20°C/min under nitrogen atmosphere. Differential scanning calorimetry (DSC) measurements were done on a DSC Q 100 TA Instrument at a heating rate of 10°C/min under nitrogen atmosphere Cyclic voltammetry (CV) measurements were carried out with Eco Chemie Company’s AUTOLAB potentiostat “PGSTAT20” and a glassy carbon working electrode in a three electrode cell. The measurements were performed in 0.1 M nBu4NPF6 solution in anhydrous dichloromethane at room temperature under nitrogen atmosphere. OLEDs were fabricated by spin-coating / evaporation hybrid method. Hole injection layer (Heraeus Clevios HIL 1.3N), electron blocking/hole transport layer (PVKH), and emitting layer (PVK:PBD + dopant) were spin-coated, whereas electron transport layer (TPBi) and cathode (LiF/Al) were evaporated. Devices of 4x2mm pixel size were fabricated. PVK – poly(9-vinylcarbazole) (MW = 90 000, Acros Organics), PVKH – poly(9-vinylcarbazole) (MW = 1 100 000, Sigma Aldrich), PBD - 2-(biphenyl-4-yl)-5-(4-tertbutylphenyl)-1,3,4-oxadiazole (99%, Sigma Aldrich), TPBi - 2,2',2"-(1,3,5-Benzinetriyl)-tris(1-phenyl-1H-benzimidazole) (sublimed, LUMTEC), LiF (99.995%, Sigma Aldrich), and Aluminium wire (99.9995%, Alfa Aesar) were purchased from the companies indicated in parentheses. OLED devices were fabricated using pre-cleaned indium-tin-oxide (ITO) coated glass substrates after ozone plasma treatment with a sheet resistance of 20 Ω cm-2 and ITO thickness of 100 nm. Heraeus Clevios HIL 1.3N was spun-coated and annealed onto a hotplate at 200 ˚C for 3 min to give 45 nm film. Electron blocking/hole transport layer (PVKH), was spun from chloroform:chlorobenzene (95:5 v/v) (3 mg/mL) and annealed at 50 ˚C for 5 min to give 10 nm film. Emitting layer was spun from toluene solution of PVK:PBD (60:40 w/w) with total concentration of host 10 mg/mL. The dopant was dissolved in the host solution in order to obtain final 5% concentration of the emitting layer. The solution was spun onto the PVKH layer and then annealed at 50 ˚C for 5 min giving 32 nm film. All solutions were filtrated directly before application using a PVDF or PTFE S2

syringe filter with 0.45 µm pore size. All other organic and cathode layers were thermally evaporated using Kurt J. Lesker Spectros II deposition system at 10-6 mbar. All organic materials and aluminum were deposited at a rate of 1 Å s-1 The LiF layer was deposited at 0.1-0.2 Å s-1. Characterisation of OLED devices was conducted in 10 inch integrating sphere (Labsphere) connected to a Source Measure Unit. 2. Synthetic procedures 2,3-Dichloroquinoxaline (I), 3,6-dimethoxy-9H-carbazole,1 3,6-di-tert-butyl-9H-carbazole, 2 3-bromo-9Hcarbazole3 and 3,6-dibomo-9H-carbazole4 were prepared according to the procedure reported previously.

Figure S1. Synthetic procedure for preparation of 2,3-dichloroquinoxaline

Figure S2. Synthesis route for preparation of target products

3-Methoxy-9H-carbazole Na (2.2 g) was dissolved in dry MeOH (40 mL) under N2, and then 3-bromo-9H-carbazole (2.23 g, 9.1 mmol), CuI (1.73 g, 9.1 mmol) and DMF (60 mL) were stirred at 125 oC for 12 h. After cooling to room temperature, the crude product was filtered through silica gel, washed in EtOAc and evaporated. The combined organic layers were purified by column chromatography on silica gel (n-Hexane/DCM 1:2) to afford white powder (1.23 g, 69%). Mp= 146-148 oC. 1H NMR (400 MHz, CDCl3): δH = 3.83 (3H, s, MeO), 6.97 (1H, dd, 3J= 8.7 Hz, 4J= 2.5 Hz, CH), 7.11 (1H, t, 3J= 7.8 Hz, CH), 7.17 (1H, d, 3J= 8.7 Hz, CH), 7.25 (1H, d, 3J= 7.9 Hz, CH), 7.31 (1H, t, 3J= 7.9 Hz, CH), 7.46(1H, d, 4J= 2.4 Hz, CH), 7.74 (1H, br s, NH), 7.93 (1H, d, 3J= 7.7 Hz, CH) ppm. 13C NMR (100 MHz, CDCl3): δC = 56.1, 103.2, 110.7, 111.3, 115.1, 119.0, 120.2, 123.3, 123.7, 125.8, 134.4, 140.3, 153.9 ppm.

S3

2,3-Di(9H-carbazol-9-yl)quinoxaline (CzQx):

Procedure A: A mixture of carbazole (2.1 mmol) and K2CO3 (2.2 mmol) dissolved in 7 mL DMSO and stirred for 20 min, then 2,3-dichloroquinoxaline (1 mmol) was added and the reaction mixture was stirred at 90 oC for 7 h. After completion, the reaction mixture was cooled to room temperature, added 100 mL brine to the mixture and filtered. The crude compound was purified by flash column chromatography (nHexane/ethyl acetate 10:1) to give green crystals. Yield (369 mg, 78%). 1H NMR (400 MHz, CDCl3): δH = 6.94 (4H, t, 3J= 8.4 Hz, 4CH), 7.01 (4H, t, 3J= 8.4 Hz, 4CH), 7.26 (4H, t, 3J= 8.1 Hz, 4CH), 7.70 (4H, t, 3 J= 7.4 Hz, 4CH), 7.82 (2H, dd, 3J= 6.3 Hz, 4J= 3.4 Hz, 2CH), 8.2 (2H, dd, 3J= 6.3 Hz, 4J= 3.4 Hz, 2CH) ppm. 13C NMR (100 MHz, CDCl3): δC = 110.3, 119.8, 121.3, 124.3, 125.3, 128.6, 130.7, 138.7, 140.5, 141.9 ppm. MS, m/z = 461 ([M+H]+, 45%)

2,3-Bis(3,6-di-tert-butyl-9H-carbazol-9-yl)quinoxaline (tCzQx):

Procedure B: A mixture of 3,6-di-tert-butyl-9H-carbazole (720 mg, 2.6 mmol) and NaOt-Bu (260 mg, 2.7 mmol) dissolved in 7 mL of DMF and stirred at room temperature for 20 min, then 2,3-dichloroquinoxaline (250 mg, 1.26 mmol) was added and the reaction mixture was stirred at 90 oC for 7 h. After completion, the reaction mixture was cooled to room temperature, 50 mL brine were added to the mixture and it was extracted with EtOAc (3× 15 mL). The combined organic layers were dried over Na2SO4 and evaporated to afford the crude product which was purified by column chromatography on silica gel (n-Hexane/Toluene 2:1) to afford green powder. Yield (620 mg, 72%). 1H NMR (400 MHz, CDCl3): δH = 1.26 (36 H, s, 4CMe3), 6.82 (4H, dd, 3J= 8.6 Hz, 4J= 1.9 Hz, 4CH), 7.00 (4H, t, 3J= 8.6 Hz, 4CH), 7.52 (4H, d, 4J= 1.8 Hz, 4CH), 7.80 (2H, dd, 3J= 6.5 Hz, 4J= 3.0 Hz, 2CH), 8.18 (2H, dd, 3J= 6.5 Hz, 4J= 3.0 Hz, 2CH) ppm. 13C NMR (100 MHz, CDCl3): δC = 31.8, 34.5, 109.6, 115.3, 123.3, 124.3, 128.5, 130.1, 137.0, 140.3, 142.2, 143.9 ppm. MS, m/z = 684 ([M+H]+, 8)

S4

2,3-Bis(3-methoxy-9H-carbazol-9-yl)quinoxaline (MeOQx):

It was prepared from 3-methoxy-9H-carbazole by the procedure similar to the procedure A. The crude product was purified by flash column chromatography (n-Hexane/DCM 4:1) to give yellow powder. Yield (317 mg, 61%). 1H NMR (400 MHz, CDCl3): δH = 3.74 (6 H, s, 2MeO), 6.60 (2H, dd, 3J= 8.9, 3J= 2.5,2CH), 6.93-7.02 (4H, m, 4CH), 7.18-7.28 (6H, m, 6CH), 7.60 (2H, dd, 3J= 7.7 Hz, 4J= 4.1 Hz, 2CH), 7.78 (2H, dd, 3J= 6.4 Hz, 4J= 3.5 Hz, 2CH), 8.15 (2H, dd, 3J= 6.4 Hz, 4J= 3.5 Hz, 2CH) ppm. 13C NMR (100 MHz, CDCl3): δC = 55.8, 103.0, 110.4, 110.5, 111.2, 111.2, 114.4, 119.8, 121.0, 124.3, 125.1, 125.9, 128.4, 130.4, 133.4, 133.5, 139.2, 139.2, 140.3, 141.8 ppm. MS, m/z = 520 ([M]+, 36%)

2,3-Bis(3,6-dimethoxy-9H-carbazol-9-yl)quinoxaline (MeO2Qx):

Is prepared from 3,6-dimethoxy-9H-carbazole as mentioned in procedure B. The crude compound was purified by flash column chromatography (n-Hexane/DCM 2:1) to give the product. Yellow powder. Yield= (373 mg ,57 %). 1H NMR (400 MHz, CDCl3): δH = 3.75 (12 H, s, 4MeO), 6.61 (4H, dd, 3J= 8.8 Hz, 4J= 2.1 Hz, 4CH), 7.17-7.18 (6H, m, 8CH), 7.20 (2H, s, CH), 7.76 (2H, dd, 3J= 6.4 Hz, 4J= 3.4 Hz, 2CH), 8.11 (2H, dd, 3J= 6.4 Hz, 4J= 3.4 Hz, 2CH) ppm. 13C NMR (100 MHz, CDCl3): δC = 55.8, 102.8, 111.4, 114.7, 125.1, 128.3, 130.2, 134.0, 140.2, 141.8, 154.8 ppm. MS, m/z = 580 ([M]+, 100).

S5

3. Basic photopysical properties: 3.1 Steady state photoluminescence spectra of solutions and neat films

CzQx

film Hex Tol THF

tCzQx

0.8 0.6 0.4 0.2 0.0

0.8 0.6 0.4 0.2 0.0

400

450

500

550

600

650

700

400

Wavelength (nm)

500

600

700

Wavelength (nm)

Film Hex Tol THF

MeOQx

MeO2Qx

1.0 Normalized PL intensity


film Hex Tol THF


Normalized PL intensity

1.0

0.8 0.6 0.4 0.2 0.0

Film Hex Tol THF

0.8 0.6 0.4 0.2 0.0

400 450 500 550 600 650 700

400 450 500 550 600 650 700 750

Wavelength (nm)

Wavelength (nm)

Figure S3. Photoluminescence spectra of compounds in Hexane (Hex), Toluene (Tol), THF, thin film (prepared from a chloroform solution).

Note the fluorescence (Figure S3) in all solvents shows Gaussian shape, indicative for CT emission. The CT emission shifts with solvent polarity. Local emission is visible below 450 nm. As the PLQY of CT emission decrease with solvent polarity, the ratio between CT and local emission is more favorable to CT in solvents of lower polarity. CzQx tCzQx MeOQx MeO2Qx

1.0

0.8

0.8

0.6

0.6

0.4

0.4

0.2

0.2

0.0 200

PL intensity (a.u.)

Absorbance (a.u.)

1.0

0.0 300

400

500

600

700

Wavelength (nm)

Figure S4. Normalized UV-Vis absorption and PL spectra of investigated molecules in thin films.

S6

3.2 Time-resolved photoluminescence study of dispersions in Zeonex (1% w/w) 10

8

10

7

10

6

10

5

10

4

10

3

10

2

10

1

10

0

10

b 80 K 295 K

Photoluminescence intensity (a.u.)


a

10

8

10

7

10

6

10

5

10

4

10

3

10

2

10

1

10

0

10n

1µ 10µ 100µ Time delay (s)

100n

1m

10

10m 100m

1

DF = 0.25 ± 0.03 ms (64 %) 2

-1

1n

 PF = 1.82 ± 0.06 ns

DF = 4.8 ± 0.7 ms (36 %)

 PH = 111 ± 29 ms

Decay 295 K Decay 80 K Fit 295 K Fit 80 K

-1

1n

10n

100n


1m

10m 100m

d 0.7 ns 295 K 5 ms 295 K 0.7 ns 80 K 71 ms 80 K

Normalized intensity (a.u.)

1.0

10

8

10

7

10

6

Linear fit 2 r = 0.995

0.8 S1 = 2.87 eV T1 = 2.59 eV EST = 0.28 eV

0.6

0.4

DF intensity (a.u.)

c

0.2

0.0 400

450

500 550 Wavelength (nm)

600

650

1

10 Laser pulse energy ( J)

100

Figure S5. Time-resolved analysis of CzQx dispersed in Zeonex (1% w/w): a) photoluminescence decay at 295 and 80 K; b) photoluminescence decay with fits and decay time constants shown; c) prompt and delayed fluorescence and phosphorescence spectra; d) power dependence of delayed fluorescence at 295 K. Note the power dependence denoted as linear is a power law = 1.

10

8

10

7

10

6

10

5

10

4

10

3

10

2

10

1

10

0

b Photoluminescence intensity (a.u.)


a

295 K 80 K

-1

10

8

10

7

10

6

10

5

10

4

10

3

10

2

10

1

10

0

PF = 3.21 ± 0.07 ns

1

 DF = 0.10 ± 0.03 ms (70 %) 2

 DF = 0.9 ± 0.2 ms (30 %)


PH = 165 ± 14 ms

-1

1n

10n

100n

c


1m

10

10m 100m

d 10 1.0

1n

0.8

S1 = 2.69 eV T1 = 2.55 eV

0.6

EST = 0.14 eV

0.4

10n

100n


1m

10m 100m

9


0.7 ns 295 K 100 us 295 K 0.7 ns 80 K 71 ms 80 K

DF intensity (a.u.)


10

10

8

10

7

10

6

0.2

0.0 450

500


650

700

1

10 Laser pulse energy (J)

100

Figure S6. Time-resolved analysis of tCzQx in Zeonex (1% w/w): a) photoluminescence decay at 295 and 80 K; b) photoluminescence decay with fits and decay time constants shown; c) prompt and delayed fluorescence and phosphorescence spectra; d) power dependence of delayed fluorescence at 295 K. Note the power dependence denoted as linear is a power law = 1.

S7

a

b

8

10

8

10

7



6

10

5

10

4

10

3

10

2

10

295 K 80 K

1

10

0

10

6

10

1

 DF = 36 ± 7 s (84 %) 2

 DF = 302 ± 25 s (16 %)

5

10

4

10

PH = 130 ± 12 ms

3

10

2

10


1

10

0

10 10

PF = 3.98 ± 0.02 ns

7

10

10

-1

10 1n

10n

100n


1m

-1

10m 100m

c

1n

10n

100n


1m

10m 100m

d 10

9

0.7 ns 295 K 100 us 295 K 0.7 ns 80 K 71 ms 80 K


1.0


0.8 8

10

DF intensity (a.u.)

S1 = 2.63 eV T1 = 2.51 eV

0.6

EST = 0.12 eV

0.4

7

10

0.2

6

0.0 450

500


650

10

700

1


100

Figure S7. Time-resolved analysis of MeOQx in Zeonex (1% w/w): a) photoluminescence decay at 295 and 80 K; b) photoluminescence decay with fits and decay time constants shown; c) prompt and delayed fluorescence and phosphorescence spectra; d) power dependence of delayed fluorescence at 295 K. Note the power dependence denoted as linear is a power law = 1.

a

b 10

8

8

10

7


5

10

4

10

3

10

2

10

295 K 80 K

1

10

0

10

6

10

1

DF = 31 ± 6 s (67 %) 2

DF = 144 ± 54 s (33 %)

5

10

4

10

3

10

PH = 74 ± 5 ms

2

10


1

10

0

10

10

-1

-1

10

1n

10n

100n

c


1m

10m 100m

10

1n

10n

100n

d 1.0

0.8


1m

10m 100m


0.7 ns 295 K 100 us 295 K 0.7 ns 80 K 71 ms 80 K

8

10 S1 = 2.58 eV T1 = 2.44 eV

0.6

E ST = 0.14 eV

0.4

DF intensity (a.u.)


6

10


PF = 3.98 ± 0.02 ns

7

10

7

10

0.2

6

0.0

10 450

500


650

700

1


100

Figure S8. Time-resolved analysis of MeO2Qx in Zeonex (1% w/w): a) photoluminescence decay at 295 and 80 K; b) photoluminescence decay with fits and decay time constants shown; c) prompt and delayed fluorescence and phosphorescence spectra; d) power dependence of delayed fluorescence at 295 K. Note the power dependence denoted as linear is a power law = 1.

S8


1.0

Phosphorescence

Quinoxaline CzQx MeOQx MeO2Qx tCzQx

0.8

0.6

0.4

0.2

0.0 450

500


650

700

Figure S9. Phosphorescence spectra of quinoxaline and all reported molecules combined in Zeonex (1% w/w) at 80 K.

a

b

5 ms 80 K

1.0

1.0

5 ms 80 K



T 1 = 3.00 eV

0.8

0.6

0.4

0.8 T1 = 2.89 eV

0.6

0.4

0.2

0.2

0.0

0.0 400

450 500 550 Wavelength (nm)

400

600


550

600

Figure S10. Phosphorescence spectrum of (a) 3-methoxy-9H-carbazole and (b) 3,6-dimethoxy-9H-carbazole in Zeonex (1% w/w).

S9

3.3 Time-resolved photoluminescence study of PVK:PBD (5% w/w) Table S1. Singlet & triplet energy in PVK:PBD Compound CzQx tCzQx MeOQx MeO2Qx

S1, eV 2.87 2.69 2.63 2.58

T1, eV 2.59 2.55 2.51 2.44

 EST, eV 0.28 0.14 0.12 0.14

0

10

295 K


-1

10

-2

10

-3

10

-4

10

-5

10

CzQx tCzQx MeOQx MeO2Qx

-6

10

-7

10

1n

10n

100n

1µ 10µ Time delay (s)

100µ

1m

Figure S11. Photoluminescence decays of the studied molecules dispersed in PVK:PBD (5% w/w).

In principle the molecules behave very similar in the PVK:PBD mixed host as in Zeonex (Figures S11S15). The results show that TADF is also present in OLED host, so explains high EQE of OLED devices produced. Indeed contribution of TADF is very similar in tCzQx, MeOQx and MeO2Qx, but CzQx has significantly less delayed fluorescence (Figure S12). It is worth to note the delayed fluorescence lifetime decreased by approximately an order of magnitude compared to Zeonex. Moreover, the singlet (CT) energy is decreased, but also triplet energy decreases by very similar value. This in fact causes the singlet-triplet gap to be virtually identical in both PVK:PBD and Zeonex. The phosphorescence spectrum is not only redshifted in PVK:PBD, but their vibronic structure becomes less evident and the decay constants decrease. This suggests a relatively strong interaction of the acceptor with host molecules. It is worth to note that phosphorescence lifetime of tCzQx is the longest among all presented molecules, similarly in Zeonex. As the phosphorescence spectrum and triplet energy are different for each molecule in PVK:PBD mixed host, therefore they reflect respective dopant’s triplet rather than the triplet of the host (PBD). What is also interesting in CzQx-doped PVK:PBD blend (Figure S12) is a relatively long-lived, biexponential prompt component. As CzQx shows the lowest solubility in toluene, this might be due to formation of aggregates in microscale, thus shorter prompt fluorescence component should be attributed to excitonic emission, the latter to aggregate emission. Photophysics therefore explains why CzQx-based device is significantly less efficient than others.

S10

a

10

8

10

7

10

6

10

5

b 10

8 1

PF = 3.8 ± 0.1 ns (91 %)

10

4

10

3

10

2

10

1



2

295 K 80 K

1n

10n

100n

1µ

10µ

100µ

1m

10m 100m

10

7

10

6

10

5

PF = 23 ± 3 ns (9 %)

1

DF = 165 ± 40 s (49 %) 2

10

4

10

3

10

2

10

1

DF = 14 ± 4 s (51 %)

PH = 85 ± 6 ms


1n

10n

100n

Time delay (s)

c 1.0

S1 = 2.73 eV

0.7 ns 295 K 100 us 295 K 0.7 ns 80 K 71 ms 80 K

T1 = 2.53 eV

1m

10m 100m

d 10

8

0.8 DF intensity (a.u.)


EST = 0.20 eV


0.6

0.4


10

7

10

6

0.2

0.0 400

450


600

1

650


100

Figure S12. Time-resolved analysis of CzQx in PVK:PBD 60:40 (5% w/w): a) photoluminescence decay at 295 and 80 K; b) photoluminescence decay with fits and decay time constants shown; c) prompt and delayed fluorescence and phosphorescence spectra; d) power dependence of delayed fluorescence at 295 K. Note the power dependence denoted as linear is a power law = 1.

b 10

a 10

8

8



PF = 4.73 ± 0.05 ns 7

10

6

10

5

10

4

10

3

10

295 K 80 K

2

10

1

1

DF = 120 ± 20 s (49 %) 5

2

DF = 10 ± 2 s (51 %)

10

4

10

3

10

PH = 132 ± 30 ms


2

1

1n

10n

100n


1m

c

10m 100m

0.7 ns 295 K 100 us 295 K 0.7 ns 80 K 71 ms 80 K

S1 = 2.66 eV T1 = 2.50 eV EST = 0.16 eV

10

1n

10n

0.4


1m

10m 100m

8

10


0.8

0.6

100n

d

DF intensity (a.u.)


6

10

10

10

1.0

7

10

7

10

0.2

6

10

0.0 400

450


600

650

1


100

Figure S13. Time-resolved analysis of tCzQx in PVK:PBD 60:40 (5% w/w): a) photoluminescence decay at 295 and 80 K; b) photoluminescence decay with fits and decay time constants shown; c) prompt and delayed fluorescence and phosphorescence spectra; d) power dependence of delayed fluorescence at 295 K. Note the power dependence denoted as linear is a power law = 1.

S11

10

8

10

7

10

6

10

5

10

4

10

3

10

2

10

1

b Photoluminescence intensity (a.u.)


a

295 K 80 K

1n

10n

100n

1µ

10µ

100µ

1m

10

8

10

7

10

6

10

5

10

4

10

3

10

2

10

1

PF = 4.55 ± 0.05 ns

1

DF = 3.2 ± 0.3 s (55 %) 2

DF = 37 ± 4 s (45 %)


1n

10m 100m

10n

PH = 39 ± 2 ms


100n

Time delay (s)

c

d S1 = 2.62 eV

1.0

10

8

10

7

1m

10m 100m

0.7 ns 295 K 20 us 295 K 0.7 ns 80 K 71 ms 80 K

T 1 = 2.47 eV  EST = 0.15 eV

0.8

DF intensity (a.u.)


10

9

0.6

0.4

0.2


0.0 450

500


650

700

1

10 Laser pulse energy ( J)

100

Figure S14. Time-resolved analysis of MeOQx in PVK:PBD 60:40 (5% w/w): a) photoluminescence decay at 295 and 80 K; b) photoluminescence decay with fits and decay time constants shown; c) prompt and delayed fluorescence and phosphorescence spectra; d) power dependence of delayed fluorescence at 295 K. Note the power dependence denoted as linear is a power law = 1.

10

8

10

7

10

6

10

b 10 Photoluminescence intensity (a.u.)


a

5

10

4

10

3

10

2

10

1

295 K 80 K

1n

10n

100n

c


10

7

PF = 4.8 ± 0.1 ns

6

1

10

5

10

4

10

3

10

2

10

1

DF = 7.4 ± 0.7 s (75 %) 2

DF = 43 ± 10 s (25 %)

10m 100m


1n

10n

100n

PH = 42 ± 3 ms


1m

10m 100m

d 10

9

1.0

S1 = 2.53 eV

0.7 ns 295 K 20 us 295 K 0.7 ns 80 K 71 ms 80 K

T1 = 2.39 eV EST = 0.14 eV

0.8

DF intensity (a.u.)


1m

10

8

0.6

0.4


8

10

7

10

0.2

0.0 450

500


650

700

1


100

Figure S15. Time-resolved analysis of MeO2Qx in PVK:PBD 60:40 (5% w/w): a) photoluminescence decay at 295 and 80 K; b) photoluminescence decay with fits and decay time constants shown; c) prompt and delayed fluorescence and phosphorescence spectra; d) power dependence of delayed fluorescence at 295 K. Note the power dependence denoted as linear is a power law = 1.

S12

4. Electrochemical characterization: 20.0µ

CzQx tCzQx MeOx MeO2Qx

15.0µ

Current

10.0µ 5.0µ 0.0 -5.0µ -10.0µ

-1.8

-1.5

-1.2

0.9

1.2

1.5

1.8

Potential

Figure S16. Oxidation and reduction potential of compounds in DCM recorded with cyclic voltammetry. EHOMO = - (E1/01/2 (vs. Fc+/Fc) +4.8), ELUMO = - (E0/11/2(vs. Fc+/Fc) + 4.8)

5. Mechanochromic Properties:

5.1 Steady state photoluminescence spectra and photographs

Figure S17. Photoluminescence of (a) CzQx, (b) tCzQx, (c) MeOQx and (d) MeO2Qx forms obtained through various external stimuli (The images taken under excitation light at 365 nm).

S13

Table S2. PL properties of compounds in solid films and powders in air. Heating of MeO2Qx-d and CzQx-d did not result in a change of photoluminescence color and no –dh form was obtained. Sample

solid states

initial(i)

ground(g)

heated(h)

fumed(f)

melted(m)

tCzQx

PL (nm) PLQY (%) PL (nm) PLQY (%) PL (nm) PLQY (%) PL (nm) PLQY (%)

482 9 516 10 534 6 504 4

527 12 543 7 586 4 510 5

477 9 511 10 559 7 478 11

479 9 514 9 523 5 478 6

544 9 582 7 618 2 524 10

MeOQx MeO2Qx CzQx

0.6

0.4

0.2

0.0

tCzQx-i tCzQx-g tCzQx-h tCzQx-f tCzQx-d tCzQx-dh tCzQx-m

0.8 0.6 0.4 0.2 0.0

400

500

600

400

700

500

MeOQx-i MeOQx-g MeOQx-f MeOQx-h MeOQx-m MeOQx-dh MeOQx-d

1.0 0.8 0.6 0.4 0.2 0.0 400

d)

700

MeO2Qx-i MeO2Qx-g MeO2Qx-f MeO2Qx-h MeO2Qx-m MeO2Qx-d

1.0


c)

600

Wavelength (nm)

Wavelength (nm)


482 6 519 11

1.0



0.8

534 11 592 10 598 6 519 7

b)

CzQx-i CzQx-g CzQx-h CzQx-f CzQx-m CzQx-d

1.0

heated film(dh)

neat film(d)

0.8 0.6 0.4 0.2 0.0

500

600

700

400

Wavelength (nm)

450

500

550

600

650

Wavelength (nm)

Figure S18. PL spectra of molecules in different forms.

S14

700

750

5.2 Time-resolved photoluminescence study 1.2

1.2 tCzQx-dh S1 = 2.80 eV

PF CzQx-d PH CzQx-d 15 ms 80 K

1.0

1.0

T1 = 2.38 eV

0.8

T1 = 2.40 eV EST = 0.26 eV

0.6

0.4



EST = 0.42 eV

CzQx-d S1 = 2.68 eV

0.2

T1 = 2.41 eV EST = 0.16 eV

0.6

0.4

0.0

450

500


650

700

400

450


650

700

1.2

1.2

T1 = 2.37 eV EST = 0.30 eV

PF MeOQx-d PF MeOQx-dh PH MeOQx-d 15 ms 80 K PH MeOQx-dh 60 ms 80 K MeOQx-d S1 = 2.50 eV

0.8

T1 = 2.34 eV EST = 0.16 eV

0.6

0.4

1.0

MeO2Qx-d S1 = 2.41 eV

PF MeO2Qx-d PH MeO2Qx-d 15 ms 80 K

T1 = 2.35 eV


MeOQx-dh S 1 = 2.67 eV


tCzQx-d S1 = 2.57 eV

0.8

0.2

0.0

1.0

PF tCzQx-d PF tCzQx-dh PH tCzQx-d 15 ms 80 K PH tCzQx-dh 15 ms 80 K

EST = 0.06 eV

0.8

0.6

0.4

0.2

0.2

0.0

0.0 450

500


650

700

450

500


650

700

Figure S19. Prompt fluorescence and phosphorescence spectra of molecules in films (d and dh forms).

It is worth to note the phosphorescence spectrum is affected by CT state if S1 and T1 are close. In MeOQxd the PH spectrum is distorted by mixing with CT state as it lies nearby, similar effect explains the distorted spectrum of MeO2Qx-d phosphorescence. The fact that phosphorescence spectra of tCzQx-d and tCzQxdh are nearly identical indicates there is no signs of aggregation, nor excimer formation. Even in neat film the molecule behaves like isolated system, but D-A dihedral angle is different between the forms (Figure S19). All powder luminescent forms of CzQx are mixtures (Figures S20, S21). It is evident from the photophysics that CzQx-g contains majority of amorphous phase, while CzQx-h and CzQx-f, which are in fact very similar, contain mostly crystalline phase. As all powders exhibit emissions from at least two distinct forms of CzQx therefore it is difficult to precisely determine their properties. CzQx-m is in principle very similar to CzQx-g, but CzQx-d is different. This shows that a film deposited from a chloroform solution and obtained via melt-casting show different packing. All powders and films show delayed fluorescence which is a mixture of the crystal/amorphous phase emission with a dominating, red-shifted emission at late times (i.e. > 10 µs) which is attributed to a relaxed CT-state emission of the molecule or excimer emission. As all the discussed forms show a mixed contribution of emissive forms, further analysis is inconclusive. It can be stated that most likely all forms show TADF emission originating from the relaxed CT state/excimer, which is nearly identical in all forms, with contribution of delayed fluorescence from other forms. Note S15

CzQx-d shows weak signal of delayed fluorescence of crystal forms (blue shifted in respect to prompt) which is consistent with XRD results. 0

10

-1



CzQx-i

CzQx-i CzQx-g CzQx-h CzQx-f CzQx-m CzQx-d

CzQx-g CzQx-h CzQx-f CzQx-m CzQx-d

10

-2

10

-3

10

-4

10

-5

10

-6

10

-7

1n

10n

100n 1µ 10µ Time delay (s)

100µ

1m

10

1n


10n

100µ

1m

Figure S20. Photoluminescence decays of different emissive forms of CzQx: (left) stacked, (right) normalized.

a

1 s

b

20 s

10

8

8

10

10

8

CzQx-g

CzQx-m

CzQx-h


CzQx-g

1 s

10

7

10

6

DF intensity (a.u.)

40 s

DF intensity (a.u.)

1 s

DF intensity (a.u.)

CzQx-i

7

10

10

7

10

6

50 s CzQx-h Power law 0.83 2 r = 0.998 1 10 100 Laser pulse energy (J)

1 s

10 s

CzQx-d

DF intensity (a.u.)

10

CzQx-m

10

CzQx-d

CzQx-f

1 s 20 s

1 10 100 Laser pulse energy (J) 10

CzQx-f

10

8

Power law 0.76 2 r = 0.996

9

10

7

Power law 1.10 2 r = 0.93

450


650

10

1 10 100 Laser pulse energy (J) 8 10 CzQx-i

DF intensity (a.u.)

50 s

DF intensity (a.u.)

1 s

Power law 0.83 2 r = 0.999

Power law 0.77 2 r = 0.996 6

10

6

7

10

6

Power law 0.91 2 r = 0.997 8

10

1 10 100 Laser pulse energy (J)

10


Power law 0.89 2 r = 0.998 1 10 100 Laser pulse energy (J)

Figure S21. (a) Time-resolved spectra of luminescent forms of CzQx. Solid line: prompt fluorescence; dotted and dashed line indicate delayed fluorescence. Delay at which the delayed fluorescence spectrum was recorded is indicated in the figure next to the respective spectrum. (b) Power dependence of delayed fluorescence recorded in different photoluminescent forms of CzQx.

S16

0

10

tCzQx-g tCzQx-h tCzQx-f tCzQx-m tCzQx-d tCzQx-dh

1n

10n

10

-1

10

-2

10

-3

10

-4

10

-5

10

-6

10

-7



tCzQx-i


100µ

tCzQx-i tCzQx-g tCzQx-h tCzQx-f tCzQx-m tCzQx-d tCzQx-dh

1n

1m

10n


100µ

1m

Figure S22. Photoluminescence decays of different emissive forms of tCzQx: (left) stacked, (right) normalized.

10

9

tCzQx-g

tCzQx-i

DF intensity (a.u.)

10 s DF/PF = 1.5

10

16 s DF/PF = 2.7 TADF

tCzQx-g

10

tCzQx-f 8


10

8


tCzQx-m 8

10

DF intensity (a.u.)

b

TADF + TTA

DF intensity (a.u.)

a

7

10

7

10


Power law 1.18 2 r = 0.999

7

TADF

tCzQx-f

10 s DF/PF = 1.8 tCzQx-m

10

10

tCzQx-h

tCzQx--m 10


10

7

10

6


8

8


Power law 1.36 2 r = 0.995

6

10 1 10 100 Laser pulse energy (J)

tCzQx-dh Power law 1.12 2 r = 0.999 9

10

7

10

16 s DF/PF = 4.0

1

10 100 Laser pulse energy (J)

8

10

Power law 1.31 2 r = 0.998 10 100 Laser pulse energy (J)

tCzQx-i

TADF

tCzQx-d

5 s

tCzQx-dh

450 500 550 600 650 700 Wavelength (nm)

10

8

10

7

DF intensity (a.u.)

DF/PF = 0.9 TTA


DF intensity (a.u.)

DF/PF = 0.8 16 s

1

10 10

TTA


DF intensity (a.u.)

TTA

DF intensity (a.u.)

N orm alized intensity (a.u.)

6

1

10 s DF/PF = 0.3 tCzQx-h


1


Figure S23. (a) Time-resolved spectra of luminescent forms of tCzQx. Solid line: prompt fluorescence; dotted and dashed line indicate delayed fluorescence. Delay at which the delayed fluorescence spectrum was recorded is indicated in the figure next to the respective spectrum. Delayed fluorescence to prompt fluorescence ratio (DF/PF) is shown next to each spectrum. Note visibly larger DF contribution in amorphous forms. (b) Power dependence of delayed fluorescence recorded in different photoluminescent forms of tCzQx.

S17

tCzQx shows a definite distinction between crystal and amorphous forms (Figures S22, S23). In fact tCzQx-g, tCzQx-d, and tCzQx-m represent nearly identical amorphous phase, while tCzQx-f, tCzQx-h, and tCzQx-dh are of the same crystalline character. tCzQx-i is therefore a mixture of amorphous and crystalline phase with crystalline being dominating. In all amorphous forms of tCzQx a strong TADF can be observed, while crystalline phase shows very little delayed fluorescence of triplet-triplet annihilation (TTA) origin (supralinear power dependence regime). MeOQx behaves similarly to other materials (Figures S24, S25), however no significant change in delayed fluorescence is observed between the forms (all show TADF with comparable intensity). This shows the change of CT energy is less pronounced, thus S-T gap facilitates TADF emission in all cases. This is most likely due to the stronger electron-donating properties of the donor (compared to tCzQx), so the CT is less affected by the D-A dihedral angle than in the case of tCzQx. MeOQx-i, MeOQx-h, and MeOQx-f appear to be crystal, with prompt and delayed fluorescence spectra matching perfectly. On the other hand MeOQx-g is amorphous, also showing TADF. MeOQx-m and MeOQx-d show different, more red-shifted emission than MeOQx-g. This shows that if amorphous phase is produced directly from liquid it contains more free volume than when the crystalline powder is ground. MeOQx-m and MeOQx-d show lower CT energy than MeOQx-g which is due to a larger D-A dihedral angle (more free volume and more freedom of D-A orientation). MeOQx-d when left on hotplate, crystallizes to give MeOQx-dh which is similar to other crystal forms of MeOQx.

10

MeOQx-g MeOQx-h MeOQx-f MeOQx-m MeOQx-d MeOQx-dh

1n

10n


10

-1

10

-2

10

-3

10

-4

10

-5

10

-6

10

-7



MeOQx-i

100µ

1m

0

MeOQx-i MeOQx-g MeOQx-h MeOQx-f MeOQx-m MeOQx-d MeOQx-dh

1n

10n


100µ

1m

Figure S24. Photoluminescence decays of different emissive forms of MeOQx: (left) stacked, (right) normalized.

S18

b

MeOQx-d

2 s

10

MeOQx-g

MeOQx-m

8

MeOQx-h 10

Power law 0.88 2 r = 0.999

Power law 0.83 2 r = 0.998 7

10

7

DF intensity (a.u.)

MeOQx-i

DF intensity (a.u.)

10

DF intensity (a.u.)

a

8

Power law 0.69 2 r = 0.999

10

7

1 s 6


10

1 10 100 Laser pulse energy (J) MeOQx-dh

MeOQx-f

8

Power law 0.75 2 r = 0.998

DF intensity (a.u.)

1 s MeOQx-m 2 s

10

MeOQx-g 10

MeOQx-f


8

10

Power law 0.63 2 r = 0.999

10

DF intensity (a.u.)

MeOQx-h 2 s DF intensity (a.u.)


1

10

10

7

Linear fit 2 r = 0.998 9

7

10 1 10 100 Laser pulse energy ( J)

MeOQx-d



MeOQx-i

2 s

1 s

450 500 550 600 650 700 Wavelength (nm)

10

7

DF intensity (a.u.)

MeOQx-dh

10

8

Power law 0.76 2 r = 0.997

1


Figure S25. (a) Time-resolved spectra of luminescent forms of MeOQx. Solid line: prompt fluorescence; dotted and dashed line indicate delayed fluorescence. Delay at which the delayed fluorescence spectrum was recorded is indicated in the figure next to the respective spectrum. (b) Power dependence of delayed fluorescence recorded in different photoluminescent forms of MeOQx.

MeO2Qx-g, MeO2Qx-d, and MeO2Qx-m are amorphous, while MeO2Qx-i, MeO2Qx-f, and MeO2Qx-h are crystalline. Like MeOQx, also MeO2Qx-d and MeO2Qx-m show more relaxed CT emission than MeO2Qx-g. MeO2Qx-i is much more similar to MeO2Qx-f, which can be explained, as both powders were obtained under influence of a solvent (Figures S26, S27). Interestingly, MeO2Qx-h, while also being crystalline, shows completely different emission than the other crystalline forms. Please also note that X-ray difractogram of MeO2Qx-h shows significantly different patterns than the other two crystalline powders. This suggests the molecule can produce two distinct crystal structures with different D-A dihedral angles. It is particularly interesting that MeO2Qx-h shows a relatively narrow emission, compared to MeO2Qx-i or MeO2Qx-f, suggesting a very rigid structure. In fact the D-A dihedral angle can be fixed in MeO2Qx-h, giving freedom to only minor oscillations of the donor unit, which makes the emission spectrum very narrow, but still of CT nature. In MeO2Qx-i and MeO2Qx-f the donor has a little more freedom to oscillate, which results in broader (but blue-shifted) emission spectrum.

S19

0

10

MeO2Qx-i MeO2Qx-g MeO2Qx-h MeO2Qx-f MeO2Qx-m MeO2Qx-d

-1



MeO2Qx-i MeO2Qx-g MeO2Qx-h MeO2Qx-f MeO2Qx-m MeO2Qx-d

10

-2

10

-3

10

-4

10

-5

10

-6

10

-7

10 1n

10n


100µ

1m

1n


10n

100µ

1m

Figure S26. Photoluminescence decays of different emissive forms of MeO2Qx: (left) stacked, (right) normalized.

b 10 DF intensity (a.u.)

MeO2Qx-i

5 s

10

MeO2Qx-g

7

MeO2Qx-g 5 s

10

8

10

7

10

6

10


MeO2Qx-f

2 s MeO2Qx-m

2 s MeO2Qx-d

450 500 550 600 650 700 Wavelength (nm)

10

6

7

10

6


8

MeO2Qx-m

10

10


MeO2Qx-h

10

7

10 Power law 0.83 2 r = 0.995 10 100 Laser pulse energy (J)

MeO2Qx-f 10

8

10

7

8

DF intensity (a.u.)

1 s

8

Power law 0.82 2 r = 0.998

DF intensity (a.u.)

MeO2Qx-h

MeO2Qx-i 10

Power law 0.98 2 r = 0.999

Power law 0.85 2 r = 0.999

DF intensity (a.u.)


MeO2Qx-d

8

DF intensity (a.u.)

2 s

DF intensity (a.u.)

a

7

Power law 0.75 2 r = 0.999 1 10 100 Laser pulse energy (J)

Power law 0.81 2 r = 0.997 10

6


Figure S27. (a) Time-resolved spectra of luminescent forms of MeO2Qx. Solid line: prompt fluorescence; dotted and dashed line indicate delayed fluorescence. Delay at which the delayed fluorescence spectrum was recorded is indicated in the figure next to the respective spectrum. (b) Power dependence of delayed fluorescence recorded in different photoluminescent forms of MeO2Qx.

S20

5.3. PXRD in different solid states Powder XRD pattern of all forms are depicted in Figure S28. The initial powders (-i) exhibited diffraction peaks, demonstrating well-ordered and crystalline structures. However, after grinding with spatula the signals are either weaker or disappear, rendering crystallinity of starting forms is disrupted by external stimulus. Upon either heating (-h) or fuming (f) of ground powders (-g) new diffraction peaks arise, manifesting the ground solids have reassembled from amorphous form into a new crystalline lattice, corresponding to a blue-shifted emission in relation to the ground form (-g). Melt- and drop-cast films are amorphous. The XRD patterns of MeO2Qx-h received by heating of MeO2Qx-g, are different from other crystalline forms, suggesting a second crystalline form is obtained by heating. MeO2Qx-h thus shows different crystal structure to the crystal form obtained by the contact with a solvent (i.e. MeO2Qx-f).

a

b

tC zQ x-i tC zQ x-g tC zQ x-h tC zQ x-f tC zQ x-m tC zQ x-d tC zQ x-d h

Intensity

In te nsity

C zQ x- i C zQ x- g C zQ x- h C zQ x- f C zQ x- d C zQ x- m

10

20

30

40

10

2 theta ( degree)

c

20

30

40

2 theta (degree)

d Me2O Qx- d Me2O Qx- i Me2O Qx- g Me2O Qx- h Me2O Qx- f Me2O Qx- m

Inten sity

Intens ity

MeO Qx -i MeO Qx -g MeO Qx -h MeO Qx -f MeO Qx -m MeO Qx -d

10

20 30 2 theta (degr ee)

40

10

20

30

40

2 th et a (deg ree )

Figure S28. Powder XRD patterns of different forms of reported molecules: a) CzQx, b) tCzQx, c) MeOQx, and d) MeO2Qx.

5.4. Summary and description of photophysical results Photophysical properties of all compounds in solid state have altered by mechanical stimulation and exhibited MCL behavior, moreover MeOQx and MeO2Qx represent multicolor altering in response to external forces (Figure S17). In all cases grinding with spatula perturbs solid forms, leading to red-shifted and broadened photoluminescence spectrum, which reflects conversion from crystal to the amorphous state. On the contrary, fuming by CH2Cl2 for 5 min or heating at 180 oC for 10 min, resulted in blue-shift, narrower photoluminescence spectra and typically higher quantum yields (except for tCzQx where it is opposite due S21

to TADF contribution in amorphous forms). These observations imply well-ordered and rigidified molecular structure upon heating and fuming. Consequently, variation in packing orientation and dihedral angle between donor-acceptor lead to different emission spectra due to change in CT energy. Furthermore, D-A dihedral angle alteration switches between charge transfer (CT) and/or locally exited (LE) states. While CT always causes broadened spectra, LE state provides narrower spectra. Moreover, melt-casting also causes significant luminescence bathochromic shift. These parameters reflect the formation of excimer forms after melting or a further relaxation of the CT due to a larger free volume. The drop-coast films in comparison to solid forms demonstrate red-shifted emission in relation to the crystal forms. The films of tCzQx and MeOQx can be transformed by heating, giving blue-shifted photoluminescence and made crystalline (PL spectra are depicted in Table S2). The pristine film of CzQx (CzQx-i) shows bluish-green emission with maximum located at 504 nm with PL = 3%. After mechanically grinding using a spatula, the green emission red-shifts giving emission at 510 nm and PL = 5%. These changes are attributed to disarranging the molecular assembly, which implies decreasing the crystallinity. After annealing at 180 oC or exposure the ground sample to DCM vapor for 5 min, the emission blue-shifts to sky blue with maximum at 478 nm and 479 nm, and PL = 11% and 6%, respectively. The hypochromic shift and higher quantum yields imply that intermolecular interactions between adjacent molecules are suppressed. CzQx-h has narrower photoluminescence spectrum and exhibits higher PL than CzQx-f which shows the CzQx-h has higher degree of crystallinity, while CzQxf still contains traces of amorphous form. The drop-cast film of CzQx exhibits the emission at 519 nm, and PL = 6%. (Figure S28 and Table S2) As-prepared form tCzQx-i presents sky blue emission band with maximum at 477 nm and upon grinding (tCzQx-g) the fluorescence maximum shifted by 50 nm towards red with a peak maximum of 527 nm. This was accompanied by the PL increase from 9% to 12%. Heating and fuming of tCzQx-g powder produces powder of sky-blue fluorescence, tCzQx-f and tCzQx-h, the emission bands show maxima of 477 nm and 479 nm, respectively. Photoluminescence of tCzQx-i, tCzQx-h, and tCzQx-f, is more blue-shifted than in hexane (488 nm). This may suggest the singlet state has more localized character than the emission in non-polar solvent. The red shift of photoluminescence after grinding (tCzQx-g) implies transformation of the excited state from the LE (or HLCT) state into an intramolecular charge transfer (ICT) state, upon relaxation of the D-A dihedral angle. In amorphous form there is more free space which allows the donor units to adapt a conformation promoting charge transfer. Dropcast film (tCzQx-d) shows similar emission to the ground powder (tCzQxg) with maximum at 534 nm and a similar PL of 11%. MeOQx-g red-shifted by 27 nm (543 nm) compared to MeOQx-i (516 nm). Upon grinding the PL and full width at half maximum (FWHM) increased, corresponding to the CT state emission. Initial state was recovered by thermal annealing (MeOQx-h) with PL = 10%. Fuming of MeOQx-g with DCM vapor turned its photoluminescence to green with maximum at 514 nm and PL = 9%. Notably, fluorescence spectra of drop-cast films emit orange photolminescence (λPL = 576 nm) with PL = 10%. MeO2Qx-i has green photoluminescence color (λPL = 534 nm, PL = 6 %). The photoluminescence color of MeO2Qx-i undergoes a change to orange-red through grinding (MeO2Qx-g), with λPL = 586 nm, and PL = 4%. Exposure of MeO2Qx-g to DCM vapor leads to MeO2Qx-f with λPL = 523 nm. Notably, thermal annealing causes slight blue-shift and conversion to a yellow emissive powder (MeO2Qx-h) λPL = 559 nm, accompanied with the increase of PL to 7 %. Thus the MeO2Qx-h does not transform into MeO2Qx-f upon fuming, but can be recrystallized to obtain MeO2Qx-i. S22

Remarkably bathochromic shift occurrs upon melt-casting. CzQx-m and tCzQx-m present green and yellow emission, λPL = 524 nm, PL = 10% and λPL = 544 nm, and PL = 9%, respectively. Additionally, the same done for MeOQx and MeO2Qx gives films with orange and red photoluminescence, λPL = 582 nm and λPL = 618 nm, respectively (indicated in Figure S17).

6. DSC and TGA CzQx tCzQx MeOQx MeO2Qx

100

Weight (%)

80

Ethyl acetate

60

40

20

0 100

200

300

400

500

600

Temperature (oC)

Figure S29. The TGA thermograms of reported materials (5% weight loss).

Exo

CzQxCzQxCzQxCzQx-

o

Tg= 66 C

i g h f

b)

tCzQx- i tCzQx-g tCzQx-h tCzQx-f

1st heating - - - 2nd heating

Exo

a)

o

Tg= 141 C

o

322 C

2nd heating o

Tg=145 C 177 oC o

Tc=188 C

o

Tg= 142 C

294 oC

o

Tg= 68 C

o

324 C

1st cooling

2nd heating

o

Tc= 145 C Tg= 127 C

o

T c=161 C

Endo

Endo

1st cooling

295 oC

50

100

150

200

250

o

326 C

50

300

100

150

o

250

300

350

d) 1st heating - - - 2nd heating

MeOQxMeOQxMeOQxMeOQx-

Tg= 104 oC

i g h f

Exo

Exo

200

Temperature (oC)

Temperature ( C)

c)

321 oC

o

o

176 C

MeO2Qx-i MeO2Qx-g MeO2Qx-h MeO2Qx-f

1st heating - - - 2nd heating

o

o

Tc=221 C

262 C o

242 C Tg= 109 oC

o

262 C

o

Tg= 104 C

o

242 C 234 o C

125 oC o

Tc = 144 C

o

262 C

o

Tg= 106 C

o

Tc= 207 C

150

200

Endo

Endo

o

262 C

100

242 oC

Tg= 110 o C

250

50

o

Ethyl acetate

o

234 C

128 o C

o

243 C o

235 C

100

150

200

Temperature (oC)

Temperature ( C)

Figure S30. DSC analysis of samples in solid state.

S23

250

7. Device fabrication Due to the very good solubility of presented molecules in toluene, the devices were produced by solution processing. Due to their sufficient solubility in toluene it was possible to use high molecular weight PVK (PVKH) as electron blocking and hole transport layer to improve charge confinement. Note that high molecular weight PVK (PVKH) is not soluble in toluene at room temperature. Therefore PVKH was spun from chloroform-chlorobenzene (95:5) mixture and PVK:PBD + dopant (5%) was spun from toluene. Electron transport layer TPBi was then evaporated, LiF and Al layers were evaporated later in this order.

Dev Dev Dev Dev

2

1 3 4 2

10

4

10

3

10

2

10

3x10

2

2x10

2

1x10

EQE (%)

-2

-2

Current density (mA cm )

4x10

2

Brightness (cd m )

5x10

1 Dev 1 Dev 3 Dev 4 Dev 2

2

10

1

0.1 2 10

0 0

2

4

6 8 Bias (V)

10

12

10

4

-1

Current efficiency (cd A )

100

-1

Current efficiency (cd A )

100

3

10 -2 Brightness (cd m )

10

1

0.1 0

1 Dev 1 Dev 3 Dev 4 Dev 2

Dev 1 Dev 3 Dev 4 Dev 2 10

10

1

10 -2 Current density (mA cm )

10

2

0.1 2 10

Figure S31. Additional OLED device characteristics.

S24

3

10 -2 Brightness (cd m )

10

4

8. References 1) Louillat, M.L.; Patureau, F.W. Toward Polynuclear Ru–Cu Catalytic Dehydrogenative C–N Bond Formation, on the Reactivity of Carbazoles. Org. Lett. 2013, 15, 164–167. 2) Liu, Y. Nishiura, M.; Wang,Y.; Hou, Z. Rapid Freeze-Quench ENDOR Study of Chloroperoxidase Compound I: The Site of the Radical. J. Am. Chem. Soc. 2006, 128, 5592–5593. 3) Baryshnikov, G. V.; Gawrys, P.; Ivaniuk, K.; Witulski, B.; Whitby, R. J.; Al-Muhammad, A.; Minaev, B.; Cherpak, V.; Stakhira, P.; Volyniuk, D.; Wiosna-Salyga, G.; Luszczynska, B.; Lazauskas,

A.;

Tamulevicius,

S.;

Grazulevicius,

J.

V.

Nine-Ring

Angular

Fused

Biscarbazoloanthracene Displaying a Solid State Based Excimer Emission Suitable for OLED Application. J. Mater. Chem. C 2016, 4, 5795-5805. 4)

Ku, C. –H.; Kuo, C. –H.; Chen, C. –Y.; Leung, M. –K.; Hsieh, K. –H. PLED Devices Containing Triphenylamine-Derived Polyurethanes as Hole-Transporting Layers Exhibit High Current Efﬁciencies. J. Mater. Chem. 2008, 18, 1296-1301.

S25

9. 1H and 13C NMR of compounds

1

H and 13C NMR of 3-methoxy-9H-carbazole

S26

1

H and 13C NMR of 2,3-di(9H-carbazol-9-yl)quinoxaline

S27

1

H and 13C NMR of 2,3-bis(3,6-di-tert-butyl-9H-carbazol-9-yl)quinoxaline

S28

1

H and 13C NMR of 2,3-bis(3-methoxy-9H-carbazol-9-yl)quinoxaline

O

N

N

N

N

O

S29

1

H and 13C NMR of 2,3-bis(3,6-dimethoxy-9H-carbazol-9-yl)quinoxaline

S30