Supporting Information

 Wiley-VCH 2013
69451 Weinheim, Germany

Highly Photoluminescent Carbon Dots for Multicolor Patterning,

Sensors, and Bioimaging**
Shoujun Zhu, Qingnan Meng, Lei Wang, Junhu Zhang, Yubin Song, Han Jin, Kai Zhang,
Hongchen Sun, Haiyu Wang, and Bai Yang*

anie_201300519_sm_miscellaneous_information.pdf
S. Zhu and Q. Meng conceived and designed the experiments. S. Zhu and Y. Song carried out most of the experiments
and data analysis. L. Wang and H. Wang performed the ultrafast spectroscopy. H. Jin and H. Sun performed the
bioimaging experiments. J. Zhang, K. Zhang and B. Yang discussed the results and commented on the manuscript. S.
Zhu wrote and revised the manuscript.

Table of contents in Supporting Information:

1. Experiment section.
2. Figure S1. The picture of CDs sample: the applied method can produce CDs with large scale; the solid powder shows
no fluorescence under the UV light.
3. Figure S2. AFM of CDs and their height distribution.
4. Figure S3. Raman spectra of CDs.
5. Figure S4 The 1H-NMR (a) and 13C-NMR (b) spectra of CDs.
6. Figure S5. FTIR spectra of CDs. It contains the -OH, epoxy, C=O, C-N (C=N), NH and CH groups.
7. Table S1. XPS analysis of CDs.
8. Figure S6. The fluorescent images of CDs aqueous solution (0.01 mg/mL) under UV (a), blue light (b) and green
light (c) excitations.
9. Figure S7. Electronic quenching on the excited state and photothermal effect of CDs aqueous solution.
10. Figure S8. The PL of CDs can be observed under day light excitation.
11. Figure S9. Stability of CDs.
12. Figure S10. TA spectra of CDs after 2000 W UV light. The decay tendency of CDs at different wavelength sections.
13. Table S2. Other reaction conditions for fluorescent carbon dots.
14. Table S3. Different carbonization conditions for fluorescent carbon dots: CD 1-4.
15. Figure S11. The UV/Vis absorption of CD 1-4.
16. Table S4. Element analysis of CD 1-4 by tuning the carbonization degree.
17. Figure S12. The excitation dependent behaviors of CD 1 to CD 4.
18. Figure S13. The possible formation mechanism of CDs was schemed from carbon precursor (polymer-like) to
carbonic core.
19. Figure S14. The resume and different graphic pattern obtained by CDs ink.
20. Figure S15. Fluorescence microscopy images of CDs micropattern, PVA/CDs nanofibers before and after 30 min
2000 W UV light exposure.
21. Figure S16. The scheme of fluorescence quenching of CDs/Fe3+.
22. Figure S17. The time-correlated single-photon counting (TCSPC) of CDs and CDs/Fe3+.
23. Figure S18. Effect of CDs on MC3T3 cells viability.
24. Figure S19. Cellular imaging of CDs to MC3T3.
25. Figure S20. The MC3T3 cells imaged under UV light after incubating in CDs/Fe2+ and CDs/Fe3+ culture solution,
respectively.
Experimental Section

Preparation of carbon dots. Carbon dots were prepared as follows: citric acid (1.0507 g) and ethylenediamine (335 µL)
was dissolved in DI-water (10 mL). Then the solution was transferred to a poly (tetrafluoroethylene) (Teflon)-lined
autoclave (30 mL) and heated at 150, 200, 250 and 300 ºC for 5 h. After the reaction, the reactors were cooled to room
temperature by water or naturally. The product, which was brown-black and transparent, was subjected to dialysis in
order to obtained the CDs. The production yield was ca. 58%. Other reaction conditions are listed in Table S2 and S3.
Printing of photoluminescent CD ink. A kind of commercial paper was chosen as printed paper which is proved a
suitable substrate for the CDs to adhere to and featured no background fluorescence under the UV lamp. Aqueous CDs
acting as colorless ink were injected into a vacant cartridge of a commercial inkjet printer (1*10-3 mg/mL). The desired
words or images were printed onto a piece of “paper” by a common printer connected to a laptop. The multi “color” inks
for “multi-color” printing were tuned to red, yellow and blue and loaded into vacant cartridge with CDs concentrations
of 0.08, 0.8 and 8 mg/mL. CD ink was drop-casted on the hydrophilic photoetched stripe for patterning micro-scale
structures. The fluorescence microscopy images were obtained under UV light (blue and green light excitations). The
patterns retained their PL intensity after 2000 W UV exposure for 30 min.
PVA/CDs nanofibers prepared by electrospinning. 0.15 g PVA (Mn=77000, 99% hydrolyzed) was first dispersed in
0.85 g of cool water for 2 h, then the mixture was heated at 95 °C in a water bath for 30 min to fully dissolve the PVA.
Finally, 0.001 g of the CDs and 0.004 g of dioctyl sulfosuccinate sodium salt (AOT) were added to the PVA solution
under vigorous stirring. Here, the presence of AOT was used to lower the surface tension and the viscosity of the PVA
solution. The resulting clear homogenous solution was used for electrospinning. A burette with an inserted Cu rod to
connect the high-voltage supply was filled with the PVA/CDs aqueous solution. An Al sheet connected to the ground
was used as the receiver. The distance between the burette tip and receiver was fixed at 15 cm and the high-voltage
supply was fixed at 15 kV. Pieces of Si sheet about 2 mm× 2 mm on the Al sheet for collecting the samples.
PDMAA/CDs nanocomposites. 0.005 g of CDs were dissolved in 10 mL N,N’-dimethylacrylamide (DMAA), Then,
100 µl of Darocur 1173 (as an initiator) was added to the solutions. The mixture was poured into the designed moulds,
and then exposed to UV radiation of a medium pressure mercury lamp of 1 kW for 2 min for polymerizing. Bulk
PDMAA/CDs nanocomposites retained their PL intensity after 2000 W UV exposure for 30 min.
Cellular toxicity test. Mouse osteoblastic cell line (MC3T3-E1) cells (104 cells/150 μL) were cultured first for 24 h in
an incubator (37 ℃, 5% CO2), and for another 24 h after the culture medium was replaced with 100 μL of Dulbecco's
modified Eagle's medium (DMEM) containing the CDs at different doses (0, 10, 20, 50, 100, 200, 400 μg/mL). Then, 20
μL of 5 mg/mL MTT solution was added to every cell well. The cells were further incubated for 4 h, followed by
removing the culture medium with MTT, and then 150 μL of DMSO was added. The resulting mixture was shaken for ca.
5 min at room temperature. The optical density (OD) of the mixture was measured at 490 nm. The cell viability was
estimated according to the following equation:
ODTreated
Cell Viability (%)  ( ) 100%
ODControl
(Where ODControl was obtained in the absence of CDs, and ODTreated obtained in the presence of CDs.)
Cellular imaging. The cells were cultured in DMEM supplemented with 10% fetal bovine serum and 1%
penicillin/streptomycin. Suspensions (0.25 mg/mL) of CDs from the stock solution were prepared with Dulbecco’s
phosphate buffer saline (DPBS). After sonication for 10 min to ensure complete dispersion, an aliquot (typically 0.1 mL)
of the suspension was added to the well of a chamber slide, then incubated at 37 ℃ in a 5% CO2 incubator for 20 h. Prior
to fixation of the cells on the slide for inspection with a fluorescence microscope, the excess CDs were removed by
washing 3 times with warm DPBS. The Fe2+ and Fe3+ ions were also added to incubate before bioimaging. The
bioimaging was taken at Olympus BX51.
Characterization. High-resolution transmission electron microscope (HTEM) was recorded on JEM-2100F and FEI
Tecnai F20 (The decreased electro-beam intensity and increased exposure time can be beneficial to the obtained TEM images
of CDs). AFM images were recorded in the tapping mode with a Nanoscope IIIa scanning probe microscope from Digital
Instruments under ambient conditions. Fluorescence spectroscopy was performed with a Shimadzu RF-5301 PC
spectrophotometer. UV-vis absorption spectra were obtained using a Shimadzu 3100 UV-vis spectrophotometer. IR
spectra were taken on a Nicolet AVATAR 360 FT-IR spectrophotometer. The fluorescent images were taken at Olympus
BX51 at UV, blue and green light excitations, the color filters have been integrated in the fluorescence microscope, no
extra filters need be used (See http://www.olympusmicro.com/primer/java/lightpaths/bx51fluorescence/index.html). IR
spectra were taken on a Nicolet AVATAR 360 FT-IR spectrophotometer. Raman spectra were measured with a
Renishaw Raman system model 1000 spectrometer with radiation at 633 nm. X-ray Photoelectron Spectroscopy (XPS)
was investigated by using ESCALAB 250 spectrometer with a mono X-Ray source Al Kα excitation (1486.6 eV).
Binding energy calibration was based on C1s at 284.6 eV. Elemental analysis was performed on Elementar Vario
MICRO CUBE, each data was parallel at least twice and the average values were obtained by measuring three kinds of
sample batches. NMR was done on AVANCEIII500 (Bruker).
Time resolved photoluminescence. Nanosecond fluorescence lifetime experiments were performed by the time-
correlated single-photon counting (TCSPC) system under right-angle sample geometry. A 379 nm picosecond diode
laser (Edinburgh Instruments EPL375, repetition rate 2 MHz) was used to excite the samples. The fluorescence was
collected by a photomultiplier tube (Hamamatsu H5783p) connected to a TCSPC board (Becker&Hickel SPC-130).
Time constant of the instrument response function (IRF) is about 300 ps.
Femtosecond transient absorption setup. The TA setup consisted of 400 nm pump pulses doubled from 800 nm laser
pulses (~100 fs duration, 250 Hz repetition rate) generated from a mode-locked Ti: sapphire laser/amplifier system
(Solstice, Spectra-Physics) and broadband white-light probe pulses generated from 2-mm-thick water. The relative
polarization of the pump and the probe beams was set to the magic angle. The TA data were collected by a fiber-coupled
spectrometer connected to a computer. The group velocity dispersion of the transient spectra was compensated by a chirp
program. All the measurements were preformed at room temperature. Pump-power dependent measurements are carried
out. In the acceptable range, no pump intensity dependent dynamics are observed.
Quantum yields (QY) measurements.
Quinine sulfate (0.1M H2SO4 as solvent; QY=0.54) were chosen as standards. The QYs of CDs (in water) were
calculated according by two kinds of methods. The average error was less than 5% for these two kinds of methods. The
QYs in Table S2 and S3 are based on the second method.
1. The QY was determined by reference point method.
 x  st ( I x / I st )( x2 /  st2 )( Ast / Ax )
Where φ is the QY, Ι is the measured integrated emission intensity, η is the refractive index of the solvent, and A is the
optical density. The subscript "st" refers to standard with known QY and "x" for the sample. In order to minimize re-
absorption effects, absorption in the 10 mm fluorescence cuvette was kept below 0.10 at the excitation wavelength (344-
360 nm).
2. The QY was determined by slope method[1] by the reference of quinine sulfate: compared the integrated
photoluminescence intensity and the absorbance value [several values (less than 0.1 at excitaion wavelength) gave the
curve] of the samples with that of the references. Then used the equation:
φx=φst(Kx/Kst)(ηx/ηst)2
Where φ is the QY, K is the slope determined by the curves and η is the refractive index. The subscript “st” refers to the
standards and “x” refers to the unknown samples. For these aqueous solutions, ηx/ηst=1.
Figure S1. The picture of CDs sample: the applied method can produce CDs with large scale; the solid powder shows no
fluorescence under the UV light.

Figure S2. AFM of CDs and their height distribution.

Figure S3. Raman spectra of CDs. G band at 1598 cm-1 and D band at 1350 cm-1 were not obvious. The high
fluorescence may also disturb the Raman characterization.
4500

4000

Intensity 3500

3000

2500

2000
1200 1500 1800 2100 2400
-1
Raman shift (cm )
Figure S4 The 1H-NMR (a) and 13C-NMR (b) spectra of CDs.
Figure S5. FTIR spectra of CDs. It contains the -OH, epoxy, C=O, C-N (C=N), NH and CH groups.

Table S1. XPS analysis of CDs.

Peak Binding
CDs
Energy

C-C/C=C (%) 284.75 55.81

Oxygenated
286.15 25.02
carbon (%)
Nitrous
287.7 19.17
carbon (%)

Figure S6. The fluorescent images of CDs aqueous solution (0.01 mg/mL) under UV (a), blue light (b) and green light
(c) excitations. The fluorescence microscopy images were obtained through band-pass filters of different wavelengths:
450 nm, 550 nm and 580 nm.
Figure S7. a) Electronic quenching on the excited state. b) Photothermal effect of CDs aqueous solution (5 mg/mL);
illuminated by 3 W/cm2 808 nm laser.

Figure S8. The PL of CDs can be observed under day light excitation (fresh solutions in DMF and water).
Figure S9. Stability of CDs. a) Effect of ionic strengths on the fluorescence intensity of CDs (0.01 mg/mL) (ionic
strengths were controlled by various concentrations of KCl). b) Effect of pH on the fluorescence intensity of CDs (0.01
mg/mL). c) Dependence of fluorescence intensity on UV excitation time for CDs in DI water (0.01 mg/mL).
Figure S10. TA spectra of CDs after 2000 W UV light. The decay tendency of the CDs at different wavelength sections.

Data notes:
The CDs exposed by high power UV light show a drastic degree of decay in all wavelength ranges, electron at excited state left
only 20%, leading to the decrease of the QY.
For prepared CDs, during the time over 1300 picosecond, the electron at the excited state still remained over 80% in the 430-
475 nm wavelength range, which led to the subsequent radiative fluorescence. This ultrahigh stable electron at the excited state
is conducive to the high QY. Actually, for CDs with low QY, the electron at the excited state decay drasticly in the first several
picosecond sections (J. R. Lakowicz, Principles of Fluorescence Spectroscopy, Third Edition; Springer: New York, 2006;
Chem. Commun. 2012, 48, 10889.).
Table S2. Other reaction conditions for fluorescent carbon dots (In (Teflon)-lined autoclave/200 ºC for 5 h).

QY Photos of CDs Peak

standard by aqueous position
Serial CDs from different reaction ratios
quinine (360 nm
visible UV
sulfate (%) excition)
light light

Citric acid, Water, 10

1 Ethylenediamine, 33.5 µL 20.9 447 nm
1.051 g mL

Citric acid, Water, 10

2 Ethylenediamine, 268 µL 65.3 443 nm
0.42 g mL

Citric acid, Water, 10

3 Ethylenediamine, 536 µL 80.6 445 nm
0.42 g mL

Citric acid, Water, 10

4 Ethylenediamine, 1275 µL 55.9 444 nm
0.42 g mL

Water, 10
5 Citric acid, 1.051 g 7.2 451 nm
mL

Water, 10
6 Ethylenediamine, 335 µL 3.8 442 nm
mL

Citric acid, Water, 10

7 Ethylamine, 335 µL 8.4 459 nm
1.051 g mL

Citric acid, n-Heptylamine, Water, 10

8 7.7 452 nm
1.051 g 600 µL mL

Sodium citrate, Water, 10

9 Ethylenediamine, 335 µL 21.6 442 nm
1.471 g mL

Citric acid, Water, 10

10 Urea, 0.3 g 19.4 435 nm
1.051 g mL

Citric acid, p-phenylenediamine, Water, 10

11 2.7 498 nm
1.051 g 0.541g mL

Data notes:
1. Through lots of experiment, we found that the –OH, –COOH and –NH2 groups are very important for the formation of CDs.
In detail, if the reactant only contained –OH and –COOH groups, the PL QYs of prepared CDs are always less than 10%. Then,
if amino molecule especial di-amine molecule or multi-amine molecule existed (such as citric acid and ethylenediamine), the
PL QY of most prepared CDs can be over 10%. Single amino connected molecules are also suffered the insufficiency of
condensation polymerization.
2. The ratio of citric acid and ethylenediamine (1.051g and 335µL) was chosen (Table S3) for detailed discussion and
application for three reasons:
(1) Though the PL QY of CDs (Serial 2 and 3 in Table S2) prepared by increased the ratio of ethylenediamine and citric acid
was higher than that of CD 2 in Table S3, the prepared CDs (Serial 2 and 3 in Table S2) are viscous and will be dried for very
long time, it is disadvantageous for practical application, the possible reason was schemed in Figure 1a. When the molar ratio
of ethylenediamine and citric acid is less than 3:2, due to the residual –COOH and –OH, the further carbonization proceeded
after the formation of polymer-like CDs (step 4 in Figure 1a). When the molar ratio of ethylenediamine and citric acid is more
than 4:2, the transitional polymer-like CDs would be further carbonized hardly;
(2) The rate of production was higher compared with other ratio;
(3) As a result, for practical application and prove universality, it’s not necessary to use the CDs with the highest QY.
3. Maybe amide-containing fluorophores affect the PL center a lots, but it still lacks sufficient evidence, for example, CDs
from other di-amine molecules (e.g. butanediamine, p-phenylenediamine) possess the QY less than 20%, the PL center was
controlled by the synergy of multi-factors such as, chemical groups, surface state, surface passivation and carbogenic core.
Table S3. Different fluorescent carbon dots by tuning the carbonization degree.
PL intensity
QY standard
after 9 min
Serial CDs from different reaction ratio Reaction conditoins by quinine
2000 W UV
sulfate (%)
exposure
Citric
Ethylenediamine, 335 Water, In (Teflon)-lined
CD 1 acid, 75.2 8.75%
µL 10 mL autoclave/150 ºC for 5 h
1.051 g
Citric
Ethylenediamine, 335 Water, In (Teflon)-lined
CD 2 acid, 60.2 10.63%
µL 10 mL autoclave/200 ºC for 5 h
1.051 g
Citric
Ethylenediamine, 335 Water, In (Teflon)-lined
CD 3 acid, 24.7 20.34%
µL 10 mL autoclave/250 ºC for 5 h
1.051 g
Citric
Ethylenediamine, 335 Water, In (Teflon)-lined
CD 4 acid, 17.3 45.25%
µL 10 mL autoclave/300 ºC for 5 h
1.051 g

Data notes:
1. The CDs prepared below 250 ºC always have difficult TEM imaging, maybe due to the unshaped nanoparticles especially in
low synthesis temperature.
2. The QY was determined by slope method by the reference of quinine sulfate: compared the integrated photoluminescence
intensity (360 nm excitation) and the absorbance value [several values (less than 0.1 at excitaion wavelength) gave the curve]
of the CDs samples with that of the references. Then used the equation:
φx=φst(Kx/Kst)(ηx/ηst)2
Where φ is the QY, K is the slope determined by the curves and η is the refractive index. The subscript “st” refers to quinine
sulfate and “x” refers to the CD 1-4. For these aqueous solutions, ηx/ηst=1. So the equation was simplified to:
φx=φst(Kx/Kst)
quinine
Serial CD 1 CD 2 CD 3 CD 4
sulfate
K 959 1335 1069 438 307
Φ (%) 54 75.2 60.2 24.7 17.3

The QYs in Table S2 were obtained by the same method.

Figure S11. The UV/Vis absorption of CD 1-4, the peak at 344 nm (surface/molecule center) decreases from CD 1 to
CD 4 while the peak at 240 nm (carbonic core center) increases from CD 1 to CD 4.

Table S4. Element analysis of CD 1-4 by tuning the carbonization degree. The oxygen content decresces from CD 1 to
CD 4. It indicated the sufficient carbonization in high reaction temperature.

Test molecule formula C (%) H (%) O (Calculated) (%) N (%)

CD 1 C4H5.036O2.11N0.945 48.01 5.036 33.724 13.23

CD 2 C4.26H5.814O1.68N1.16 51.13 5.814 26.806 16.25

CD 3 C4.72H6.771O1.28N1.152 56.60 6.771 20.499 16.13

CD 4 C5.85H7.148O0.75N0.759 70.25 7.148 11.972 10.63

Figure S12. The excitation dependent behaviors of CD 1 (a), CD 2 (b), CD 3 (c) and CD 4 (d).
Figure S13. The possible formation mechanism of CDs was schemed from carbon precursor (polymer-like) to
carbogenic CDs.

Data notes:
The reaction temperature is important for the further carbonization of CDs (Table S3), from low temperature to high
temperature, the polymer-like CDs were changed to carbogenic CDs:
1. From CD 1-4, the peak intensity at 344 nm (surface/molecule state section) decreased while the intensity at 240 nm
(carbogenic core section) increased in UV-Vis spectra (Figure S11);
2. The percentage of oxygen decreased while the percentage of carbon increased (Table S4);
3. The increasing illumination stability and obvious excitation-dependent PL were also observed in CD 4 (Table S3,
Figure S12).
Figure S14. The resume and different graphic pattern (a) obtained by CD ink, the below (b) pictures are designed pattern
printed on the same papers under the bright light and UV light excitations.
Figure S15. Fluorescence microscopy images of CDs micro stripe (a) and PVA/CDs nanofibers (b) before and after 30
min 2000 W UV exposure, the exposure times of taking fluorescent image are the same for all pictures. The PL intensity
nearly un-changed after 30 min 2000 W UV exposure. All the images for (a) and (b) have the same scale bars. The
fluorescence microscopy images were obtained through band-pass filters of different wavelengths: 450 nm, 550 nm and
580 nm.
Figure S16. The scheme of fluorescence quenching of CDs/Fe3+.

Data notes:
To estimate their HOMO and LUMO energy levels, cyclic voltammetry (CV) was carried out by using a standard three-
electrode system, which consists of platinum sheet as the working electrode, a platinum wire as the counter electrode, and
Ag/AgCl as the reference electrode. CV was recorded in DMF containing saturated CDs and 0.1 M (Bu)4NBF4 as the
supporting electrolyte. The HOMO and LUMO energy levels in eV of CDs were calculated according to the following
equations:
E(HOMO)= -e(Eox + 4.4) (eV)
E(LUMO)= -e(Ered + 4.4) (eV)

Where Eox and Ered are the onset of oxidation and reduction potential, respectively.
The Ered was determined to be -0.96 V.
The corresponding LUMO level was -3.4 eV. However, the HOMO level could not be obtained due to the irreversible of
positive scanning of CV (b in the following picture). The HOMO level from ultraviolet photoelectron spectroscopy (UPS) is
thin film state instead of solution state, so we calculated the HOMO through the following equations:[2]
E(HOMO) = E(LUMO) -∆E (3.6eV)
Where ∆E is the absorption peak of CDs in UV-Vis spectrum.

The cyclic voltammogram of CDs in CDs and 0.1 mol/L (Bu)4NBF4 DMF solution. a) negative scanning. b) positive
scanning.

The crystal field stabilization energy (CFSE) was determined to 2.7 eV as complexation between Fe3+ ions
and phenolic hydroxyl of CDs.[3]
Figure S17. The time-correlated single-photon counting (TCSPC) of CDs and CDs/Fe3+. The average lifetime of CDs is
13.7 ns and contains two lifetime components of 42.3 ns (~17%) and 7.9 ns (~83%); The average lifetime of CDs/Fe3+ is
2.3 ns and contains two lifetime components of 7.8 ns (~20%) and 0.9 ns (~80%) (360 nm excitation, delay time at 443
nm emission).

Figure S18. Effect of CDs on MC3T3 cells viability.

Figure S19. Cellular imaging of CDs to MC3T3. The washed cells imaged under bright field, UV light, blue light and
green light excitations after incubating in CDs culture solution.
Figure S20. The MC3T3 cells imaged under UV light after incubating in CDs/Fe2+ and CDs/Fe3+ culture solution,
respectively.

Data notes:
1. The PL intensity for using CDs/Fe2+ and CDs/Fe3+ in cells are higher than that of using CDs system, the reason was
still not clear at present.
2. For different cells, The PL intensity may change little for using CDs/Fe2+ and CDs/Fe3+ system, may be the
complexity of cells affect the PL quenching of CDs/Fe3+ in cells.

References:
[1] H. Zheng, Q. Wang, Y. Long, H. Zhang, X. Huang, R. Zhu, Chem. Commun. 2011, 47, 10650.
[2] V. Gupta, N. Chaudhary, R. Srivastava, G. D. Sharma, R. Bhardwaj, S. Chand, J. Am. Chem. Soc. 2011, 133, 9960.
[2] Burrows et al. Student's solutions manual, Oxford University press, 2009