Subscriber access provided by UNIV OF CALIFORNIA SAN DIEGO LIBRARIES
Article
Boron precursor dependent evolution of differently emitting carbon dots Jayasmita Jana, Mainak Ganguly, Kuttay R. S. Chandrakumar, Gowravaram Mohan Rao, and Tarasankar Pal Langmuir, Just Accepted Manuscript • DOI: 10.1021/acs.langmuir.6b04100 • Publication Date (Web): 26 Dec 2016 Downloaded from http://pubs.acs.org on December 28, 2016
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Langmuir is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Page 1 of 28
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Langmuir
Boron precursor dependent evolution of differently emitting carbon dots Jayasmita Jana,a Mainak Ganguly,b Kuttay R. S. Chandrakumar,c Gowravaram Mohan Rao,d Tarasankar Pala* a
b
Department of Chemistry, Indian Institute of Technology, Kharagpur-721302, India
Department of Chemistry, Furman University, Greenville, South Carolina 29613, United States c Theoretical Chemistry Section, Chemistry Group, Bhabha Atomic Research Centre, Mumbai 400085, India d
Department of Instrumentation, Indian Institute of Science, Bangalore 560 012. India
*
Corresponding author. Phone: +91-03222 283320. E-mail:
[email protected] ABSTRACT Attention has been directed towards electron deficient boron doping in carbon dots (CDs) with the expectation of revealing new photophysical aspects in accordance with varying amount of boron content. It has been emphatically shown that boron uptake in CDs varies with different boron precursors evolving altered emissive CDs. Boron doping in CDs causes definite surface defect due to generation of electron deficient states. Modified hydrothermal treatment of a mixture of ascorbic acid (AA) and different boron precursor compounds (borax/boric acid/sodium borate/sodium borohydride) produces different kinds of boron doped CDs (BCDs). These BCDs (< 6 nm) differ in size, emission maxima (~15 nm) and fluorescence intensity but carry unchanged excitation maxima (365 nm). These differences are related to the nature of boron precursor compounds. The most fluorescing BCD (quantum yield ~ 5%) is identified from borax mediated reaction and is used for detection of Fe(III) in nanomolar level in water via fluorescence ‘Turn Off’ phenomenon. Again, Fe(III) infested CD solution regains its lost fluorescence with AA paving the way of nanomolar level AA detection from the same pot. The proposed method has been tactfully made interference free for the quantitative measure of Fe(III) and AA in real samples. Furthermore, new photophysical properties of the CDs with variable boron content supplement information which is hitherto unknown. Theoretical calculations also justify the observed optical behaviour of the as-synthesized BCDs. Calculation describes variable amount of boron doping related huge charge polarization within the carbon surface, leading to the formation of surface defects. Thus subsequent electronic transition related red shift in the absorption spectrum authenticates experimental findings. INTRODUCTION
ACS Paragon Plus Environment
Langmuir
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
In recent days fluorescent carbon dots (CDs) have emerged as a potential fluorescent probe for various applications including chemical sensing,1 bioimaging,2 catalysis,3 fluorescent markers4 etc. Scientists have described CDs as highly fluorescent surfacepassivated small carbon nanoparticles5 with sp2 and sp3 hybridized carbons atoms with plenty of oxygen-containing groups.6 In general, CDs are nontoxic, cytocompatible, water soluble and photostable.7 These features make CDs preferable over other fluorescent quantum dots. CDs can be easily synthesized through microwave treatment,8 hydrothermal treatment,9 plasma treatment,10 laser ablation/passivation,11 and electrochemical method.12 Generally CDs lie within a diameter range of 10 nm.13 Tiny size of CDs causes quantum confinement of emissive energy traps on particle surface14 which is a reason of intense fluorescence. Some other reasons behind the fluorescence of CDs is radiative recombination of excitons, HOMOLUMO energy gap, triplet carbene at zigzag ends, surface passivation etc.6 The emissive property can also be improved by further doping of hetero atoms like boron, nitrogen, phosphorous etc.15 Doping of different heteroatoms in a green chemistry strategy ways has become the area of much interest in recent days. Generally doping of nitrogen as hetero atom is more explored as nitrogen is sufficiently electron rich than carbon which provides n-type doping characteristics.16 Sulfur, phosphorous are also used for the same reason. Doping of electron deficient boron has not been thoroughly surveyed yet except one or two cases.16,17 However, doping of boron, along with nitrogen, phosphorous has been reported.15,18,19 Just like nitrogen and sulfur, boron can be incorporated into carbon dots where boron embraces across covalent interaction and becomes an integral dopant. Iron (Fe) is an essential element where it prevails mainly in two common oxidation states, +2 [ferrous, Fe(II)] and +3 [ferric, Fe(III)]. The redox Fe(III)/Fe(II) couple has an essential role in the living systems but high concentration of Fe ions endangers cells and tissues by readily participating in oxidation-reduction reactions forming reactive oxidative intermediates.20 So determination of Fe in trace level is always important. There are colorimetric,21 potentiometric,22 fluorimetric,23,24 flame atomic absorption spectroscopic,25 mass spectrometric26 etc. techniques for Fe(III) sensing. However, fluorometric technique is much more sensitive for routine sensing, as in this technique the signal to noise ratio is high. Another important biologically important compound is ascorbic acid, a water soluble compound mostly found in citrus fruits and plants.27 It works as antioxidant, enzyme factor, nutrition cofactor in biological system.28 Also it prevents free radical-induced diseases like cancer and Parkinson's disease, acting through a free radical pathway.29 Lack of ascorbic acid
ACS Paragon Plus Environment
Page 2 of 28
Page 3 of 28
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Langmuir
causes scurvy. However, excess ascorbic acid in body causes urinary stone, diarrhea, and stomach convulsion.30 Hence detection of ascorbic acid is always important. Electrochemical sensing of ascorbic acid31 is often found to be hampered by coexisting compounds dopamine and uric acid32 because of the same oxidation peak position.33 There are other reports of high performance liquid chromatography (HPLC),34 capillary zone electrophoresis,35 flowinjection spectrophotometry36 and colorimetric determination37 techniques. Here also the fluorescence technique is proved to be very efficient. Most of the fluorescence techniques are based on turn off phenomenon for ascorbic acid determination.38-42 Thus designing of a fluorescence probe that undergoes Turn On phenomenon in presence of ascorbic acid becomes important and sensitive. In our study we have reported the syntheses of four types of fluorescent CDs with different amount of boron doping and elaborately accounted their photophysical properties. It has been shown from detailed investigation that precursor controls the nature of CDs. Also we have designed a fluorescence Turn Off/Turn On sensor for Fe(III) and ascorbic acid from a single pot. EXPERIMENTAL SECTION Chemicals and materials All the reagents used throughout the experiment were of AR grade. Triple distilled water was employed during the experiment. Ascorbic acid, borax (anhydrous), sodium borohydride, boric acid, sodium borate, dopamine (DA), L-3,4-dihydroxyphenylalanine (L-DOPA), catechol, glucose, lactose, fructose, sucrose and all metal salts [copper sulphate pentahydrate CuSO4. 5H2O, cadmium sulphate hydrate CdSO4.H2O, cobalt nitrate Co(NO3)2.6H2O, ammonium iron(II) sulphate hexahydrate (NH4)2Fe(SO4)2.6H2O, iron(III) chloride FeCl3, iron(III) sulfate Fe2(SO4)3, iron(III) nitrate nonahydrate Fe(NO3)3 9H2O, mercury chloride
HgCl2, magnesium sulphate MgSO4, nickel sulfate hexahydrate NiSO4.6H2O, lead nitrate Pb(NO3)2, zinc sulfate hexahydrate ZnSO4.6H2O] were purchased from Sigma-Aldrich. Sodium hydroxide (NaOH) was obtained from HiMedia Laboratories Pvt. Ltd. All the reagents were used without further purification. All glassware were cleaned with freshly prepared aqua regia, rinsed with sufficient amount of distilled water, and dried well before use. Instrumentation All UV−vis absorption spectra were recorded in a SPECTRASCAN UV 2600 digital spectrophotometer (Chemito, India). The fluorescence measurement was done at room
ACS Paragon Plus Environment
Langmuir
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
temperature using a LS55 fluorescence spectrometer (Perkin-Elmer, Waltham, MA). Fluorescence lifetimes were measured with Easy life V (Optical Building Blocks Corporation) equipped with a 380 nm LED excitation source. A nonlinear least square (χ2) fit was tested to determine the fit of the decay rate to a sum of exponentials and a visual inspection of the residuals and the autocorrelation function were used to determine the quality of the fit. Powder X-ray diffraction (XRD) was done with a PW1710 diffractometer, Philips, Holland, instrument. Surface chemical analysis was examined by X-Ray photoelectron spectroscopy, XPS (SPECS PHOIBOS 100 MCD energy analyser), in an ultra-high vacuum environment (1.9 × 10−9 mbar) using Al Kα anode (1486.6 eV). Pass energy of 40 eV for survey scan and 30 eV for high resolution scan have been used during acquisition of the XPS spectra. High-resolution scans were recorded for understanding the chemical environment of the constituent elements present at the surface. Each scan was repeated 3 times to reduce the signal to noise ratio. Fourier transform infrared (FTIR) studies were carried out with a Thermo-Nicolet continuum FTIR microscope. Transmission electron microscopy (TEM) analysis was performed with a H-9000 NAR instrument, Hitachi, using an accelerating voltage of 300 kV. Raman spectroscopic analyses were done using a Raman spectrophotometer (Jobin Yvon Horiba, excitation source: 514 nm Ar ion gas laser). Preparation of CDs Carbon dots (CDs) were prepared under modified hydrothermal treatment (MHT) of an aqueous mixture of ascorbic acid and one boron compound like borax/boric acid/ borate/ borohydride. In a typical synthesis 0.6 mL of 10-1 M ascorbic acid and 0.4 mL of 10-1 M borax/boric acid/borate/borohydride (as per the choice) were mixed and the final volume was made up to 6 mL by distilled water. Then the reaction mixture was sonicated for 30 minutes. After that the homogeneous reaction mixture, taken in a 15 mL screw-cap test tube, was subjected to MHT treatment in our laboratory made set up for 7 hours.43 The temperature was maintained 1800C throughout the reaction using a 200 W electrical bulb heating. After 7 hours of MHT treatment, a pale yellow solution resulted in. The solution was centrifuged to remove any gritty non fluorescent particles. Then the transparent solutions were reserved for future use. Measurement of quantum yield (Φ) The quantum yield of CDs was measured according to a reference point method.44 Quinine sulfate in 0.1M H2SO4 (literature quantum yield 0.54) was used as the standard
ACS Paragon Plus Environment
Page 4 of 28
Page 5 of 28
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Langmuir
reference. The quantum yield of a sample (sl) was measured with respect to the known quantum yield of the standard reference (st) using the following formula, Φsl=Φst(Fsl/Fst)( η2 sl/ η 2st)(Ast/Asl) Where Φ is the quantum yield, F is the fluorescence intensity, η is the refractive index of the solvent, and A is the optical density. The subscript "st" and "sl" refer to standard with known quantum yield and the sample, respectively. Detection of Fe(III) In a typical assay, 100 µL 10-2 M ferric chloride solution was added into 300 µL borax mediated CD solution. The solution was then diluted to 3 mL for fluorescence studies. Selectivity was tested in presence of other different metal salts of equivalent amount. The interference due to Fe(II) was removed by adding 100 µL 10-2 M fructose to the solution containing both Fe(III) and Fe(II). Also the effect of fructose was observed on the solutions containing Fe(III) and Fe(II) individually. The fluorescence of the fructose added solutions were measured within 10 minutes of addition. Detection of ascorbic acid In a typical assay, 100 µL 10-2 M ascorbic acid solution was added to 400 µL Fe(III) mixed CD solution. The solution was then diluted to 3 mL. Measurements were done within 30 minutes. Selectivity was tested in presence of dopamine (DA), L-3,4-dihydroxyphenylalanine (L-DOPA), catechol, glucose, lactose, fructose, sucrose. Theoretical Methods and Computational Details The geometry of the model carbon dots (CD) was constructed with 19 benzenoid fused ring structures, without any symmetry constraints, as shown in Figure 3. All the terminal carbon atoms were saturated with hydrogen atoms.
The boron atoms were
introduced in the CD framework as boron atoms and boron oxides (B2O and B2O3) at the centre of the model CD systems. In addition, as the experiments suggested the presence of hydroxyl groups at the peripheral part of the CD, four hydroxyl groups were introduced at the terminal carbon atoms of the CD. For the case of closed shell systems, the RHF-types of Kohn-Sham methods have been applied. Herein, the def 2-SVP basis set was employed for all atoms and the energy calculations as well as geometry optimization was performed at the level of density functional theory (DFT) using the BP86 exchange-correlation functional. The time dependent DFT (TDDFT) based calculations for the evaluation of electronic excitation spectrum were performed by the Half-and-half hybrid functional by Becke’s
ACS Paragon Plus Environment
Langmuir
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
exchange functional combined with Lee-Yang-Parr correlation functional (BHHLYP).45-48 All the theoretical calculations were performed with the use of ORCA system of programs.49 The grid based DFT was used as present in ORCA, that employed a typical grid quadrature to compute the integrals. During the SCF procedure, the grid contained 96 radial shells with 36 and 72 angular points. The charge on each atom was obtained by the Mulliken population analysis. RESULTS AND DISCUSSION Highly fluorescent boron doped carbon dots (BCDs) have been synthesized reproducibly from aqueous mixture of ascorbic acid (AA) and different boron precursor (borax/borate/boric acid/borohydride) compounds under hydrothermal treatment in 7 hours. It may be mentioned that microwave assisted synthesis of boron doped carbon dots is no way superior to our adopted and popular cost effective MHT43 method. Here AA works as the carbon source and borax/borate/boric acid/borohydride (individually) supplements boron. CDs synthesized by these precursors are marked as B1CD, B2CD, B3CD and B4CD, respectively. TEM studies reveal that the particles are quasi-spherical and remain almost aggregated. Particle diameter varies within the range of ~5-6 nm for B1CD, 3-4 nm for B2CD, 3-5 nm for B3CD and 4-5 nm for B4CD (Figure 1A-D). HRTEM image (Figure 1EH) shows that for B1CD, B2CD, B3CD and B4CD there exist a crystalline region that corresponds to (002), (002), (222) and (002) plane of carbon.50,51,52 From XRD we found peaks at 23.3º, 23.6º, 22.3º and 23.5º for B1CD, B2CD, B3CD and B4CD respectively (Figure 1I). These XRD peaks indicate the presence of (002),53 (002),53 (222) (JCPDS card no. 85-0409) and (002)53 planes of carbon for B1CD, B2CD, B3CD and B4CD respectively. However, broad nature of peak shows the presence of some sort of disorderness and amorphous region as expected. We have characterized the as synthesized CDs using XPS like the reported literature.51-52 Full range XPS spectrum of B1CD shows peaks at 284.05 eV, 532.34eV and 192.37 eV due to the presence of C1s, O1s and B1s respectively (Supporting Information, Figure S1). Further study of C1s core shows peaks for C-B, C=C and C-O at 283.9 eV, 284.98 and 288.33 eV respectively (Supporting Information, Figure S1B). The same study for B1s core shows peaks for B-C, B-CO and B-O at 190.7 eV, 191.99 eV and 192.54 eV respectively (Supporting Information, Figure S1C). Such peaks for definite bonds (C-B, C=C, C-O, B-C, B-CO and B-O) are obtained for B2CD, B3CD and B4CD also (Supporting Information, Figure S1D-L). The carbon standard for XPS measurement is 284.6
ACS Paragon Plus Environment
Page 6 of 28
Page 7 of 28
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Langmuir
eV. The XPS data agrees well with the results of FTIR spectra. In FTIR spectrum peaks at 678 cm-1, 1050 cm-1, 1122 cm-1, 1343 cm-1, 1404 cm-1, 1637 cm-1, 3445 cm-1 peaks for B1CD corresponds to =C-H, B-O-C, C-B, C-O, C=O and -OH bands respectively.15,50,63 B2CD, B3CD and B4CD also contain =C-H, B-O-C, C-B, C-O, C=O and -O-/-OH bonds (Figure 1J). All these information suggest successful doping of boron in all these CDs. Boron being an electron deficient atom introduces defects in the energy states of CDs. This will give rise to the emissive traps at the particle surface. Thus an intense emission resulted in. Also the transition modes get changed and we noticed a weak absorption motif associated with asymmetric shoulder peaks.50 In the absorption spectra, two weak peaks are observed at 256 nm and 341 nm for B1CD. These peaks correspond to π-π* and n-π* transition respectively. B2CD has two very weak absorption peaks at 224 and 256 nm. For B3CD and B4CD the characteristic peaks are too weak to detect (Figure 2A). Raman spectroscopic analysis shows the presence of weak and broad peaks (Supporting Information, Figure S2). Such weak Raman signals appear for highly fluorescent particles.54 It is now well understood that doping in CDs is a fascinating area of photophysical research. The as-synthesized boron doped CDs show most intense fluorescence emission peak when excited at 365 nm wavelength. Trial run for excitation wavelength scan allows us to choose 365 nm as the appropriate excitation wavelength (Supporting Information, Figure S3). However, the fluorescence intensities and emission maxima differ significantly for as obtained four types of CDs when excited at 365 nm (Figure 2B). The order of emission maxima is B1CD>B4CD>B2CD>B3CD (λem= 460 nm, 453 nm, 452 nm, 445 nm, respectively). Also the position of peak maxima differs for the CDs. This situation can be explained by considering density of the doped hetero atom, boron. It may be mentioned that electron deficient boron brings surface defect in CDs and the effect is the generation of electron deficient states.55 As in B1CD higher unit of B is present, electron deficient states are expected somewhat at higher energy because of the most pronounced interaction of B with CDs. Thus, it shows highest red shifted emission maxima. But for a particular precursor, accommodation of higher amount of B in CDs only causes intensity variation of the emission peak without any change in peak position. Again, higher the atomic percentage of boron, higher is the expected surface defect. Here comes the importance of incorporation of B from different precursor compounds during BCD syntheses. In aqueous solution borax, sodium borate, sodium borohydride dissociate fully while boric acid, being a very weak acid, dissociates to a negligible extent. Thus the former three compounds take part in the reaction
ACS Paragon Plus Environment
Langmuir
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 8 of 28
with their corresponding anionic counterpart B4O72-, BO33-, BO2-/BH4- respectively. Borohydride forms metaborate in water but complete conversion does not occur, so we will consider the average % boron from these two forms, while boric acid enters as total undissociated unit i.e. H3BO3. Thus the atomic % of boron in B3CD becomes the lowest in case of boric acid mediated synthesis. The estimated % boron in the as obtained CDs also authenticates the fact and that B1CD, B2CD, B3CD and B4CD contain 1.16, 1.12, 1.025 and 1.14 % boron respectively. The result is corroborated from XPS studies. Thus the intensity variation of BCDs with the nature of precursor compounds is authenticated. The precursors, AA and borax/borate/boric acid/borohydride possess no fluorescence property at the normal condition. However, hydrothermal treatment of these compounds has been
done individually.
Although
the
products
of
MHT
of borax/borate/boric
acid/borohydride are non fluorescent at the experimental condition, the product of AA shows maximum fluorescence at 440 nm when excited at 360 nm (Supporting Information, Figure S4) wavelength of light. From this study it is clear that boron doping causes much enhancement of the emissive property of the products B1CD, B2CD, B3CD and B4CD with red shifting in emission maximum. This red shift in emission peak is due to the presence of hetero atom, boron. As boron is an electron deficient atom, an electron deficient state (V0) is generated above the actual ground state (S0). Thus the excited electron has to travel a short distance towards the ground state during radiative decay (Supporting Information, Figure S4). Hence the emission peak will be red shifted. Quantum yields of B1CD, B2CD, B3CD and B4CD are 5.45 %, 3.43%, 2.13% and 4% respectively with respect to quinine sulphate prepared in 0.1M H2SO4 solution. Lifetime measurement shows that B1CD, B2CD, B3CD and B4CD have average lifetimes 16.12 ns, 9.27 ns, 7.1 ns and 14.7 ns respectively (Supporting Information, Figure S5). A comparative table (Table 1) has been presented for all these four types of boron doped CDs. However, further investigation is warranted for elaborate explanation. A comparative account for the optical properties of some other boron doped carbon dots have also been provided in Table S1. The as synthesized BCDs are very stable in terms of emissive property. We have studied the fluorescence spectral profile of these CD solutions over two months at room temperature. Virtually no change in emission intensity and emission wavelength is observed, indicating the long term stability of the assynthesized fluorescent particles (Supporting Information, Figure S3).
ACS Paragon Plus Environment
Page 9 of 28
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Langmuir
The chemical or molecular structures of the as-synthesized BCDs could not be defined properly. However FTIR and XPS studies indicate that they possess C-B, C=C, C-O, =C-H, B-O-C, C=O and -O-/-OH bonds. The DFT studies have been performed where the model structures have been considered. The geometry of the model carbon dots (CD) has been constructed with 19 benzenoid fused ring structures, without any symmetry constraints. The boron atoms have been introduced in the CD framework as boron atoms and boron oxides (B2O and B2O3) at the centre of the model CD systems. The terminal carbons are saturated with hydrogen atom and also four OH groups are introduced at the terminal position (Figure 3A). This model qualifies for highest stability. Based on this theoretical model further studies are performed that agree well with the experimental results. Simultaneous theoretical calculations also justify the optical behaviour of the as-synthesized CDs. The optimized geometry of the pure CD, as obtained by the DFT based BP exchange-correlation function, is shown in Figure 3A. The calculated dipole moment, charge on boron atom and oscillator strengths, maximum wavelengths (λ) for CD, boron and boron oxide doped systems are presented in Table 2. The absorption spectrum calculated by TDDFT method for CD with four hydroxyl groups, boron and boron oxide doped CDs is shown in Figure 3B. For the undoped CD at equilibrium geometry, the spectra exhibits a single dominant peak at 340 nm and a small satellite peak at 473 nm, corresponding to the excitation of HOMO-1 and HOMO to LUMO of the complex, respectively. The dominant peak undergoes red shifts systematically from 340 to 918 nm as the boron is introduced gradually in the form of boron atoms and boron oxides into the CD. It is also worthwhile to note that although such red shifts are observed for the case of boron doped compounds, the maximum shift is exhibited by the system with higher amount of boron. For instance, in case of B2CD and B4CD the observed absorption wavelength maximum is 504.7 and 593.6 nm, respectively; similarly for the case of B2O-CD and 2-B2O-CD is 918 and 842 nm and for the case of B2O3-CD and 2B2O3-CD is 553 and 701 nm, respectively. Interestingly, the experimentally observed red shift at higher wavelength is at 460 nm for the case of B1CD and for other cases, the variation in the red-shift is only within 15 nm. The theoretically predicted results qualitatively demonstrate that the increase in concentration of boron causes further shift in the absorption peak. The observed red shift in the absorption wavelength presumably arises due to the introduction of boron atoms in the CD. In case of pure CD, there are sp2 carbon atoms with delocalized electrons within the fused aromatic ring and introduction of any hetero atom, such as boron or boron oxides can induce the formation of surface defects due to the charge
ACS Paragon Plus Environment
Langmuir
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
polarization or creation of possible electron-hole pairs. Since boron complexes are electron deficient species, it is expected that boron atoms/groups can accept electron from the electron rich carbon centres. Evidently, while analyzing the charge population on boron atoms, it can be seen that boron atoms carry a partial negative charge which leads to the formation of positive charge in the framework, leading to the formation of defects on the hexagonal carbon surface which proportionately increases with respect to the boron ion concentration. Although in case of B2O3 case, boron appeared to have positive charge, the oxygen atom of B2O3, being more electronegative than boron, becomes more negatively charged centre which also causes the formation of positively charged carbon surface. The charge polarization caused by the introduction of the dopant hetero atoms is likely to be one of the important factors for the observed red shift in the UV absorption spectrum. In addition, it can also be noticed from Table 2 that the calculated dipole moments of all the compounds substantially increase upon boron doping and it can also be seen that the higher the concentration of boron, higher is the dipole moment. For instance, the change in dipole moment is almost six times more for the boron oxides systems compared to that of pure carbon systems. For the case of merely boron atom doped CD, the increase in dipole moment is not very significant and it changes from 2.497 (CD) to 3.754 (4B-CD) Debye. The highest dipole moment is observed for the case of 2B2O3-CD, 16.07 Debye. The substantially enhanced dipole moments for the case of boron doped systems lucidly demonstrate the existence of the charge polarization which can have a dramatic influence on the UV absorption behavior of these boron doped CDs. It should be noted that the dipole moment values are appeared to be quantitatively very high which may be due to the usage of medium level basis sets. Nevertheless, the observed dipole moment variations can be perceived qualitatively. The summary of the theoretical calculation apparently suggests that the electronic transitions and the associated red-shifts in the absorption spectrum can be correlated with the concentration of boron atoms and these effects are mainly caused by the huge charge polarization within the carbon surface, leading to the formation of surface defects. Here, one point should be mentioned that the feed boron ratio have an effect on the surface and optical properties of BCDs. Although the initial concentrations of boron precursor compounds are same but the atomic % of boron is different in the compounds. Also boron is being doped as boron oxide moiety. The extent of doping increases as the boron content of the precursor compounds is increases. However from XPS and FTIR studies it is found that in all four BCDs C-O, C-B, B-O bonds are commonly present but the optical
ACS Paragon Plus Environment
Page 10 of 28
Page 11 of 28
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Langmuir
properties differ greatly. As optical properties especially fluorescence depends largely on the electronic states of the substrate, the difference in the fluorescence intensity and peak position indicates that the BCDs are in different electronic environment. The BCDs become different due to the varied amount of doped boron that causes different charge polarization on carbon surface leading to surface defects. From the experimental findings we conclude that B1CD is the most effective and pronounced boron doped product amongst the as-synthesized CDs. Henceforth, we choose to use B1CD for further investigation. It is also worth noting that B1CD exhibits highest quantum yield (Φ) with highest boron content. Emission profile of B1CD is found to be pH dependent. In acidic medium, the emission peak shows a red shifting of ~15 nm from 460 nm. Again, with the increase in pH of the medium, blue shift occurs and it approaches towards 460 nm (Supporting Information, Figure S6A). However, intensity of emission does not follow any proper order. From pH =1 the intensity gradually increases till pH =9, then again intensity decreases (Supporting Information, Figure S6B). This shift can be explained by reversible binding of protons to the emissive sites at low pH. At low pH protonation shows blue shift whereas on increase in pH deprotonation causes red shifting.56 However, at high pH, there may be noncovalent molecular interactions through hydrogen bond between – OH groups causing quenching of fluorescence.57 Hence, by varying pH, the fluorescence can be tuned. The experimental pH has been set to 9. B1CD contain oxygenated functional groups at the surface that can bind metals easily. That encouraged us to use B1CD as metal detection probe. Different metal salts [Cd(II), Co(III), Cu(II), Cr(III), Fe(II), Fe(III), Hg(II), Mg(II), Ni(II), Pb(II), Zn(II)] have been added to B1CD solution. The fluorescence has been measured within 30 minutes of metal addition. When Fe(III) is added, the fluorescence is quenched greatly (Figure 4). This quenched solution is termed as Fe(III)B1CD. Gradual quenching of fluorescence of B1CD with the increased concentration of Fe(III) ion paves the way of Fe(III) sensing in aqueous medium. Using this method we can detect Fe(III) down to 3.1 nM level (Figure 4). This is much lower than the permissible limit of Fe(III) in water i.e., 0.3 mg/L or ~ 5.357 µM, as prescribed by World Health Organization. The relative fluorescence intensity varies linearly over two concentration ranges. The correlation coefficient (R2) is 0.989 for 3-1000 nM concentration range, and is 0.985 for 3-300 µM range. Thus we are able to measure the Fe(III) for a wide range of concentrations. Also the interference study indicates that quenching capability of
ACS Paragon Plus Environment
Langmuir
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Fe(III) is not significantly affected when the other cations are present in the solution as interfering agents (Figure 4). Different environmental water samples (drinking water, tap water and river water) have been used to establish the utility of the as synthesized B1CDs. To these water samples certain extent of Fe(III) was spiked using the standard addition method. Different interfering cations of fixed concentrations have been added while the concentration of Fe(III) was varied spiking the water samples. The detected concentration, recovery %, and relative standard deviation (%) of Fe(III) in water samples have been listed in Table S2. The recovery % of Fe(III) ions in these samples agrees well justifying the applicability of the present method. The quenching B1CD solution in the presence of Fe(III) can be explained by considering the facile binding of Fe(III) with the surface groups of B1CD particles. Fe(III) has an inherent and strong affinity towards -O- as reported by Zhang et al.58 The quenching of B1CD solution in the presence of Fe(III) can be explained by considering the facile binding of Fe(III) with the surface groups of B1CD particles (zeta potential is -2.12 mV). It is assumed that upon the addition of Fe(III) ion into a solution of B1CD, there happens a fast electron transfer reaction (Figure 5A). Here electronic interaction of terminal O- with Fe(III) causes a hindrance in the radiative recombination of carbon dots which in turn quenches the fluorescence of B1CD. And also the peak for the quenched fluorescence shifts to 445 nm. Earlier we observed that with the decrease in medium pH, the fluorescence peak shifted towards ~ 450 nm i.e., blue shifting of the original (λem = 460 nm) fluorescence due to binding of proton to the emissive sites of CDs. Blue shift of emission spectra of B1CDs upon the addition of iron can also be explained by the fact that the quencher shortens the excited state lifetime. This provides more weightage to the emission process from only relaxed excited states (which are higher in energy than totally solvent–relaxed states). Smaller size of Fe(III) cation helps in such chelation. TEM image (Figure 5B) shows aggregated form of Fe(III)B1CD. Also we have varied salts of Fe(III) to check the quenching phenomena and it is found that all the salts of Fe(III) exhibit the same quenching effect confirming that Fe(III) cation is the root cause of quenching. Thus the quenching effect of Fe(III) is independent of the counter anion. Figure 5C shows the effect of different Fe(III) salts on B1CD. The lifetime of B1CD and Fe(III)B1CD differ greatly (16.12 ns and 3.16 ns) so also the UV-vis spectra differ for both the species (Figure 5D). Thus there lies a possibility of dynamic quenching.24 The quenching efficiency can be fitted to Stern-Volmer equation,
ACS Paragon Plus Environment
Page 12 of 28
Page 13 of 28
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Langmuir
I0/I=1+Ksv.[Q] where, I0 and I are fluorescence intensities of B1CD without and with Fe(III) ion, respectively, Ksv is Stern-Volmer constant and [Q] is the concentration of quencher, Fe(III). The I0/I vs. [Q] plot (Figure 4) shows linearity over a wide range (3-1000 nM and 3-300 µM with R2= 0.989 and 0.985 respectively) indicating the static quenching. Hence, it is concluded that both static and dynamic quenching occurs during quenching of fluorescence of B1CD. It is worth mentioning that B2CD, B3CD and B4CD show comparatively less pronounced fluorescence quenching in presence of Fe(III) ions and the measure LOD values are 12.9 nM, 15.4 nM and 7.5 nM for B2CD, B3CD and B4CD, respectively (Supporting Information, Figure S7). On the other hand B1CD is chosen as the most effective probe because of its highest sensitivity for Fe(III) quantification with a LOD value of 3.1 nM. It is also seen that undoped carbon dots (CDa, product obtained after MHT of AA) undergo fluorescence quenching in presence of Fe(III). Fe(III) can be sensed in solution using CDa down to 21 nM level with a linear range of detection 1-300 µM level (R2=0.984) (Supporting Information, Figure S8). It is proposed that the quenching is prompted by chelation of Fe(III). In fact, facile chelation occurs due to the presence of -O-/hydroxyl groups at the edges of carbon dots as mentioned earlier. Higher fluorescent intensity for boron doped carbon dots causes much effective and more sensitive detection of Fe(III) in water samples. A comparative table (Table S3) has been accounted to show the efficiency of our prescribed method. However, Fe(II) interferes to some extent for Fe(III) sensing. To eliminate this interference we have serendipitously found an effective reagent, fructose. Upon the addition of fructose to Fe(II)B1CD [Fe(II) induced quenched B1CD solutions] and Fe(III)B1CD individually, two solutions showed different behaviours while fluorescence intensity was measured within 10 min of fructose addition. Interestingly it was observed that fluorescence of Fe(III)B1CD is selectively increased with fructose while that for Fe(II)B1CD solution remained quenched/unaltered (Figure 6A). It is well established that Fe(III) has a tendency to bind with D-fructose. Tonkovic̍ 59 has shown the Fe(III)-fructose complex formation is evident at pH=11. We have observed that as the pH is lowered down (pH=9) there occurs a competitive interaction of Fe(III) with fructose and B1CD, while the latter remains present in solution. At this pH condition (
