Characterization and Analytical Separation of Fluorescent Carbon Nanodots

Carbon nanodots (C-dots) are recently discovered fluorescent carbon nanoparticles with typical sizes of <10 nm. The C-dots have been reported to have excellent photophysical and chemical characteristics. In recent years, the advances in the development and improvement in C-dots synthesis, characterization, and applications are burgeoning. In this review, we introduce the most commonly used techniques for the characterization of C-dots. The characterization techniques for C-dots are briefly classified, described, and illustrated with applied examples. In addition, the analytical separation methods for C-dots (including electrophoresis, chromatography, density gradient centrifugation, differential centrifugation, solvent extraction, and dialysis) are included and discussed according to their analytical characteristics. The review concludes with an outlook towards the future developments in the characterization and the analytical separation of C-dots. The comprehensive overview of the characterization and the analytical separation techniques will safeguard people to use each technique more wisely.


Introduction
During the past few decades, the combination of nanotechnology with numerous disciplines of science have driven the search of new advanced nanoscale materials as a particular interest for scientists [1].In particular, the functionalized carbon nanodots (C-dots), as an emerging photoluminescent (PL) nanomaterial, have gained considerable attention of the scientific communities owing to their superior structural, photophysical, and chemical properties over other materials.In fact, they have been a multidisciplinary researched nanomaterial in recent years.
The discovery of C-dots stretches back to 2004 when Xu et al. [2] performed gel electrophoretic separation of singlewalled carbon nanotubes (SWCNT) synthesized by arcdischarge methods.During the course of their investigation, the suspension was separated into three classes of materials, among which, they observed a fast moving band of nanomaterial that displayed fluorescence under the ultraviolet (UV) light.The fluorescent nanomaterial was further analyzed and fractionated into three components which displayed different fluorescence colors under the UV light.Since this serendipitous discovery, much research efforts have been dedicated to the study of C-dots.
Many scientists have reported the unique properties of C-dots that exceed those of any previously existed materials.They include diameters below 10 nm with discrete, quasi-spherical, nanocrystalline, or amorphous carbon structures; broad excitation wavelengths ( ex ); tunable fluorescence emission; excellent photostability; high quantum yield (QY); high sensitivity and selectivity targeting specific analytes; favorable biocompatibility; low cytotoxity (a promising nontoxic alternative to heavy-metal-based traditional semiconductor quantum dots); and large Stokes shifts [51][52][53].Regarding their surface chemistry, it has been widely described that the C-dots surface is typically attached with many carboxylic acid moieties, which on one hand largely enlarges their potential for further functionalization with small (inorganic/organic) molecules, and on the other hand ideally imparts them with excellent water solubility, facilitating their purification and characterization.
Building upon the great interest in developing novel synthetic methods and searching for new applications for this newly found nanomaterial, one has to fully understand the fundamental properties of C-dots products.Due to the intensive research efforts carried out in the past decades, the scientific community nowadays understands much better the structural, photophysical, and chemical properties of C-dots; however, the complete characterization of C-dots mixture still remains unachieved and faces noteworthy challenge.This is due to the fact that a C-dots product exists as a complex mixture comprising components with various sizes and different surface functionalities, and a polydisperse C-dots sample only represents the summation or average properties of all individual C-dots species [95,100,101].In this sense, there is an urgent need for the development of efficient analytical separation methods to isolate high-purity C-dots fractions for more precious investigation of their unique chemical and photophysical properties.Since the first isolation of C-dots from SWCNT by polyacrylamide gel electrophoresis (PAGE) in 2004 [2], the reports associated with analytical separation techniques of C-dots are emerging.Impressive successes in this area include not only electrophoresis [94,[102][103][104][105][106], but also chromatography [91, 95-97, 100, 101, 107-109], density gradient centrifugation [29], differential centrifugation [98,110], solvent extraction techniques [99,[111][112][113], and dialysis [20,21,87,98,114,115].
In this current review, without pretending to being exhaustive, we for the first time provide a broad vision of characterization and analytical separation techniques for Cdots by means of 190 references published in 2004-2017, leaving aside the synthesis and applications of C-dots which have been comprehensively reviewed by a set of thorough reviews [52,[116][117][118].In addition, some of the merits and drawbacks of the given separation and analytical techniques for C-dots are also highlighted.The main objective of this review is to arouse the interest of the nanoparticles (NP) community in the significance and the importance of characterization and analytical separation of NP so as to truly reveal the NP identity.

Characterization Techniques
The chemical composition, size, shape, and structure are important factors that determine the particular and unique properties of C-dots; therefore, numerous efforts have recently been made to explore the reliable and robust characterization tools for C-dots.In this section, as indicated by Scheme 2, all the past and updated characterization techniques for C-dots are discussed in detail.

Microscopic and Diffraction
Techniques.Different nondestructive imaging and microscopic techniques have been developed for the morphology and size characterization of nanoparticles.For the particle size measurement, microscopy is the only method that can directly observe and measure the individual nanoparticles.The microscopic techniques that have been developed for measurements of C-dots include electron microscopic and diffraction techniques such as scanning electron microscopy (SEM), transmission electron microscopy (TEM), atomic force microscopy (AFM), and Xray diffraction (XRD).

Scanning and Transmission Electron Microscopies.
Both SEM and TEM are useful electron microscopy techniques that provide information about the particle size, size distribution, and morphology of C-dots [83,84,[119][120][121][122].The SEM and TEM images can also be used to investigate whether the agglomeration of particles is present or whether the good   [83], (B) TEM [84], and (C) HRTEM [84] image of a bulk sample of C-dots.dispersion of particles is achieved.SEM produces images by scanning the surface of a C-dots sample with a focused electron beam which interacts with the atoms of C-dots during which the charges are accumulated to form an image.Figure 1(A) shows an example of SEM image of C-dots [83].In cases where the measurements exceed the resolution of SEM (1-20 nm), TEM which affords a much higher resolving power (about 0.2 nm) is favored [123,124].In TEM, a beam of high-energy electrons is applied to obtain images of the sample based on the electron transmission of C-dots.The average diameter of C-dots can be estimated by randomly counting the particle size on the TEM images.Recently, highresolution TEM (HRTEM) which uses both the transmitted and the scattered beams to create an interference image has been extensively and successfully used for analyzing C-dots structures and lattice imperfections.Figures 1(B) and 1(C) display an example of TEM and HRTEM image of C-dots [84], which shows that the C-dots have crystalline structure and the lattice spacing distance was about 0.21 nm, which is close to that of the graphite (100) plane.

Scanning Probe Microscopies.
AFM is a high-resolution scanning probe microscopy which provides dimensional surface images of C-dots at resolution lower than 1 nm [85,[125][126][127][128].Compared to SEM and TEM techniques, AFM not only produces two-dimensional (2D) images of C-dots from which the C-dots dimension can be determined by randomly counting the height of particles on the images but also provides three-dimensional (3D) information about the surface morphology of C-dots.Figures 2(A) and 2(B) show that both 2D and 3D topographic AFM images of C-dots were achieved through the interaction between the AFM cantilever tip and the C-dots under investigation [85].Herein, it is important to emphasize that although these microscopy-based techniques are highly accurate and reliable, the electron microscopy requires elaborate sample preparation and acquiring excellent images using these techniques is really challenging.This is because of the fact that this microscopy-based technique has large potential in leading to  imaging artifacts due to previous sample preparation process (agglomeration of the C-dots sample on the grid) and the vacuum conditions in sample chamber [129].

X-Ray Diffraction.
XRD is a rapid analytical technique used for phase identification and characterization of crystalline materials based on their diffraction patterns.When the X-rays interacts with a crystallite C-dots sample, the constructive interference is produced, and these diffracted X-rays are then detected, processed, and counted, revealing the average structure of C-dots [86,87,89,[130][131][132][133][134][135][136]. Figure 3(A) displays a typical XRD pattern of polyethylenimine (PEI) functionalized C-dots (C-dots-PEI) derived from microwave-assisted pyrolysis of glycerol in the presence of branched PEI [86].In addition, XRD also provides the information about the purity of an as-synthesized C-dots sample.A recent paper [28] reported that XRD can be applied to estimate the purity of C-dots synthesized from glucose by neutralization heating method.As indicated by Figure 3(B), several impurity peaks were observed in the XRD pattern of C-dots due to the incomplete carbonization of glucose.Although the XRD is a valuable characterization tool for obtaining the critical features of C-dots with a crystallite structure, it is not applicable to characterizing amorphous Cdots.

Dynamic Light Scattering. DLS technique has been used
to determine hydrodynamic particle size using liquid phase analytical systems equipped with detectors that respond to C-dots in solution.The average C-dots radius can be determined by measuring the diffusion rate of C-dots in liquids.Although there are several typical examples of using DLS for particle size determination of C-dots [24,69,120,127,136], the application of DLS for the C-dots characterization is not widely utilized.This is due to the fact that the size determination of nanoparticle in a liquid phase using DLS is not reliable since the DLS method provides only a qualitative analysis of the size distribution from the observed photon correlation function [137,138].

Ultraviolet-Visible and Photoluminescence Spectroscopy.
UV-vis and PL spectroscopy are common and widely available spectroscopic techniques that have been used to measure the optical properties of C-dots.Taking together all the information from literatures which characterized C-dots by UV-vis and PL spectroscopy [31,37,41,85,87,103,139], all types of C-dots are active in the UV-vis region of the electromagnetic spectrum, and the fluorescent emission of C-dots shows  ex -dependent behavior.Recently, the timeresolved PL spectroscopy has been reported to be useful in measuring the photoluminescent lifetime of C-dots [37,41,87,103,139].A typical example of UV-vis absorption spectra, PL spectra at different  ex and time-resolved PL spectra of waste frying oil-derived sulfur-doped carbon dots (S-C-dots) [87], is depicted in Figure 4.In addition, it is important to note that the UV-vis and PL spectroscopy can be used together to determine the QY of C-dots.In detail, the QY of a C-dots sample is determined by a comparative method with quinine sulfate with known QY as the reference.Essentially, the PL spectra of solutions of quinine sulfate and C-dots with identical UV absorbance at the same  ex are recorded.Hence, a simple ratio of integrated PL intensity of the two solutions will yield the QY of the C-dots [87].

Infrared Spectroscopy.
IR is a widely used characterization tool to identify functionalities of the solid materials.In the case of IR analysis of C-dots samples, apart from the evaluation of hydroxyl (-OH) and carbonyl (C=O) functional groups on the C-dots surface [41,49,69,127], IR is also able to examine the doping of heteroatoms into the C-dots framework.Important examples include the identification of the presence of amide/amine (-CN/NH 2 ) [15,20,21,[37][38][39][40][41][42][43][44][45][46], alkyl sulfide (C-S) [33,[46][47][48][49], organosiloxane (Si-O-Si/Si-O-C) [89,127,140,141], phosphates (P=O and P-O-R) [38,49,93], and boronic acid (B-O and B-N) [114,142,143] moieties attached on the surface of C-dots, providing evidence for introduction of nitrogen (N), sulfur (S), silicon (Si), phosphorus (P), and boron (B) heteroatoms into Cdots.The merits of this technique for the characterization of surface functionalization of C-dots are being low cost, simple, rapid, and easy for sample preparation.However, the IR cannot give the fine structure information of C-dots, and doping with metal heteroatoms such as aluminum (Al), magnesium (Mg), and nickel (Ni) into C-dots cannot be revealed by IR.Fortunately, other techniques such as XPS provide supplementary information in this aspect, which will be covered in the later part of this review.

Raman Spectroscopy.
RS is one of the commonly used nondestructive and noninvasive spectroscopic techniques that have been used for the identification of the state of carbon of a given C-dots sample.Raman spectra of C-dots usually present two main first-order bands [3,27,32,41,88,129,[144][145][146][147][148][149].The D band represents the vibrations of carbon atoms with dangling bonds in the termination plane of disordered graphite or glassy carbon.The G band (graphitic band) is associated with the vibration of sp 2 -bonded carbon atoms in a 2D hexagonal lattice.A common way to estimate the purity (degree of disorder or graphitization) of a C-dots sample is the ratio of the intensities of the disordered D band and crystalline G band (D/G).The C-dots sample with amorphous nature has a high D/G ratio.High degree of graphitization of the C-dots sample is indicated by a relatively lower D/G ratio.A typical RS spectrum of C-dots from thermal decomposition of Trapa bispinosa peel is shown in Figure 5(A) [32].In this case, the C-dots had both D (1331 cm −1 ) and G (1578 cm −1 ) bands, and the resulting D/G ratio of 0.59 indicates the nanocrystalline graphite structure of C-dots.An additional D  band located at 2654 cm −1 explains the sp 2 hybridization pattern.RS has also been used for size determination of Cdots based on the inverse relationship between the grain size and the D/G ratio.Khanam et al. [88] demonstrated the possibility of using RS technique to determine the grain size of C-dots and C-dots functionalized with hydroxyl and methoxy (F-C-dots) (Figure 5(B)).The relationship between the grain size and the intensity ratio is given by grain size = 44[ (D) / (G) ] (in Å).By using this equation, the grain sizes of C-dots and F-C-dots were determined to be 11 nm and 7.3 nm, respectively.Herein, it should be noted that although RS technique is powerful in the characterization of C-dots, several reports stated the difficulty [150] or failures [52,151,152] in acquiring a high-quality RS of C-dots sample due to the intense fluorescence of C-dots, resulting in covering the characteristics of the Raman signal.

X-Ray Photoelectron Spectroscopy.
XPS is a surface chemical analysis technique that provides information on surface properties, elemental composition, and electronic state of the elements that exist within a material.A XPS spectrum is obtained by irradiating the sample with a beam of X-rays while simultaneously measuring the number of electrons and kinetic energy.For the characterization of Cdots, XPS has generally been used for the assessment of the elemental composition and the chemical bonds of C-dots in combination with IR.Doping of C-dots with nonmetallic heteroatoms such as N [43,49,56,99,114,121,122,134], S [14,24,46,49,87], Si [50,127,140,141], P [39,49,93], and B [114,142] has been characterized by XPS.A typical example of XPS analysis employed to estimate the surface states and chemical composition of nitrogen and sulfur-codoped C-dots (SNCNs) was displayed in Figure 6 [46].In addition, the incorporation of C-dots with metal heteroatoms such as Al [153], Mg [154], and Ni [155] that cannot be detected by IR has also been demonstrated by XPS.Although XPS is very valuable in the characterization of elemental composition and functionalities of C-dots, it does not generally have the spatial resolution to examine individual nanoparticles.
2.2.6.Energy Dispersive Spectroscopy.EDS, which is also called energy dispersive X-ray analysis, is another commonly used technique for the elemental analysis of a sample.In EDS, a high-energy beam of X-rays is focused onto the sample to be analyzed, and an X-ray spectrum is obtained by measuring the number and energy of the X-rays emitted from a specimen with an energy dispersive spectrometer.In general, all of the chemical elements are active in EDS [from beryllium (z = 4) to uranium (z = 92)], and qualitative (from the peak energy) and quantitative (from the peak intensity) analyses can be carried out on the basis of the lines present in the simple X-ray spectra [156].The main application of EDS for C-dots characterization is to analyze the elemental composition of C-dots, for example, the contents of carbon (C), oxygen (O), and other doped elements such as Si [89,137,139].In addition, EDS is able to provide information about the purity of C-dots [32,157,158].Figure 7 displays an  EDS spectra of C-dots passivated with N-(-aminoethyl)-aminopropyl methyldimethoxy silane (AEAPMS) [89].Apart from obtaining the C, O, and Si contents, small impurity peaks of tungsten (W) were found to be coexisting in the assynthesized C-dots sample.structural insights into C-dots.The main application of NMR for the characterization of C-dots aims to precisely identify the surface functionality of C-dots, to determine the element states (e.g., C, N, and P) in C-dots, and to examine the nature of their bonding onto the particle surface.Tian et al. [90] firstly reported the utility of 13 CNMR spectroscopy in liquid state to identify the carbon states in a C-dots sample derived from candle shoot (Figure 8).Followed by this, numerous efforts have been contributed to using NMR spectroscopy in both solid state and liquid state for C-dots characterization [21,38,56,85,109,115,135,143,[159][160][161][162][163][164].In addition, NMR has been proved to be powerful in the establishment of the chemical alterations that happened to the surface modifiers during carbonization.For instance, Khanam et al. [88] applied 1 H NMR characterization to follow the conversion of methoxy group to hydroxyl group on the surface of F-C-dots during the reaction process, and Nandi et al. [165] applied 1 HNMR to confirm the transformation of the glucose residues into elemental carbon and the presence of coating alkyl chains.Although the NMR technique has merits in nondestructive nature and easy sample preparation, it is very expensive, time-consuming, and less sensitive in comparison to mass spectrometric techniques.

Mass Spectrometric Techniques. Mass spectrometry (MS)
has been reported to be powerfully useful in elucidating the chemical structures of smaller-sized nanoparticles (<5 nm) [166].The mass spectrometric techniques that have been applied to characterize C-dots include inductively coupled plasma-mass spectrometry (ICP-MS), matrixassisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS), and electrospray ionization quadrupole time-of-flight tandem mass spectrometry (ESI-Q-TOF-MS/MS).

Inductively Coupled Plasma-Mass Spectrometry. ICP-
MS is a type of mass spectrometry that has been widely used for determination and quantification of inorganic nanoparticles such as gold and nickel particles [167][168][169].However, ICP-MS is not widely used for the measurement of inorganic nanoparticles.For the ICP-MS measurement of C-dots, only Bourlinos et al. [114] reported the use of ICP-MS for the characterization of B-doped C-dots so far.In their study, the exact amount of the B-doped C-dots sample was determined by ICP-MS using boronic acid as the calibration standard.The content of B (3%) was in accordance with the result obtained by XPS.

Matrix-Assisted Laser Desorption/Ionization Time-of-
Flight Mass Spectrometry.MALDI is a soft ionization technique that has been widely used for the structural characterization of NP [170][171][172].TOF analysis is based on accelerating a set of ions to a detector where all of the ions are given the same amount of energy, and the determination of the mass of ions is according to their different flight times.MALDI-TOF MS analysis gives mass accuracy better than 0.1.Based on this aspect, it has recently been used to characterize C-dots.Khanam et al. [88] for the first time applied MALDI-TOF MS to acquire the molecular weight of rare C-dots and F-C-dots.The determination was carried out by using three different matrix materials, namely, sinapinic acid (SA), 2,5-dihydroxy benzoic acid (DHBA), and -cyano-4 hydroxycinnamic acid (HCCA) (Figure 9).The molecular weight obtained with these matrices was the same and the molecular weights for C-dots and F-C-dots were 267 / and 392 /, respectively.Soon afterward, our group reported the utility of MALDI-TOF MS for the characterization of the surface-attached functionalities on Cdots by examining their fragmentation patterns [91,96,108].One typical example is the structural elucidation of C-dots obtained by microwave-assisted pyrolysis of citric acid and 1,2-ethylenediamine [91] (Figure 9).The C-dots mixture was firstly separated by RP-HPLC, after which, the fractions were collected and further characterized by MALDI-TOF MS in positive ionization mode under a pulsed N 2 laser (337 nm).We demonstrated that each of the C-dots fractions shows their unique fragmentation pattern, closely relating to their surface-attached carboxylic acid and amide/amine moieties.The MALDI-TOF MS technique is powerful in identifying the functionalities of C-dots; however, this technique exhibits lower sensitivity with increasing mass and is not capable of capturing the larger ions in high mass range.

Electrospray Ionization Quadrupole Time-of-Flight
Tandem Mass Spectrometry.The electrospray ionization quadrupole time-of-flight tandem mass spectrometry (ESI-Q-TOF-MS/MS) combines the use of soft ionization technique and sensitive MS detection.Till now, our group is the only one that successfully applied ESI-Q-TOF-MS/MS to characterize the chemical composition of C-dots [92,173].By coupling the ultraperformance liquid chromatography (UPLC) with ESI-Q-TOF-MS/MS detection, the MS and MS/MS spectra of each separated C-dots fraction were simultaneously captured (Figure 10).Based on high-accuracy MS and MS/MS analyses, we elucidated the chemical structures and obtained the molecular formulas of each Cdots species simultaneously.In particular, our study revealed for the first time that the C-dots species exist as supramolecular clusters with their individual monomers units linked together through noncovalent bonding forces.

Thermal Techniques. Thermogravimetric analysis (TGA)
is the only thermal technique that has been applied to characterize C-dots [27,32,56,69,88,93,158,174,175].In TGA measurement, the instrument continuously weighs a C-dots sample by coupling with FTIR or MS for gas analysis.A weight loss curve was plotted by measuring the amount and the change rate of the weight loss of a C-dots sample with respect to temperature, providing information about the thermal stability of C-dots.For example, Bandosz et al. [93] reported TGA/MS for the characterization of surface functionality of three kinds of C-dots samples.They acquired the TGA curves under the condition that the Cdots samples were heated up to 1000 ∘ C (10 ∘ C/min) under a constant helium flow (100 mL/min) (Figure 11).The Cdots derived from poly(4-ammonium styrene-sulfonic acid) were determined to have the richest surface groups, especially those decomposing as SO 2 (m/z 64), CO 2 (m/z 44), and CO (m/z 28).Another interesting case is the employment of thermal analysis coupled with FTIR (TGA/FTIR) to record the formation of C-dots from plasma treatment of egg-white [27].The TGA method has advantages in fast load and low sample amount; however, it only provides meaningful data to certain types of C-dots which show changes in mass during measurements.

Analytical Separation Techniques
As mentioned previously, from the viewpoint of better understanding the fundamental properties of C-dots, numerous analytical techniques have been developed for the separation of C-dots obtained from various starting materials.A summary of the separation techniques described in this section is presented in Scheme 3.

Gel Electrophoresis. GE is a fractionation technique
which allows the analytes to be separated according to the different migration behavior of analytes by sieve effects under the influence of an electric field.Xu et al. [2] firstly reported the separation of C-dots by PAGE when they were purifying SWCNT derived from arc-discharge soot.The oxidized candle-soot mixture was separated in a 20% denaturing gel (8 M urea, 1x TBE running buffer) by applying 600 V at 55 ∘ C in an electrophoresis unit.As indicated by Figure 12(A), when observed under UV light, the candle-soot dispersion was separated into three classes of nanomaterial, that is, agglomerates which did not penetrate the gel, slow moving dark bands of short tubular carbons, and fast moving multicolor bands of fluorescent nanomaterials.The multicolor fluorescent bands were further separated into three discrete bands, displaying different colors in order of their elution and increasing size (Figure 12(B)).After separation, the isolated C-dots fractions were passed through a 0.45 m PVDF filter and extensively dialyzed against distilled water for purification and subjected to further characterization by elemental analysis, FTIR, PL  which limits its application in separation of C-dots with a wide range of size.

Capillary Zone Electrophoresis. CZE, with remarkable
separation efficiency, has rapidly emerged as a useful separation approach for NP [94, 104-106, 183, 184] according to their different electrophoretic mobilities (on the basis of their charge/size ratio) through an electrolyte solution contained in a fused silica capillary under the influence of an electric field.The utility of CZE in conjunction with a diode array detector (230 nm) was firstly reported by Baker and Colón [104] for the separation of C-dots obtained from the flame of an oil lamp.In this study, they comprehensively examined the influence of buffer composition on the electrophoretic pattern of the mixture of negatively charged C-dots by varying buffer type, pH, and concentration.Although CE is a powerful technique that is simple to perform and achieves satisfactory separation efficiency, the benefits provided by the high number of theoretical plates obtained with CE for C-dots separation are overshadowed by its low sensitivity of the UV detection systems owning to the small injection volumes of sample [185].To overcome the limitation of the low sensitivity and ensure the accuracy for C-dots analysis, our research group separated C-dots by use of commercial CZE apparatus coupled with a diode array detector This CE method was successfully used to study the reaction time-associated kinetics of C-dots formation and identify the functional groups-associated charge states.Afterwards, our group applied CE coupled with UV detection (250 nm) and LIF detection ( ex / em 488/550 nm) for the separation of hollow C-dots synthesized from glacial acetic acid and diphosphorus pentoxide without external heating [106].The separation was achieved using 40.0 cm capillary (50 m i.d. and 325 m o.d.) with 10 mM SDS and 30 mM phosphate (pH 9.0) as run buffer at an applied voltage of 15 kV.The separated fractions were also collected and further analyzed by UV-vis and PL spectroscopy, MALDI-TOF MS, and TEM.We can see that CZE is certainly useful in identifying the different charge states of C-dots species present in a complex C-dots mixture.However, due to the low sample injection volume of CZE, the collection of enough amounts of individual CEfractionated C-dots fractions is extremely time-consuming.Besides, CE can only separate the charged C-dots species rather than neutral C-dots species which eluted out as one single intense peak.Isoelectric focusing (IEF) is commonly used for the determination of the isoelectric point of macromolecules and bioparticles.The colloidal NP of different sizes can be size-selectively separated from their mixtures in a homemade miniscale IEF unit [186].This method is rapid, highly sensitive, and inexpensive.However, so far, there is no report on the separation of C-dots using IEF.

Chromatographic Techniques.
The chromatographic techniques, previously proven to be highly efficient in the separation of metal-based quantum dots [92,172,173] based on their sizes and chemical properties, have recently been applied to separate and analyze C-dots.Currently, chromatographic methods available for the separation of Cdots are anion-exchange high performance liquid chromatography (AE-HPLC) [95,100,101,107], reversed-phase-(RP-) HPLC [91,96,108], and size exclusion chromatography (SEC) [97,109].

Anion-Exchange High Performance
Liquid Chromatography.AE-HPLC is a separation technique which allows the separation of negatively charged ions or molecules according to their affinity to the ion exchange resin containing positively charged groups.Vinci and Colon [95] firstly used AE-HPLC in conjunction with UV detection at 250 nm and laserinduced photoluminescence detection (LIP) at  ex / em of 325/350 nm to separate C-dots products obtained starting with the soot generated during combustion of paraffin oil in a flame.Solutions of C-dots mixture was fractionated using a strong anion-exchange column and ammonium acetate (NH 4 Ac) as eluent (Figure 14(A)).The fractions of Cdots were collected for further characterization by PL spectroscopy (Figure 14(B)) and TEM (Figure 14(C)).The C-dots were demonstrated to be comprised of C-dots species with different optical and electronic properties.Later, the same group employed the previously well-established AE-HPLC method for the high-resolution separation of another kind of C-dots derived from the oxidation of graphite nanofibers [100,101].Apart from TEM and spectroscopic studies [100], the authors examined the surface composition of the Cdots fractions that exhibit stronger PL by means of XPS and FTIR [101] and identified that the luminescent NP with the higher luminescence possesses less oxygen content.Similarly, Liu et al. [107] applied AE-HPLC with UV detection at 250 nm to fractionate C-dots capped with PEI.Four fractions were collected, characterized by TEM and FTIR, and finally used for separation and preconcentration of Cr(VI).The Cdots species in smaller size behavior served as promising adsorbents for the separation and preconcentration of Cr(VI) and worked more efficiently as signal-enhancing agents for flame atomic absorption spectrometric determination of Cr(VI) in real water samples.

Reversed-Phase
High Performance Liquid Chromatography.RP-HPLC is a separation technique with the polarity of mobile phase higher than that of stationary phase.Compared to CZE, in addition to the preparative scale property of HPLC which allows for the collection of individual fractions of C-dots for more precious study, HPLC using a C18 column can efficiently retain and separate the neutral C-dots species which migrate as one single peak in CZE.Our group [96] firstly proposed the RP-HPLC method for high-resolution separation of C-dots prepared by hydrothermal carbonization of chitosan with UV detection at 300 nm and fluorescence detection at  ex / em of 300/405 nm.The separation was achieved by using C18 column with methanol (MeOH) and Milli-Q water as the mobile phase (Figure 15(A)).In both cases, the electronic and optical properties of C-dots were measured by UV-vis and PL spectroscopy (inset of Figure 15(A)) and TEM (Figure 15(B)).Similar to the observations of Vinci et al. [95,100,101], the as-synthesized C-dots samples were demonstrated to be comprised of numerous C-dots species, displaying their unique optical and electronic properties.Additionally, of particular importance, the MALDI-TOF-MS was applied to elucidate the surface-attached functionalities of C-dots fractions by examining their fragmentation patterns (Figure 15(C)).Later, we adopted RP-HPLC coupled with fluorescence detection (340/440 nm) for the fractionation of C-dots prepared from citric acid and 1,2-ethylenediamine using 10 mM NH 4 Ac (pH 4.5) and MeOH as the mobile phase [91].Our group further reported the separation of hollow C-dots by the above established RP-HPLC method using UV detection (300 nm) with a binary solvent mixture of MeOH and 10 mM NH 4 Ac buffer (pH 5.5) as the mobile phase [108].The fractions collected were found to be multicolor emissive, displaying red, green, and blue fluorescence under UV irradiation (365 nm).Selected fractions with strong PL were proved to be in low cytotoxicity by MTT assay and, therefore, finally used for living cell imaging.Recently, we employed RP-UPLC coupled with either UV detection [187] or MS [92,173] detection for the separation and analyses of different types of C-dots.A typical example is the utilization of ultraperformance liquid chromatography (UPLC) coupled with ESI-Q-TOF-MS/MS to separate and characterize C-dots derived from microwaveassisted pyrolysis of citric acid and 1,2-ethylenediamine [173].By using UPLC, the C-dots product was well separated into ten fractions within 4.0 min.By using ESI-Q-TOF-MS/MS, the characterization of the chemical structures and the elucidation of the molecular formulas of C-dots species were achieved simultaneously.

Size Exclusion
Chromatography.SEC, also known as gel filtration chromatography, is a size-based separation process with a column with flow channels formed by packing with porous materials.The separation of particles is achieved by their differences in flow velocity.The particles with smaller or equal size of pore in packing materials permeate deeply into the column while larger particles that diffuse freely through it are excluded, resulting in longer retention time for the smaller particles.Recently, SEC has also been used for the separation of C-dots based on their size differences.For example, Wang et al. [109] simply fractionated C-dots passivated with PEG1500N by an aqueous gel column packed with commercially supplied Sephadex G-100 gel with water as the eluent.The fractions were collected and further characterized by TEM, AFM, 13 CNMR, and FTIR analysis.Jiang et al. [97] isolated fluorescent C-dots products from Nescafe5 Original instant coffee powder by Sephadex G-25 SEC (Figure 16).In this interesting case, the coffee powder was dissolved in distilled water at 90 ∘ C, vigorously stirred to mix, and centrifuged at 14,000 rpm for 15 min.The resulting supernatant was filtered through a 0.22 m-membrane to remove the large or agglomerated particles and then loaded onto the gel column with distilled water as eluent.The fluorescent C-dots fractions were collected and applied for cell imaging.As indicated by Figure 16, the obtained C-dots fractions display multiple colors when excited at different  ex .

Density Gradient Centrifugation.
The density gradient centrifugation method which is a routine technique used to separate biomacromolecules also exhibits great potential in isolating NP [188].Briefly, samples are added on the top of a density gradient formed by sequentially layering solutions of different densities.Upon centrifugation, the particles deposit in the density gradient at different speeds on the basis of their sizes, shapes, or densities and eventually form different hands in the gradient.Usually, the density gradient is prepared with sucrose, glycerol, cesium chloride, and other aqueous solutions.The advantage of this technique is its nontoxicity.However, for NP prepared in organic medium, transferring from organic solvent to aqueous solution is challenging  due to the coagulation possibility of some particles [189].Fortunately, this is not a big problem for C-dots sample which is usually synthesized in aqueous medium and possesses excellent water solubility.Pandey et al. [29] proposed the sucrose density gradient centrifugation (SDGC) method to separate C-dots from other carbonaceous materials.The sample mixture was fractionated in a density gradient formed by overlaying 50-100% of pure sucrose in test tube.Figure 17 shows that fine separation can be achieved based on the size of C-dots species.The obtained fractions exhibited green and blue fluorescence under UV light.

Differential Centrifugation.
Differential centrifugation is used to separate the multiple fractions within a NP sample by size and density; the larger and denser NP precipitate at lower centrifugal forces while the smaller and lighter NP remain in the supernatant and require more force and greater times to precipitate [98].In a typical differential centrifugation process, a NP mixture is centrifuged multiple times, and after each run the precipitate is removed while the supernatant is centrifuged at a higher centrifugal force.Sahu et al. [98] for the first time reported the use of differential centrifugation method to obtain different size-range C-dots species synthesized by hydrothermal treatment of orange juice (Figure 18).In the centrifugation procedure, the aqueous brown solution containing C-dots was firstly centrifuged at 3000 rpm for 15 min to get the less-fluorescent deposit coarse nanoparticles (CP).Then, an excess amount of acetone was added to the upper brown solution and centrifuged at a high speed of 10,000 rpm for 15 min.The supernatant containing the highly fluorescent nanoparticles (CD) has an average size of 1.5-4.5 nm.The deposit CP exhibiting weak fluorescence have larger size in the range of 30-50 nm.

Solvent Extraction Process.
Solvent extraction, also known as liquid-liquid extraction or partitioning, is a method used to separate a compound based on its solubility in two liquids that do not mix (e.g., water and an organic solvent).This method has been widely used in the isolation  of C-dots from the reaction mixtures.Typically, organic solvents (e.g., ethyl acetate, acetone, chloroform, toluene, and n-hexane) are used to extract the amphiphilic C-dots and eliminate unreacted reagents in aqueous solution [111][112][113].Recently, Han et al. [99] proposed a novel "gradient extraction" method to separate C-dots synthesized from hydrothermal treatment of cow milk based on their surface polarity of the C-dots species (Figure 19).Four organic solvents with different polarities: hexane (polarity: 0.06), carbon tetrachloride (polarity: 1.6), mixture of carbon tetrachloride and dichloromethane (v/v = 3 : 2), and dichloromethane (polarity: 3.4), were used to extract C-dots fractions.The four fractions obtained were determined to have different surface polarities and show surface polarity-dependent photoluminescence.The solvent extraction process is certainly good in roughly isolating C-dots from the reaction mixtures; however, it does not possess the ability to completely separate and exactly examine the hidden properties of each individual C-dots species.
3.6.Dialysis.Dialysis is a widely used purification method to separate the C-dots from the potential small fluorescent molecules that can be byproducts of the synthetic route [20,21,87,98,114,115].The dialysis process works on the principles of the diffusion of small fluorescent molecules across a semipermeable membrane based on the principle that the substances in water tend to move from an area of high concentration to an area of low concentration.Usually, the as-synthesized fluorescent product was dissolved in distilled water and dialyzed against a semipermeable membrane in purified water with stirring and recharging with fresh distilled water at different time intervals over a certain time period.The purification by dialysis is certainly very useful to remove the potential small fluorescent species from C-dots sample.However, similar to the solvent extraction process, it is not capable of efficiently separating each individual C-dots species.

Conclusion and Outlook
In this review, we outline the general approaches for the characterization and analytical separation of C-dots.Given the complexity of C-dots components and the inherent drawbacks of each technique, only one technique is not enough to capture the complete picture of the C-dots properties.For most of time, multiple characterization and analytical separation techniques described here are used in complementary ways.Though significant progress has been made in recent years, we are still far from achieving full understanding of C-dots properties.Bearing this in mind, much effort is required.On one hand, despite the size and chemical structure of C-dots having been estimated by a number of researchers, the exact molecular weight of Cdots still remains debated, and, therefore, developing new methods with improved resolution and sensitivity based on new or original concepts is essential to progress the characterization of C-dots.On the other hand, carefully checking and refining the existing methods to overcome the limitations and to further improve their working efficiency is of equal importance.While any tiny improvement may not significantly matter, the cumulative effects of a series of such advances will make a substantial change.Furthermore, among the numerous techniques applied to separate or to characterize C-dots, efforts have been seldom used in coupled ways.It is anticipated that the coupled techniques could be further explored in the near future to achieve integrative separation-characterization of C-dots in an efficient way.

Scheme 2 :
Scheme 2: Schematic illustration of the characterization techniques for C-dots.
Figure 2(C) depicts the height profiles along lines I, II, and III in the 2D AFM image (Figure 2(A)) of C-dots.

Figure 2 :
Figure 2: (A) AFM topography image of a bulk sample of C-dots on quartz substrate.(B) AFM 3D image of (A).(C) Height profiles along lines I, II, and III in (A) [85].

Figure 4 :
Figure 4: (A) UV-vis absorption spectra, (B) PL spectra at different  ex , and (C) time-resolved PL spectra of S-C-dots derived from waste frying oil.Insets in (A): photographs of the aqueous solution of the C-dots under daylight (left) and UV light (right).Inset in (B): the normalized PL spectra at different  ex of the S-C-dots [87].

Figure 10 :
Figure 10: Representative (A) MS and (B) MS/MS spectra from ESI-MS analysis of fraction 2. Shown in the insets are the MS spectra of the same fraction in the higher m/z ranges [92].

Figure 12 :
Figure 12: (A) Electrophoretic profile of C-dots in 1% agarose gel under 365 nm UV light.(Wells A-C are crude SWNT suspension, fluorescent carbon, and short tubular carbon, resp.Wells D and E are the further separation of fraction C. Well F is the cut SWNT).(B) Photographic images of the 3 fractions of fluorescent C-dots species under 365 nm [2].

Figure 14 :
Figure 14: (A) AE-HPLC chromatograms of the soot-derived C-dots sample monitored by UV detection at 250 nm and LIP detection at  ex / em of 325/350 nm.(B) Photographic images of the 29 separated C-dots fractions under a hand-held UV lamp.(C) TEM images and electron diffraction patterns of selective C-dots fractions (images A-C are attributed to fractions 3, 9, and 28, resp.)[95].

Figure 15 :
Figure 15: (A) Absorption chromatograms of (a) chitosan and acetic acid in MeOH/water (1 : 19 v/v) and (b) C-dots solution monitored at 300 nm.(c) Fluorescence chromatogram of C-dots solution monitored at  ex / em of 300/405 nm.The inset displays the absorption spectra of the 12 separated C-dots fractions.(B) and (C) are the TEM image and MALDI-TOF MS spectrum of fraction 1, respectively [96].

Figure 16 :
Figure 16: Schematic illustration of the preparation of fluorescent C-dots from Nescafe Original instant coffee by SEC technique and its application for cell imaging [97].

Figure 17 :
Figure 17: (A) SDGC separation of C-dots from sonication of sugar cane juice.(B) The images of the separated C-dots bands under ambient and UV light [29].

Figure 18 :
Figure 18: Schematic illustration of the formation and the differential centrifugation of C-dots from hydrothermal treatment of orange juice [98].