Journal of Colloid and Interface Science 437 (2015) 28–34

Contents lists available at ScienceDirect

Journal of Colloid and Interface Science

www.elsevier .com/locate / jc is

Carbon dots functionalized by organosilane with double-sided anchoringfor nanomolar Hg2+ detection

http://dx.doi.org/10.1016/j.jcis.2014.09.0130021-9797/� 2014 Elsevier Inc. All rights reserved.

⇑ Corresponding author at: Queensland Micro- and Nanotechnology Centre,Griffith University Nathan Campus, Brisbane, Australia. Fax: +61 7 555 28226.

E-mail address: qin.li@griffith.edu.au (Q. Li).

Wentai Wang a,b,c, Tak Kim a,b, Zifeng Yan c, Guangshan Zhu a, Ivan Cole d, Nam-Trung Nguyen a, Qin Li a,b,⇑a Queensland Micro- and Nanotechnology Centre, Griffith University, Brisbane, QLD 4111, Australiab Environmental Engineering, Griffith University, Brisbane, QLD 4111, Australiac State Key Laboratory of Heavy Oil Processing Key Laboratory of CNPC, China University of Petroleum, Qingdao, Chinad CSIRO Materials Science and Engineering, Victoria, Australia

a r t i c l e i n f o a b s t r a c t

Article history:Received 13 July 2014Accepted 7 September 2014Available online 16 September 2014

Keywords:Carbon dotsFluorescent nanoparticlesHg2+ detectionWater quality

Surface functional groups on carbon dots (CDs) play a critical role in defining their photoluminescenceproperties and functionalities. A new kind of organosilane-functionalized CDs (OS-CDs) were formedby a low temperature (150 �C) solvothermal synthesis of citric acid in N-(b-aminoethyl)-c-aminopro-pylmethyl-dimethoxysilane (AEAPMS). Uniquely, the as-synthesized OS-CDs have dual long chain func-tional groups with both ANH2 and ASi(OCH3)3 as terminal moieties. Double sided anchoring of AEAPMSon CDs occurs, facilitated by the water produced (and confined at the interface between CDs and solvent)when citric acid condenses into the carbon core. The resultant OS-CDs are multi-solvent dispersible, andmore significantly, they exhibit excellent selectivity and sensitivity to Hg2+ with a linear detection rangeof 0–50 nM and detection limit of 1.35 nM. The sensitivity and selectivity to Hg2+ is preserved in highlycomplex fluids with a detection limit of 1.7 nM in spiked 1 M NaCl solution and a detection limit of 50 nMin municipal wastewater effluent. The results show that the OS-CDs synthesised by the solvothermalmethod in AEAPMS may be used as an effective Hg2+ sensor in practical situations.

� 2014 Elsevier Inc. All rights reserved.

1. Introduction [15,16]. However, the above fluorescence probes possess some dis-

Despite being one of the most toxic heavy metal ions, mercury(Hg2+) ion is widespread and widely used in industry, and causesserious environmental and health concerns [1–4]. With stringentregulations on Hg2+ contamination in drinking water in place(e.g. the maximum contamination limit for Hg2+ is 2 ppb in theUSA [5], 1 ppb in Australia and the European Union [6,7]), detec-tion and remediation of Hg2+ in water has always been a field ofhigh importance. The increased necessity and practice in wastewa-ter recycling in recent years has further increased the urgency ofdeveloping facile and accurate Hg2+ detection methods. Many ana-lytical methods for Hg2+ detection have been developed includingsurface-enhanced Raman scattering (SERS) technique [8], surfaceplasmon resonances [9], inductively coupled plasma mass spec-trometry [10], fluorescence chemosensors [11] and electrochemi-cal methods [12], etc. Most of the Hg2+ fluorescence probes aremetal-based, such as gold and silver nanoparticles and nanowires[13–15]. Organic molecules and semiconductor quantum dotswere also applied as fluorescence probes for Hg2+ detection

advantages, which greatly limit their practical applications, such ashigh cost, toxicity of the probe materials, poor stability and com-plex synthesis procedures. Therefore, new Hg2+ fluorescent probesthat can overcome the above limitations are highly desirable.

Carbon dots (CDs) are a class of carbon-based nanoparticles thatcomprise discrete carbogenic nanoparticles with sizes below10 nm. CDs have emerged as versatile fluorescent nanoparticlespossessing unique features such as high quantum yields [17], non-toxicity, nonblinking, high photostability and vast accessibility[18–20], with strong potential to be applied in bio-imaging, sens-ing and optoelectronic devices [19,21–24]. CDs can be synthesizedthrough a number of methods including laser ablation [25], elec-trochemical exfoliation [26], carrier-supported aqueous route[19], combustion route [27], hot injection [28], hydrothermal[29] and microwave treatment [30] etc. These methods generallyresult in hydrophilic CDs with abundant ACOOH and AOH groupson surface, which are amenable for further functionalization.

Owing to the rich surface functional groups, CDs have beendemonstrated as an effective fluorescence probe for the detectionof copper ions [31], ferric ions [32], silver ions [33], as well as mer-cury ions [34] in water. The presence of the cation analytequenches the CD fluorescence with the fluorescence intensitybeing proportional to the concentration of analytes, most likely

W. Wang et al. / Journal of Colloid and Interface Science 437 (2015) 28–34 29

due to the effect of electron transfer. In terms of Hg2+ sensing, CDspotentially offer advantages such as high sensitivity, economic andgreen synthesis routes, convenient detection procedures amongstothers [35].

However, there is still much room for improvements in order tobring CDs closer to practical applications, such as simpler and moreefficient synthesis methods, tunable emission bands, enhancedquantum yield (QY), heightened sensitivity, specificity and durabil-ity in complex fluids, such as wastewater effluent. Wastewatereffluent contains large amount of organic matters, bacteria andviruses; it has high fluorescence background and is prone to inter-act or contaminate nanoparticle surfaces, representing a challeng-ing sample type for fluorescence probes.

Herein, we report a new kind highly photoluminescent (PL)organosilane functionalized CDs (OS-CDs) for Hg2+ detection,which were synthesized by a low temperature solvothermal treat-ment of citric acid in coordination solvent, N-(b-aminoethyl)-c-aminopropylmethyldimethoxysilane (AEAPMS), in an enclosedsystem. In addition to high QY and excellent stability, the resultantOS-CDs show multi-solvent dispersibility, being dispersable inwater and most of the common organic solvents. Importantly,the as-prepared OS-CDs is an effective fluorescence probe withsuperior selectivity and sensitivity to Hg2+ with a detection limitof 1.35 nM (0.27 ppb) and a linear range of 0–50 nM. Furthermore,the OS-CDs’ sensing ability to Hg2+ in real municipal wastewaterand water of high salinity is preserved, still achieving nanomolarsensitivity. This work may broaden the potential of OS-CDs inpractical use for Hg2+ detection.

2. Material and methods

2.1. Materials

Citric acid anhydrous was purchased from Sigma–Aldrich. N-(b-aminoethyl-c-aminopropylmethyldimethoxysila (AEAPMS) waspurchased from Beijing Shenda Fine Chemical Co., Ltd. Sodiumhydroxide, sodium chloride, potassium chloride, calcium chloride,magnesium sulfate, zinc sulfate heptahydrate, cadmium chloridehydrate, chromium trichloride and sodium hypochlorite were pur-chased from Chem-Supply. Hydrochloride acid, silver nitrate, mer-cury (II) chloride and cobaltous oxalate dehydrate, and solventsincluding dimethylsulfoxide (DMSO), methanol, dimethyl formam-ide (DMF), acetone, ethanol, tetrahydrofuran (THF), toluene andhexane were all purchased from Alfa Aesar. All chemicals andreagents were used as received without any further purification.

2.2. Synthesis of organosilane-functionalized CDs

The typical procedure of solvothermal synthesis of OS-CDs is asfollows: 0.5 g citric acid anhydrous was added into 10 ml AEAPMSwith continuous stirring. The mixture was then transferred into anautoclave with a PTFE inner vessel and placed in oven at 150 �C for4 h. Brownish liquid was obtained after the reaction process. Theproduct was dispersed in Milli-Q water or other appropriate sol-vents, followed by purifying three times with an Al2O3 filled chro-matographic column for removing the residue reactants. Thecollected fraction was further filtered by a 0.22 lm syringe filterto remove the large particles. Finally, the solution was centrifugedfor 30 min at 12,000 rpm for further purification, and the superna-tant was collected as the product.

2.3. Characterization

The hydrodynamic particle size and zeta potential were mea-sured by dynamic light scattering (DLS) on a Malvern Instrument

Zetasizer Nano-ZS. Atomic force microscopy (AFM, Dimension3000) analysis was carried out with tapping mode on a platinumcoated mica substrate. FT-IR spectra were collected on a Perkin-Elmer Spectrum 100 with a resolution of 4 cm�1 in transmissionmode. A baseline correction was applied after the measurement.X-ray photoelectron spectroscopic (XPS) measurements were per-formed on a Kratos Axis Ultra photoelectron spectrometer whichuses Al Ka (1253.6 eV) X-rays. The UV–vis absorption and fluores-cence emission were measured by a Jasco V670 UV–VIS spectrom-eter and a Thermal Scientific Lumina fluorescence spectrometer,respectively. The concentration of CDs was determined by a gravi-metric method (details in Supporting Information).

2.4. Multi-solvent dispersibility test

100 lL of OS-CDs was dropped into 5 ml of various solvents,such as DMSO, methanol, DMF, acetone, ethanol, THF, tolueneand hexane, as well as Milli-Q water, respectively, and mixed uni-formly. The samples were kept at the room temperature for2 weeks.

To observe the transfer of CDs from toluene to water, 100 lL ofOS-CDs was first dispersed into 5 ml toluene, and then 5 ml ofMilli-Q water was slowly added into above solution. An interfacewas clearly observed between water and toluene. The vial wasplaced under a 365 nm UV lamp to observe the movement ofOS-CDs between the organic and water phases.

2.5. Procedures for Hg2+ sensing

Detection of Hg2+ in pure water was performed at room temper-ature. OS-CDs solution with a given concentration was preparedbefore measurement. 2 ml of OS-CDs solution was transferred intoa quartz cuvette followed by addition of calculated amount of Hg2+

solution. After mixing uniformly and incubating for 30 min, the PLemission spectra were collected.

Detection of Hg2+ in municipal wastewater effluent was testedin wastewater effluent which was collected from Redland Waste-water Treatment Plant, Brisbane, Australia, and has gone throughsecondary treatment. The wastewater sample was filtered with a0.22 lm syringe filter firstly to remove large particles. A given con-centration of OS-CDs in wastewater was then prepared. 2 ml ofabove OS-CDs in wastewater solution was transferred to a quartzcuvette for PL measurement. Hg2+ solution was then added intothe vial step-wisely to increase the concentration from 1 nM. Eachtime after adding Hg2+ solution, a 30 min time interval was givento allow a good diffusion-driven mixing in the vial, before the PLmeasurement.

3. Results and discussion

3.1. Synthesis and physiochemical characterization of OS-CDs

The analysis of the OS-CDs morphology by transmission elec-tron microscopy was challenging because the AEAPMS passivatedOS-CDs tend to draw moisture from the air and turn into a gel-typematerial, similar to previous reports [28]. The AFM images of theOS-CDs in Fig. 1A provide the two-dimensional (2D) and 3D mor-phology. The size monodispersity of OS-CDs by counting the heightof 150 particles was shown in Fig. 1B, indicating the size of OS-CDsis mainly distributed in the range of 1–2.5 nm. The DLS data(Fig. 1C) also shows a narrow size distribution in the range of0.5–2 nm. The smaller size derived by DLS can be attributed tothe better dispersion of CDs in water, whilst agglomerationmay have occurred during the drying process of AFM samplepreparation.

Fig. 1. (A) AFM image of OS-CDs. (B) Histogram of OS-CDs particle size. (C) Size distribution of OS-CDs measured by DLS.

30 W. Wang et al. / Journal of Colloid and Interface Science 437 (2015) 28–34

Fig. 2A shows the FT-IR spectra of OS-CDs in comparison withthe reactants, namely citric acid and AEAPMS. It is clear thatC@O stretching vibration of the ACOOH groups appeared at1745 cm�1 after the reaction. The broad absorption between3200 and 3600 cm�1 can be attributed to AOH or NAH. The peaksat 1630, 1565 and 1460 cm�1, which belong to the C@O, NAH andCAH stretching of amide bond, respectively, suggest the formationof RAC@ONR between AEAPMS and carbon core, as illustrated bythe red branches in Scheme 1. It should be highlighted that thereis a distinctive, broad peak between 856 cm�1 and 1140 cm�1

ascribed to SiAOASi and SiAOAC peaks, suggesting the formationof siloxane groups by hydrolization, which may lead to the attach-ment of organosilane long chain onto the carbon core surface, asillustrated by the blue branches in Scheme 1. The vibrational fin-gerprints of CAN (1180, 1250 cm�1) and ANH2 (3300 cm�1)stretching vibration belonging to the amine-terminated longchains were both observed in the spectra of AEAPMS and OS-CDs.

The XPS data shown in Fig. 2B reveals the elementary composi-tion of OS-CDs, namely C 61.17%, N 13.93%, O 12.95% and Si 11.95%,confirming the framework of the OS-CDs are mainly constructed bycarbon. High resolution spectra of C 1s, Si 2p, O1s and N 1s detail-ing the chemical bonding states are shown in Fig. 2C–F, respec-tively. The C 1s core level spectrum (Fig. 2C) can be decomposedinto three contributions at 284.6 eV (46.5%), 285.6 eV(48.5%), and287.8 eV(5.0%), which can be assigned to CAC/C@C bonds, CAN/CAO, and C@O bonds, respectively [36]. The Si 2p spectrum(Fig. 2D) can be deconvoluted into two components, namelySiAOACCDs at 101.9 eV [37] (84.93%), suggesting the covalentattachment of silane end to the carbon core, and SiAOASi at104.9 eV [38] (15.07%) of the silica network formed by hydrolyza-tion of organosilane. The O1s spectrum can be decomposed intofour contributions (Fig. 2E): the peak at 531.2 eV can be ascribedto the ACOO groups (14.12%) [39], peaks at 532.2 eV and533.0 eV are attributed to C@O (68.11%) and CAOH/CAOAC(14.34%), respectively [40]; a small peak at 535.2 eV can beassigned to SiAO (3.44%). In the N 1s spectrum (Fig. 2F), CAN bond(87.34%) at 399.1 eV and ANH2 (12.66%) at 400.0 eV can beidentified.

The zeta potential (f) of OS-CDs in pure water was measured tobe slightly negative, around �2 mV (Fig. S2). This suggests thatthere may be comparable amount of positive charged (such asANH3

+) and negative charged moieties (such as ACOO�, AOH andthe hydrolysed organosilane terminals) on the surface of OS-CDs,as depicted in Scheme 1.

Overall, the chemical analyses have suggested that the solvo-thermal synthesized OS-CDs are constructed by mainly by carbonwith dual long chains attached through RAC@ONR bond originatedfrom amidation as well as SiAOAC and SiAOASi bonding due tosilane hydrolization on the surface of CDs as illustrated inScheme 1. The co-existence of both amine terminated andSi(OCH3)2CH3 terminated long chains on the surface of as-preparedOS-CDs impart them unique properties and functionalities.

It is well known that even small changes of reaction parametersmay lead to different properties of nanomaterials, especially in thesynthesis of CDs [41]. In this work, solvothermal method wasemployed using citric acid as the carbon source and N-(b-amino-ethyl)-c-aminopropylmethyldimethoxysilane (AEAPMS) as thecoordination solvent to produce a highly photoluminescent CDsat a low temperature of 150 �C. It is worth noting that with thesame carbon precursor and same solvent, previous study [28]which used an open reaction system at 240 �C resulted in hydro-phobic CDs. The difference between the two synthesis strategiesis that in solvothermal method the reactor, i.e. the autoclave, isenclosed, therefore, when citric acid condenses into the carboncores, the water produced is trapped in the autoclave under pres-sure. A fraction of the trapped water is very likely to stay at theinterface between carbon nucleates and organosilane solventthrough hydrogen bonds, forming a thin water layer, which facili-tates the formation and retention of these hydrophilic functionalgroups (AOH and ACOOH). Meanwhile, the organosilane mole-cules may be hydrolyzed within the thin water layer and attachonto the surface of CDs through SiAOASi or SiAOAC bonding[42]. The overall formation mechanism is illustrated in Scheme 1.

3.2. The optical properties of OS-CDs

The as-synthesized OS-CDs show a distinctive absorption peakcentred at 360 nm in the UV–Vis absorption spectrum and a max-imum emission peak at 465 nm in the fluorescence (PL) spectra asshown in Fig. 3. In contrast to earlier reported OS-CDs [28], the as-prepared OS-CDs are excitation-independent. When the excitationwavelength was varied in range of 320–420 nm, the emission max-ima remained at 465 nm. Further increasing excitation wavelengthto 440 nm and above, the OS-CDs was almost non-fluorescent,indicating that the OS-CDs have only one fluorescence centre.The excitation-independence may also be ascribed to the narrowparticle size distribution. 360 nm was selected as the excitationwavelength in the following experiment as it induced the highest

Fig. 2. (A) FTIR spectra of citric acid, OS-CDs and AEAPMS. (B) Surface scan of XPS spectra. (C) High resolution spectra of C1s. (D) Si 2p. (E) O1s. (F) N1s.

Scheme 1. Illustration of the possible formation process of OS-CDs from citric acid and AEAPMS in solvothermal synthesis.

W. Wang et al. / Journal of Colloid and Interface Science 437 (2015) 28–34 31

fluorescent intensity. The QY of the as-prepared OS-CDs in Milli-Qwater was calculated to be 51%, higher than most other reportedCDs [28,43].

The fluorescence lifetime of OS-CDs in water under the excita-tion of 360 nm can be fitted by a multi-exponential function asshown in Fig. S3. Two fitting decay times were acquired, namely

s1 = 5.45 ns (5%) and s2 = 15.49 ns (95%) with the average lifetimeof 14.99 ns. This result confirms that the fluorescence of OS-CDswas dominated by the emitting centre with lifetime ofs2 = 15.49 ns. Its long fluorescence lifetime compared to the otherCDs [44] is likely due to the abundant long chain surface functionalgroups which provide better trapping effect.

Fig. 3. UV–Vis absorption spectrum and PL emission spectra of OS-CDs at differentexcitation wavelengths; the inset is a picture of OS-CDs under room light and365 nm UV lamp.

32 W. Wang et al. / Journal of Colloid and Interface Science 437 (2015) 28–34

3.3. Multi-solvent dispersibility of OS-CDs

The OS-CDs can be well dispersed in both polar and apolar sol-vents. Fig. 4A shows that OS-CDs can be well dispersed in DMSO,methanol, DMF, acetone, ethanol, THF, toluene and hexane as wellas Milli-Q water, with no sediment or layering phenomenonobserved after two weeks, which shows the excellent multi-sol-vent solubility and stability. The multi-solvent dispersibility ofOS-CDs was vividly observed by mixing OS-CDs in toluene andwater mixture, as shown in Fig. 4B. Fluorescence disappeared com-pletely in toluene phase but transferred into water phase after1 day, indicating the better solubility of OS-CDs in water than intoluene. The strong multi-solvent dispersibility of the as-synthe-sized OS-CDs further confirms the dual long chain surface chemis-try illustrated in Scheme 1.

Furthermore, the PL intensity of as-synthesized OS-CDs (inFig. S4) does not vary in ionic solutions when ionic strength isincreased by adding NaCl up to 1 M. The excellent dispersity insolutions of high ionic strength can be ascribed to the steric effectdue to the surface long chains [45].

3.4. Hg2+ detection

3.4.1. In pure waterThe as-synthesized OS-CDs fluorescence sensitivity to metal

cations was first assessed against a series of 100 lM aqueous solu-

Fig. 4. (A) The dispersion of OS-CDs in different solvents. (B) The multi-solventdispersibility performance of OS-CDs in toluene and water phase.

tions containing Ag+, K+, Na+, Ca2+, Mg2+, Zn2+, Hg2+, Co2+, Cd2+, Fe3+

and Cr3+. As shown in Fig. 5A, the PL of the OS-CDs was insensitiveto most of the metal cations but Hg2+ and Fe3+. Although Fe3+ alsoinduced the fluorescence quenching of OS-CDs, Hg2+ can be selec-tively identified by adding sodium hexametaphosphate as themasking agent of Fe3+ ions [46] as shown in Fig. S5.

To further verify the effect of the co-presence of some commonmineral elements such as K+, Na+, Ca2+ and Mg2+ on OS-CDs’ sens-ing selectivity to Hg2+, the PL intensities of OS-CDs in pure water,and in solutions containing K+, Na+, Ca2+, Mg2+ and a mixture ofall above (100 lM) with and without Hg2+ were compared asshown in Fig. 5B. The result clearly demonstrates that the presenceof one or more above common metal ions has no adverse impact onthe sensitivity of OS-CDs to Hg2+ ions. In addition, anions such asSO4

2�, NO3�, Cl� and ClO4

� were also found to have no effect on thefluorescence of OS-CDs (Fig. S6 showed data of ClO4

�). It is worthnoting that ClO4

� is highly oxidative and has been shown to havean effect on the PL of graphene quantum dots [47]; the OS-CDsappear to be stable against oxidative conditions. All these lead tothe conclusion that the as-prepared OS-CDs possess excellentselectivity to Hg2+.

The sensitivity of the as-prepared OS-CDs to Hg2+ concentrationin the range of 0–5 lM was evaluated. As shown in Fig. 6A and B,concentration-dependent quenching was observed: with theincrease of Hg2+ concentration from 0 to 5 lM, the PL intensitydecreased gradually. The fluorescence quenching data can be fittedby the Stern–Volmer equation: F0

F � 1 ¼ KSV c, where KSV is theStern–Volmer quenching constant, c the concentration of Hg2+, F0

and F the PL intensity of OS-CDs without Hg2+ and with differentconcentration of Hg2+, respectively. As shown in Fig. 6B, a good lin-ear correlation (R2 = 0.9977) was obtained over the concentrationrange of 0–50 nM, with a quenching constant KSV of 6.49 � 10�3 L/mol. The detection limit was determined to be 1.35 nM, using theequation 3r/m, where r is the relative standard deviation and mthe slope of calibration curve as described in detail in Fig. S7.

The as-prepared OS-CDs by solvothermal method showed a sig-nificantly better sensitivity to Hg2+ than that of the OS-CDs pre-pared by ‘hot injection’ method [48]. As depicted in Scheme 1,the solvothermal synthesized OS-CDs have amine-terminated alkylchain due to the silane reaction with carbon surface with the pres-ence of water layer, in contrast to the OS-CDs by ‘hot injection’. It isknown that the binding affinity between Hg2+ and ANH2/ANHgroups is stronger than that between Hg2+ and hydroxyl or carbox-ylate groups [49]. In addition, the spatial distribution of these ter-minal amino groups and the amine groups in the middle of thealkyl chain is very likely to trap Hg2+ and forms complexations,as illustrated in Scheme 2. Such a binding event would facilitatethe non-radiative electron/hole recombination annihilationthrough an effective electron transfer process, causing fluorescencequenching.

3.4.2. In brine solutionsThe high ionic strength PL stability of OS-CDs is highly desirable

in real applications such as water quality measurement at miningsites. Therefore, we further evaluated the sensitivity of OS-CDs toHg2+ ions in salty solutions (with NaCl concentration of 1 M). Asshown in Fig. S8, the PL of the as-prepared OS-CDs maintains excel-lent response to the concentration of Hg2+ in the range of 0–50 nMwith a good linearity (R2 = 0.9987). The detection limit was deter-mined as 1.7 nM, very close to the result in pure water, confirmingthe stability of OS-CDs in solutions of high ionic strength as a resultof the steric effect.

3.4.3. In municipal wastewater effluentTo further increase the complexity of the sample composition,

wastewater after secondary treatment was employed, which con-

Fig. 5. (A) Effect of metal ions (100 lM) on the fluorescence of OS-CDs. (B) PL intensity of OS-CDs in water, K+, Na+, Ca2+ and Mg2+ aqueous (100 lM) with and without Hg2+

ions, Mixed ions including K+, Na+, Ca2+ and Mg2+, with the concentration of 100 lM.

Fig. 6. (A) PL spectra of OS-CDs in water with different Hg2+ concentrations of 0–5 lM. (B) The linear region (0–50 nM) of Stern–Volmer plot; Inset: The relationship betweenF0/F�1 and Hg2+ concentrations within the range of 0–1 lM. (C) PL spectra of OS-CDs in wastewater with different concentrations of Hg2+ in range of 0–40 lM. (D) The linearregion (0–1 lM) of Stern–Volmer plot; Inset: The relationship between F0/F�1 and Hg2+ concentrations within the range of 0–40 lM.

Scheme 2. Schematic illustration of the Hg(II) complexation on OS-CDs surface.

W. Wang et al. / Journal of Colloid and Interface Science 437 (2015) 28–34 33

tains many kinds of bacteria, viruses, metal ions and organic mol-ecules. The fluorescence quenching of OS-CDs by different concen-tration of Hg2+ in the wastewater sample were investigated. As

shown in Fig. 6C and D, the OS-CDs can still detect the Hg2+ ionsfrom 0.2 lM to 40 lM as shown in Fig. 6C. The Stern–Volmerplot shows the linear range of 0–1 lM with R2 = 0.9978 andKSV = 0.26 L/mol. The detection limit for Hg2+ in wastewater wascalculated to be 50 nM, highly sensitive for wastewater qualitymonitoring. The preserved excellent Hg2+ sensitivity in wastewatereffluent suggests the as-synthesized OS-CDs are largely non-interactive with the microorganisms and organic molecules inwastewater, likely owing to the negative charge, long chain surfacefunctional groups. The slight loss of Hg2+ sensitivity of OS-CDs inwastewater is likely due to the high background fluorescencecaused by the organic molecules in wastewater.

4. Conclusions

In summary, a new kind of OS-CDs was formed by a low tem-perature (150 �C) solvothermal synthesis of citric acid in AEAPMS.

34 W. Wang et al. / Journal of Colloid and Interface Science 437 (2015) 28–34

The as-synthesized OS-CDs have dual long chain functional groupswith both ANH2 and ASi(OCH3)3 as terminal moieties. An enclosedsynthesis system such as the autoclave traps water, a by-product ofcarbon core formation, therefore, enables the organosilane,AEAPMS, to covalently attach on carbon surface by both ends,namely the ANH2 end by amidation and the ASi(OCH3)2CH3 endby hydrolization and SiAOAC formation. The resultant OS-CDsare amphiphilic, and stable in solutions of high ionic strength withQY of 51%. More significantly, they exhibit excellent selectivity andsensitivity to Hg2+ with a linear detection range of 0–50 nM anddetection limit of 1.35 nM. The sensitivity and selectivity to Hg2+

is preserved in highly complex fluids with a detection limit of1.7 nM in spiked 1 M NaCl solution and a detection limit of50 nM in municipal wastewater effluent. Our results demonstratethat by tuning the arrangement of the surface functional groups,the as-prepared OS-CDs can be devised into an effective fluores-cence probe for Hg2+ detection in complex fluid samples.

Acknowledgments

The authors thank Dr. Barry Wood at The University of Queens-land for his assistance in acquiring the XPS data. We also thank Mr.Bradley Taylor at Redland City Council for providing access towastewater samples.

Appendix A. Supplementary material

Supplementary data associated with this article can be found, inthe online version, at http://dx.doi.org/10.1016/j.jcis.2014.09.013.

References

[1] M. Patra, A. Sharma, Bot. Rev. 66 (2000) 379.[2] H.N. Kim, W.X. Ren, J.S. Kim, J. Yoon, Chem. Soc. Rev. 41 (2012) 3210.[3] I. Onyido, A.R. Noris, E. Buncel, Chem. Rev. 104 (2004) 5911.[4] T.W. Clarkson, J. Trace Elem. Exp. Med. 11 (1998) 303.[5] EPA National Primary Drinking Water Standards, 2009.[6] European Drinking Water Directive, 1998.[7] Australian Drinking Water Guidelines, 2011.[8] W. Ren, C. Zhu, E. Wang, Nanoscale 4 (2012) 5902.[9] E.S. Forzani, H. Zhang, W. Chen, N. Tao, Environ. Sci. Technol. 39 (2005) 1257.

[10] D.S. Bushee, Analyst 113 (1988) 1167.[11] L. Qi, Y. Zhao, H. Yuan, K. Bai, Y. Zhao, F. Chen, Y. Dong, Y. Wu, Analyst 137

(2012) 2799.[12] S. Anandhakumar, J. Mathiyarasu, K.L.N. Phani, Anal. Methods 4 (2012) 2486.[13] B.C. Ye, B.C. Yin, Angew. Chem. Int. Ed. Engl. 47 (2008) 8386.[14] A. Widmann, C.M.G. van den Berg, Electroanalysis 17 (2005) 825.

[15] X. Ren, Q.H. Xu, Langmuir 25 (2009) 29.[16] M. Li, Q. Wang, X. Shi, L.A. Hornak, N. Wu, Anal. Chem. 83 (2011) 7061.[17] X. Wang, L. Cao, S.T. Yang, F. Lu, M.J. Meziani, L. Tian, K.W. Sun, M.A. Bloodgood,

Y.P. Sun, Angew. Chem. Int. Ed. Engl. 49 (2010) 5310.[18] Y.P. Sun, B. Zhou, Y. Lin, W. Wang, K.A.S. Fernando, P. Pathak, M.J. Meziani, B.A.

Harruff, X. Wang, H. Wang, P.G. Luo, H. Yang, M.E. Kose, B. Chen, L. MonicaVeca, S. Xie, J. Am. Chem. Soc. 128 (2006) 7756.

[19] R. Liu, D. Wu, S. Liu, K. Koynov, W. Knoll, Q. Li, Angew. Chem. Int. Ed. Engl. 48(2009) 4598.

[20] Q. Li, T.Y. Ohulchanskyy, R. Liu, K. Koynov, D. Wu, A. Best, R. Kumar, A. Bonoiu,P.N. Prasad, J. Phys. Chem. C 114 (2010) 12062.

[21] S. Zhu, Q. Meng, L. Wang, J. Zhang, Y. Song, H. Jin, K. Zhang, H. Sun, H. Wang, B.Yang, Angew. Chem. Int. Ed. Engl. 52 (2013) 3953.

[22] F. Wang, Y.H. Chen, C.Y. Liu, D.G. Ma, Chem. Commun. 47 (2011) 3502.[23] C. Wang, X. Wu, X. Li, W. Wang, L. Wang, M. Gu, Q. Li, J. Mater. Chem. 22 (2012)

15522.[24] T.H. Kim, F. Wang, P. McCormick, L. Wang, C. Brown, Q. Li, J. Lumin. 154 (2014)

1.[25] X. Li, H. Wang, Y. Shimizu, A. Pyatenko, K. Kawaguchi, N. Koshizaki, Chem.

Commun. 47 (2011) 932.[26] Y. Li, Y. Hu, Y. Zhao, G. Shi, L. Deng, Y. Hou, L. Qu, Adv. Mater. 23 (2011) 776.[27] H. Liu, T. Ye, C. Mao, Angew. Chem. Int. Ed. Engl. 46 (2007) 6473.[28] F. Wang, Z. Xie, H. Zhang, C.Y. Liu, Y.G. Zhang, Adv. Funct. Mater. 21 (2011)

1027.[29] S. Sahu, B. Behera, T.K. Maiti, S. Mohapatra, Chem. Commun. 48 (2012) 8835.[30] H. Zhu, X. Wang, Y. Li, Z. Wang, F. Yang, X. Yang, Chem. Commun. (2009) 5118.[31] J. Zong, X. Yang, A. Trinchi, S. Hardin, I. Cole, Y. Zhu, C. Li, T. Muster, G. Wei,

Biosens. Bioelectron. 51 (2014) 330.[32] W. Sun, Y. Du, Y. Wang, J. Lumin. 130 (2010) 1463.[33] H. Li, J. Zhai, X. Sun, Langmuir 27 (2011) 4305.[34] H. Li, J. Zhai, J. Tian, Y. Luo, X. Sun, Biosens. Bioelectron. 26 (2011) 4656.[35] L. Zhang, H. Wei, F. Gao, Y. Su, Y. Su, J. Zhang, Y. Ma, M. Xu, M. Xu, Z. Li, Z. Yang,

Nano-Micro Lett. 5 (2013) 247.[36] M. Algarra, B.B. Campos, K. Radotic, D. Mutavdzic, T. Bandosz, J. Jiménez-

Jiménez, E. Rodriguez-Castellón, J.C.G. Esteves da Silva, J. Mater. Chem. A 2(2014) 8342.

[37] L. Vast, G. Philippin, A. Destrée, N. Moreau, A. Fonseca, J.B. Nagy, J. Delhalle, Z.Mekhalif, Nanotechnology 15 (2004) 781.

[38] H. Tvardauskas, B. Bukauskaite, V Jasutis, D. Lescinskas, I. Simkiene, Semicond.Sci. Technol. 9 (1994) 1633.

[39] A. Schiffrin, A. Riemann, W. Auwarter, Y. Pennec, A. Weber-Bargioni, D. Cvetko,A. Cossaro, A. Morgante, J.V. Barth, Proc. Natl. Acad. Sci. U.S.A. 104 (2007) 5279.

[40] W. Lu, X. Qin, S. Liu, G. Chang, Y. Zhang, Y. Luo, A.M. Asiri, A.O. Al-Youbi, X. Sun,Anal. Chem. 84 (2012) 5351.

[41] W. Wei, C. Xu, L. Wu, J. Wang, J. Ren, X. Qu, Sci. Rep. 4 (2014) 3564.[42] T. Sen, I.J. Bruce, Sci. Rep. 2 (2012) 564.[43] P.C. Chen, Y.N. Chen, P.C. Hsu, C.C. Shih, H.T. Chang, Chem. Commun. 49 (2013)

1639.[44] C.X. Guo, J. Xie, B. Wang, X. Zheng, H.B. Yang, C.M. Li, Sci. Rep. 3 (2013).[45] Q. Li, U. Jonas, X.S. Zhao, M. Kappl, Asia-Pac. J. Chem. Eng. 3 (2008) 255.[46] S. Liu, J. Tian, L. Wang, Y. Zhang, X. Qin, Y. Luo, A.M. Asiri, A.O. Al-Youbi, X. Sun,

Adv. Mater. 24 (2012) 2037.[47] Y. Dong, G. Li, N. Zhou, R. Wang, Y. Chi, G. Chen, Anal. Chem. 84 (2012) 8378.[48] F. Yan, Y. Zou, M. Wang, X. Mu, N. Yang, L. Chen, Sens. Actuators, B: Chem. 192

(2014) 488.[49] N.B. Mojtaba Hadavifar, Habibollah Younesi, Qin Li, Chem. Eng. J. 237 (2014)

217.