Colloidal semiconductor nanocrystals (also called quantum dots,-QDs) have attracted much attention due to their unique optical and electronic properties. These properties include their tunable fluorescence and absorption spectra which can be attributed to quantum confinement. These properties have led to QDs being widely investigated in applications such as drug delivery [1][2] , fluorescent markers for bioimaging [3][4][5] , optical-electronic devices [6][7] , immunoassays [8][9][10] , and fluorescence-based sensing 4,9,[11][12][13][14][15][16][17][18][19][20][21][22][23][24][25][26][27][28] .
In recent years, QD-based fluorescence sensing has been targeted for the detection and measurement of biochemical conditions, pharmacological substances, and environmental pollutants with potential advantages in economy 27 , sensitivity, and simplicity. The general schemes for fluorescent detection using QDs include 'turn-on' mechanisms, where the fluorescence intensity is enhanced by the addition of analytes, and 'Turn-off' mechanisms, in which quenching of fluorophores occurs in response to an analyte via direct quenching, photoinduced electron transfer (PET), and/or fluorescence resonance energy transfer (FRET) processes. Among QD compositions, there are many types that have been used for fluorescence sensing applications to date, notably including : CdSe 15,21 , CdSe/ZnS 22,28 , CdTe 17,19,25,29 , CdTe/ZnS 19 , metal modified CdS 23 , ZnSe/ZnS 26 , CuInS2 30 , Si QDs 3-4, 9, 14, 24 .
CdSe QDs are widely used as fluorescence probes due to their high photostability and tunability across the visible spectrum. CdSe QDs can be prepared mainly from two solvent systems: high boiling point organic solvents with coordinating surfactants, and aqueous-based systems. CdSe QDs that are prepared at high-temperature organic solvents usually exhibit superior fluorescent properties, but often times a surface modification process is required to allow the QDs to be dispersible in water for further uses. In contrast, CdSe QDs prepared from water-based systems are readily water-dispersible when hydrophilic capping agents are used during the synthesis and can be natively attached to protect the QDs surface 31 . Some common stabilizing agents are 3-mercaptopropionic acid (MPA), Lglutathione (GSH), thioglycolic acid (TGA), and N-acetyl-L-cysteine (NAC). There have been many reports on the CdSe QDs that are prepared using water-based systems as it is relatively convenient in handling, but low photostability and sensitivity to variations in pH and ionic strength are of some concern.
In a similar vein, Si QDs have also been reported to be used in fluorescence sensing for the detection of various analyzes such as explosive molecules 32 , organic compounds 33 , alkaline phosphate 24 , ethyl carbonate 9 , glucose 34 , and dyes 14 due to their high photostability, biocompatibility, low cytotoxicity, and surface paint-ability. Moreover, the preparation method of aqueous-based Si QDs is simple, low-cost, and non-toxic. To fabricate Si QDs-based fluorescence sensors, other molecules that are photoactive are usually required because Si QDs themselves have very high photostability and hardly respond to the analytes. There are a few reports using the organic molecules (in complementary to Si QDs) with aza crown 4 and azo group 14 to complete the electron transfer process, but to the best of our knowledge, there is no report on using Si QDs with other types of quantum dots. Furthermore, the emission energy of Si QDs is in the blue emission range which enables them to serve as a good energy donor for the other visible-emitting fluorophores in the photoinduced electron transfer (PET) and fluorescence resonance energy transfer (FRET) processes. In this work, we made use of the high photostability of blue-emitting Si QDs to design a new fluorescent probe.
Copper, while required as a nutrient, is identified as one of the toxic metals in water that can negatively affect animals, plants, and humans. An excess amount of copper in the human body can cause the risk of many diseases, such as Alzheimer's disease, prion disease, Wilson's disease and Parkinson's disease 27,[35][36][37] . The traditional methods for the detection of copper includes the use of inductively coupled plasma-mass spectroscopy (ICP-MS), inductively coupled plasma -optical emission spectrometry (ICP-OES), and atomic absorption spectroscopy (AAS). However, these techniques are expensive, time-consuming, and required skillful users. Therefore, the development of copper detection probes has received a big interest to be the simple, sensitive, fast, accurate, and low detection limit sensing probe. In recent years, the quantum dots-based fluorescence sensor has been widely studied for Cu 2+ detection due to their high photostability 19,21,23,[25][26][27][28] . The detection of Cu 2+ by fluorometric detection of QDs has been reported. Unfortunately, other metals ions such as Hg 2+ and Ag + also leads to the quenching of CdSe/ZnS QDs in the Cu 2+ sensing system by the fluorometric detection of CdSe/ZnS. 28 This makes accurate determination of Cu 2+ in any system more difficult. In addition, Mn 2+ also "turn on" the fluorescence of Si QDs similar to Cu 2+ thus, these Si QDs were used as a dual detection probe 27 .
On the other hand, ratiometric fluorescence probes have been reported and showed significant improvement from the intensity-based systems for many reasons 38 . The ratiometric system, in which the ratio of fluorescence measured at two different emission and/or excitation wavelengths composes the signal, is self-correcting, so that interference from the probe concentration, media, and instrumental factors such as excitation power are eliminated. As a result, a separate internal standard is not required for the dual photoluminescence system, and complicated and high interference specimens such as biological samples can be readily analyzed. Also, naked-eye detection is promising for the ratiometric systems as the probes can give obvious changes of perceived emission color on exposure to the targeted condition or analyte. A ratiometric fluorescence probe for copper (II) ion detection has been reported by using CdTe QDs and metal-organic frameworks (MOFs) 17 . This ratiometric detection of CdSe/MOFs can overcome the effects of many interferences including Mn 2+ and Ag + , but the fluorescent signals can also be quenched by Hg 2+ ions. Therefore, emitter combinations are required for ratiometric probes that have lower undesired quenching from interfering ions, and high photostability.
In this work, we report a new fluorescent probe consisting of mixed-QDs: Si QDs and CdSe QDs for sensitive and selective copper(II) ion detection in an aqueous solution. X-ray photoelectron spectroscopy (XPS) was measured to understand the quenching mechanism of the CdSe QDs which plays the dominant role in the mixed-QDs probe to detect Cu 2+ . Blue-emitting Si QDs were selected as the FRET donor and yellow-emitting CdSe QDs as FRET acceptor. Naked eye detection of the analytes was done under a UV lamp by observing the change in emission colors. Our result opens door to the application of mixed Si QDs/CdSe QDs for naked eye Cu 2+ detection in various analytes.
All chemicals were used without any furthur purification. Trisodium citrate dehydrate 99%, 3-(aminopropyl)trimethoxylsilane 97% (APTES) 3-mercatopropionic acid 98% (MPA), Hydrazine hydrate 99% (N2H4 • H2O), sodium hydroxide, (NaOH), N-[2-hydroxylethyl] piperazine-N'-[2-ethane-sulfonic acid] 99+ % (HEPES) were purchased from Sigma Aldrich. Cadmium chloride anhydrous, CdCl2 > 99%, was from Fluka. Sodium selenite anhydrous 99% (Na2SeO3) was from Alfa Aesar. All the metal salts were analytical grade and purchased from Ajax Finechem, Thermo Fisher Scientific. Ultrapure water is used throughout (Milli-Q pure system, Millipore).
Absorption spectra were recorded using a HP 8453 UV-VIS spectrophotometer. Fluorescence spectra measurements were done with a Cary Eclipse Fluorescence spectrophotometer at excitation = 350 nm. PerkinElmer, EnSight multimode microplate reader was used to analyze fluorescence intensity in selectivity study. Fluorescence lifetime of QDs were measured using Horiba DeltaFlex spectrophotometer Field emission transmission electron microscopic (FETEM) images of the quantum dots were captured using field emission transmission electron microscope: JEOL, JEM-3100F after depositing them on carbon-coated Cu grids. The elemental composition analysis was carried out by Xray photoelectron spectrometer, XPS; AXIS ULTRADLD, Kratos analytical, Manchester UK. Fourier Transform Infrared, FTIR spectra were recorded by Thermo Scientific, Nicolet iS50 FTIR.
The synthesis of Si QDs was based on a reported method 3 with some modifications. Briefly, Si QDs were synthesized by the hydrothermal method. 1 g of tri-sodium citrate was dissolved in Milli Q water 21.5 mL and run through nitrogen gas for 15 min. Then 5.36 mL APTES was added into the solution. Finally, the mixture was transferred into a 100 mL Teflonlined stainless-steel autoclave and heated at a constant temperature of 200 ºC for 4 h. Si QDs were purified by washing and centrifugation in ethanol several times. Si QDs were redispersed in Milli Q water and kept in the dark at 4 ºC for further use.
The synthesis of CdSe QDs was based on a reported method 15 . Typically, 0.80 mL of 0.20 M CdCl2 solution was diluted in 50 mL of Milli Q water in the flask with a vigorous stir. Then, 34.6 µL of MPA was added into the solution followed by sufficient 1.0 M NaOH to adjust the pH to 9.0. Then 0.80 mL of 0.02 M Na2SeO3 was injected into the flask. After the solution was refluxed at 100 C for 5 min, 3.67 mL of N2H4 • H2O was loaded into the solution. The yellow CdSe QDs solution was obtained after 12 h of refluxing at 100 C under air. Assynthesized QDs were purified by washing with ethanol and centrifugation. Lastly, CdSe QDs were redispersed in Milli Q water and kept in the dark at 4 ºC for further use.
300 µL of Si QDs (absorption value = 0.2, λabs = 350 nm) and 60 µL of CdSe QDs (absorption value = 0.15, λabs = 350 nm) were dissolved in 10 mM pH 7.0 HEPES buffer under the vigorous stirring. The final volume was adjusted to 3 mL to obtain the mixed-QDs probe solution for the further study.
The optical properties of QDs were characterized by UV-VIS and fluorescence spectrometry. The sizes and planes of the dots were measured using the field emission transmission electron microscope (FETEM). The functional group of the dots was obtained by Fourier Transform Infrared (FTIR) technique.
Quenching of Si and CdSe QDs by copper (II) ion was studied by fluorescent spectroscopy. 300 µL of Si QDs (in milli-Q water, absorption value = 0.2, λabs = 350 nm) and 60 µL of CdSe QDs (in milli-Q water, absorption value = 0.15, λabs = 350 nm) were transferred to the vial. Then the total volume was brought to 3 mL by the addition of 10 mM pH 7.0 HEPES buffer Then, a slight amount of Cu 2+ ions from 10 mM Cu(NO3)2 was injected into the solution to give the final concentration of free copper ion as 0-16 µM.
The detection of Cu 2+ by the mixed-QDs probe was studied by the fluorescence titration between mixed QDs-probe and Cu (II) ion solution at various concentrations from 0 nM -200 nM.
PL lifetime analyses were conducted at room temperature using a Horiba DeltaFlex spectrophotometer with a 375nm pulsed laser diode excitation. Decay lifetimes were collected through a monochromator centered at either 485nm for Si QDs or 575nm for CdSe QDs with an emission bandwidth of 6nm and a time range of 100ns.
In order to test the selectivity of the sensor for Cu 2+ , the response to various metal ions including Co 2+ , Fe 3+ , Zn 2+ , Al 3+ , Mg 2+ , Cr 3+ , Ba 2+ , Li + , Ca 2+ , Sr 2+ , Ag + , Na + , Ni 2+ , K + , Cd 2+ , Pb 2+ , Mn 2+ , and Hg 2+ , was detected using the procedure below.
Selectivity of Si QDs, CdSe QDs, and mixed-QDs probe were studied. Separately Si QDs, CdSe QDs, and mixed-QDs probe with and without Cu 2+ were added in 96 well plate. Then 6 µL 50 µM Cu(NO3)2 solution, or one of the other metals were added into each well, and the final volume of each well was adjusted to 300 µL by 10 mM HEPES buffer pH 7.0. The concentration of all metals was set at 1 µM. The plate was shaken automatically on the microplate reader for 15 min at 100 rpm before the fluorescence measurement.
Si QDs was synthesized based on the previous report 3,24 with modification in the heating system from using a microwave oven to a hydrothermal method under a Teflon-lined stainless-steel autoclave at 200 C. After 4 h, Si QDs were purified by ethanol. APTES was used as a silicon source for Si QDs with citrate as a stabilizing ligand. Blue-emitting Si QDs were obtained after the reaction and purification process. The UV-VIS absorption band and fluorescent spectra of Si QDs were illustrated in Figure 1. The characteristic first absorption peak of Si QDs at around 347 nm and the symmetric emission spectra at 430 nm corresponding to the blue emission color (Figure 4b) were obtained. In FTIR spectra, several distinct transmittance peaks appear in the range of 1000-3500 cm -1 . Typically, the broad absorption peaks at ∼ 3120-3400 cm -1 is assigned to O-H stretching vibration. The absorption peak at ∼ 1580-1650 cm -1 can be attributed to C=O stretching vibration indicated the production of -COOH and the absorbance at ∼1390-1400 cm -1 is corresponding to C-O stretching vibration. Most importantly, the strong absorption peak at ∼1100 cm -1 is ascribed to the Si-O bonding vibration stretching, which proved that Si QDs were successfully synthesized. Moreover, FETEM and SAED images in Figure 1c demonstrated the size, morphology, and crystal lattice of quantum dots. The average diameter of Si QDs is 2.50 ± 0.30 nm with spherical particles and well dispersion. Additionally, the SAED image of Si QDs demonstrates the crystal lattice related to (111), (220), and (211) lattice planes with the d spacing equal to 3.14, 1.01, and 1.59 Å, respectively. The elemental analysis of these Si QDs by using energy-dispersive X-ray spectroscopy (EDS) proves that these Si QDs containing Si, C, O, and N. X-ray photoelectron spectroscopy (XPS) was used to study the surface composition of Si QDs. Full-scan XPS spectra of Si QDs, showed five distinct peaks at 1070 eV, 532 eV, 399 eV, 285 eV, and 102 eV which are related to Na 1s, O 1s, N 1s, C1s, and Si 2p, respectively, as shown in Figure 2a.
CdSe QDs were synthesized using hydrothermal method under the air atmosphere at 100 C for 12 h in an aqueous solution without any shell coating based on the previous report 15 . Hydrazine was used as a reducing agent to reduce Se 4+ to Se 2-without an oxygen-free atmosphere, and MPA was used as a stabilizing ligand to stabilize CdSe in an aqueous solution. A yellow-emitting CdSe QDs was obtained as shown in Figure 4. The obtained CdSe QDs were characterized by UV-VIS, fluorescence spectroscopy, FETEM, FTIR, and XPS measurement. Figure 1a shows the optical properties of CdSe QDs with the first absorption at around 549 nm and emission at 573 nm with a symmetric peak. The FETEM images of CdSe QDs showed well dispersion of spherical particles with an average diameter of 3.82 ± 0.77 nm. The SAED image illustrated the d-spacing of CdSe of 3.50, 2.14, 1.83 Å corresponding to (111), (220), and (311) lattice planes (Figure 1d.) The elemental analysis of these CdSe QDs using EDS represents the compositions of C, O, S, Se, and Cd. The FTIR spectra of CdSe QDs showed the transmittance peak at ∼ 3120-3450, ∼ 1586, ∼ 1400 cm -1 , which were described to O-H stretching, C=O stretching vibration, and C-O stretching vibration, respectively. FTIR spectra demonstrated that CdSe QDs and Si QDs have many hydroxyl groups, Figure 1b. Full-scan XPS spectra of CdSe QDs (Figure 2b) showed six distinct peaks at 1072 eV, 532 eV, 405 eV, 295 eV, 168 eV and 54 eV which were corresponding to Na 1s, O 1s, Cd 3d, C 1s, S 2p and Se 3d, respectively. It is confirming the existence of Cd, Se, S in CdSe QDs. The Stern-Volmer plot comparing the effect of Cu 2+ ions to QDs in the concentration range from 0.2-16 µM in Figure 3, indicating the clearly quenched of CdSe QDs in the presence of free copper (II) ions. Commonly, the quenching of fluorophore takes place through the static and/or dynamic interaction between fluorophore and quencher 26 . The deviations were observed at the higher concentration of Cu 2+ which is probably related to the simultaneous existence of both statistic and dynamic quenching. 29 The surface information of water dispersible CdSe QDs and Si QDs was investigated by X-ray photoelectron spectroscopy (XPS) analysis. XPS survey spectra of CdSe QDs and Si QDs were compared in the presence and absence of Cu 2+ ions as shown in Figure 2a,d. The peak of Cu 2p was obviously observed at 932 eV for CdSe QDs, indicating the significant deposition of copper on the surface of CdSe QDs. On the other hand, XPS spectra of Si QDs exhibited only a tiny peak of Cu 2p. The observation of the different in copper composition on the surface can imply that copper (II) ions have higher tendency to deposit onto CdSe QDs surface rather than Si QDs. This observation can be related to the possible mechanisms of the fluorescent quenching of CdSe QDs by Cu(II) ion as have been discussed previously as 1) the reduction of Cu (II) ion on the surface of CdSe QDs 25,[39][40] , 2) Cation exchange on the surface of CdSe QDs 22 , and 3) the Ksp of CuS, CuSe lower than CdSe 17,25 .
Detailed investigation on XPS analysis was done to better understand this phenomenon. The XPS spectra of CdSe in the presence of Cu 2+ ions demonstrated that Cd (II) ions remained on the surface, but they were disturbed by Cu (II) ion as the intensity of the main Cd peak decreased with slight redshift on the binding energy (Figure S3), while in Si QDs, only small tail loss (Figure 2b.) It can be implied that Cu (II) ions bound more effectively on the surface of CdSe QDs than Si QDs. Furthermore, at Cu 2p of high-resolution XPS spectra of CdSe (Figure 2g), various species of Cu including Cu 0 , Cu + , and Cu 2+ were detected, but they were not observed in Si QDs. This observation can be related to the partial reduction of Cu 2+ on the surface of CdSe QDs into Cu + and Cu 0 , and Cd 2+ on the surface can then be replaced by Cu ions that were reduced. Moreover, that high-resolution XPS at S 2p (Figure 2f) demonstrated the oxidation of S 2-and -SH on the surface of QDs to sulfate and thiosulphate at 168.0 and 169.3 eV. Another reason for the high ability of Cu 2+ in quenching of CdSe QDs could be the formation of CuS on the surface of CdSe QDs leading to nonradiative process for the exciton as the Ksp of CuS (6.0 × 10 -37 ) 17 is lower than Ksp of CdSe (6.3 × 10 -36 ) 22 . In comparison to the XPS analysis of Si QDs, the high-resolution XPS of Si QDs at Cu 2p illustrated the existence of Cu on the surface of QDs, but with lower intensity than on CdSe QDs. The reason for small Cu deposition could be that the functional group on Si QDs was mainly carbonyl group from sodium citrate capping agent, which has lower binding affinity with Cu 2+ . Upon the detailed investigation in combination with the fluorescence titration and Stern-Volmer's plot, it can be inferred that Cu 2+ mainly quench CdSe QDs but not Si QDs. The different effects of Cu (II) ions on Si QDs and CdSe QDs leads us to the idea of naked eyes Cu 2+ sensing probe called a mixed QDs probe. Since CdSe QDs were dramatically quenched and Si QDs were rarely affected by Cu 2+ ions, ratiometric fluorescence sensor combining Si QDs and CdSe were created. Mixed-QD probe was fabricated in 10 mM HEPES buffer (pH = 7.0) by mixing the two QDs in the buffer. The mixed-QDs probe exhibited compatibility between the two aqueous QDs without the induction of precipitation. FETEM image of the mixed-QDs probe provided information on the particle size and morphology of the probe as homogeneous mixing of the two dots and without any coating or aggregation of mixed dots observed (Figure S6.) The EDS analysis showing that this mixed probe contains Cd, Se, Si, S, C, N, and O as shown in Table S5 . The emission color of the mixed-QDs probe was yellow-green and was significantly brighter than CdSe QDs, Figure 4b.
The mixed-QDs probe showed 2 emission spectra at 430 and 573 nm belonging to Si QDs and CdSe QDs, respectively. However, we observed an enhancement of fluorescence spectra of CdSe QDs at 573 nm when Si QDs are added as shown in Figure 4.
Fluorescence resonance energy transfer, FRET, a non-radiative process between two fluorescence molecules was expected to be the reason for enhancement in CdSe fluorescence. FRET relies that the distance between the donor molecule (D) and the acceptor molecule (A) is typically of the order 1-10 nm 10,41 . For FRET to occur, there are few criteria needed. First, the spectra emission of D must overlap with the absorption of A. Also, the proximity between D and A should be about 10 Å to 100 Å [42][43] . The size of our QDs and the overlap between the emission spectra of Si QDs and absorption spectra of CdSe QDs clearly satisfied the FRET criteria. To confirm the energy transfer between Si QDs and CdSe QDs, fluorescence titration between Si QDs and CdSe QDs was studied in detail as shown in Figure S4. We carried out two titration experiments to confirm the energy transfer between Si QDs and CdSe QDs, 1) CdSe QDs were titrated with Si QDs and 2) Si QDs were titrated with CdSe QDs. The results showed that PL intensity of CdSe QDs were enhanced with the increasing amount of Si QDs, and the PL intensity of Si QDs were diminished with the increasing amount of CdSe QDs, confirming the energy transfer from Si QDs (donor) to CdSe QDs (acceptor) at the excitation of Si QDs (350 nm).
To further confirm the presence of energy transfer, time-resolved (TR) PL measurements were done using 375 nm pulsed laser diode excitation. To separately resolve emission from the Si QDs and CdSe QDs, we probed two different emission channels selected by a monochromator. Figure 4c shows the TR-PL signal at 485nm where the emission is dominated by Si QDs, while Figure 4d shows the signal at 575nm where the emission is dominated by the CdSe QDs once they are introduced. Both sets of decays were fit with biexponential functions, from which the decay lifetime τ D is expressed as the amplitude average lifetime. The initial samples of Si QDs (donor) have an amplitude average lifetime of 4.8±0.11ns, while that of the CdSe (acceptor) was determined to be 1.0±0.12ns. In the presence of energy transfer from the Si QDs (donor) to the CdSe QDs (acceptor), we will expect the average lifetime of the donor to decrease 44 . Specifically, because the Si QD lifetime is significantly longer than the CdSe QDs lifetime, we expect the lifetime measured at the CdSe QD emission channel to increase when the QDs are additionally excited via energy transfer from the Si QDs, approaching the lifetime recorded at the Si emission channel. Upon addition of CdSe QDs to the Si QDs, the amplitude average lifetime of the Si QDs (donor) continues to decrease with each aliquots addition of CdSe with a corresponding increase in the average lifetime of that of the CdSe QDs (acceptor) as shown Table S1. After the final addition of the CdSe QDs (reaching 0.25µM CdSe QDs), the amplitude average lifetime of the donor (Si QDs) has decreased to 4.3±0.17ns while that of the acceptor (CdSe QDs) has increased to 4.1±0.32ns as shown in the Figure 4c and Table S2. Given the low optical density of the sample, this is clear evidence of non-radiative energy transfer from the Si QDs to the CdSe QDs. [45][46] In addition to this and to corroborate the quenching of the CdSe in the presence of Cu 2 we added successive aliquots of Cu 2+ to the mixed QD sensor solution. As shown Figure S5, the lifetime of the CdSe QDs continued to diminish with each addition of Cu 2+ ions to the mixture. This further confirms that the CdSe QDs is strongly quenched in the presence of Cu 2+ ions.
Fluorescence technique was used to indicate the detection of Cu 2+ ions by the mixed QDs probe. The diminish of fluorescence intensity at 573 nm was observed upon the addition of Cu 2+ ions. This can be attributed to the effect of Cu 2+ ions on the CdSe QDs surface (Figure 5a.) The ratiometric plot: I430/I573 versus concentration of quencher, Cu 2+ (Figure 5b) showed that there was a linear relationship between the concentration of Cu 2+ ion and: I430/I573 value in the range of 0-100 nM (R 2 = 0.99046). The mixed-QDs probe has a detection limit of 3.89 nM, which was almost 10 times lower than the CdSe (MPA) QDs based sensor published previuosly 21 . The reasons for the high sensitivity of this probe were likely the combination from the sensitivity from using intensity ratio and the increase of CdSe QDs fluorescent signals due to FRET.
Another purpose of implementation of the mixed-QDs probe for sensing Cu 2+ is the ability for naked eye detection. In the presence of 10 µM of Cu 2+ , under a UV lamp, the mixed-QDs sensor changes the emission color from yellow-green to blue as shown Figure 4b.
In addition to Cu 2+ ions, the fluorescence quenching of mixed QDs-probe was also evaluated in the presence of some relevant biological cations and cations of environmental problem concerns: Ag + ,Fe 3+ , Ni 2+ , Cr 3+ ,Co 2+ , Ca 2+ , Li + , Mg 2+ , Sr 2+ , Al 3+ , Ba 2+ , Zn 2+ , Na + , K + , Pb 2+ , Mn 2+ , Cd 2+ , and Hg 2+ , as shown in Figure 6. The fluorescent ratio (I430/I573) of the mixed-QDs probe increased significantly when exposed to 1 µM Cu 2+ ions, whereas the exposure to the same concentration of other cations caused slightly changes (~20 times lower), shown in Figure 6a. In the presence of 1 µM Cu 2+ ions, other cations did not change the intensity ratio as well. These observations indicated the good selectivity of Cu 2+ ions compared with the other competitive cations.
The selectivity of the mixed-QDs probe to Cu 2+ ions are due to the reasons mentioned earlier including the deposition of CuS and/or substitution of Cd 2+ ion by Cu 2+ ion leading to the non-radiative recombination of exciton. From the hard-soft acid-base theory, Cu 2+ ion has Pearson border acid character which has a strong interaction with thiol groups 47 . Other metals with hard and borderline acid characters could bind to both carboxylate group of MPA on CdSe QDs and citrate group on Si QDs, and did not lead to the deposition of metal salts on QDs surface. Only in the presence of Cu 2+ ions, the thiol group of MPA will react with Cu 2+ ions to form CuS 17 leading to the quenching of luminescence 26 .
To demonstrate that the mixed-QDs probe can be used as the naked eye detection with selectivity to Cu 2+ , Figure 6b illustrates the changes in emission color comparing of Si QDs, CdSe, QDs, mixed-QDs probe (with and without Cu). The yellow-green emission of probe completely turn to blue emission upon the addition of Cu 2+ while it is remaining yellowgreen by the addition of the other cations.
This work has demonstrated the ratiometric fluorescence sensor for rapid, sensitive, and selective to Cu 2+ ions using the mixed-QDs probe in an aqueous solution. The mixed-QDs probe consists of two types of QDs are CdSe QDs and Si QDs without any modification on the surface. Mixed QDs probe has good solubility in aqueous without any accumulation. The yellow-green emission of mixed-QDs probe originated from the yellow-emitting CdSe QDs and blue emitting Si QDs. In the presence of Cu 2+ ions, the emission color would turn to blue. The detection limit was 3.89 nmol L -1 for Cu 2+ ions with the linear range from 0-100 nmol L -1 (R 2 = 0.99046). The mixed-QDs probe has a good sensitivity and selectivity toward to the biological relevant cations and cations of environmental concern: Ag + , Fe 3+ , Ni 2+ , Cr 3+ ,Co 2+ , Ca 2+ , Li + , Mg 2+ , Sr 2+ , Al 3+ , Ba 2+ , Zn 2+ , Na + , K + , Pb 2+ , Mn 2+ , Cd 2+ , and Hg 2+ . The detection limit and linear range of this mixed-QDs probe are comparable with other QDs probes reported previously, with the advantage of the large range of other metal cations that this probe can tolerate as shown in Table 1.
The mixed-QDs probe demonstrates the ratiometric fluorescent probe with only QDsbased fluorophores. Photostability, sensitivity, selectivity and ability for naked eye detection of Cu 2+ could allow this probe system to be further developed for use in analysis of water samples and biological specimens.