Electronic Supplementary Material

Ratiometric Sensing of Mercury(II) based on a FRET process on Silica Core-shell Nanoparticles acting as Vehicles

Baoyu Liuab*, Fang Zenga, Shuizhu Wua, Jiasheng Wangb, Fangcheng Tangb

a College of Materials Science and Engineering, South China University of Technology, Guangzhou 510640, China. Email: ; Tel: (+86) 20-2223-6363

b Guangzhou Lusahn New Materials Co. Ltd., Guangzhou 510530,China. *Corresponding author, Email:; Tel: (+86) 20-8226-6168

Scheme S1. Synthesis route of the probe precursor (Compound 2)

Scheme S2. Synthesis route of the donor precursor (compound 3)

Figure S1. Mass spectrum of Compound 1. m/z: [Compound 1 + H]+ 571.3

Figure S2. 1H NMR spectrum of Compound 1

Figure S3. Mass spectrum of Compound 2. m/z : [Compound 2 + H]+ 1065.8

Figure S4. 1H NMR spectrum of Compound 2

Figure S5. Mass spectrum of Compound 3. m/z: [NBD-APS + H]+ 407

Figure S6. 1H NMR spectrum of Compound 3

Figure S7. Overlap of the absorption band of the acceptor and emission band of the donor.

Figure S8. Fluorescence decay curves measured for NBD in the silica nanoparticle dispersion. Black curve: in the absence of Hg2+; red curve: in the presence of Hg2+ (5 ´ 10-5 M).

Calculation of the Förster life time(t)

In principle, the time-resolved decay curve of a fluorescence system can be fit to the exponential decay:

[Eq. (1)]

where i is the number of fluorescent components, I the time-resolved fluorescence intensity, ti the decay time (lifetime) for each component, A the additional background and αi values are the pre-exponential factors. Theoretically, there should be one decay time for the particle-based system in the absence of mercury ions since there is no FRET process involved; and there should be two decay times for the system in the presence of mercury ions, one for the unquenched donors and the other for the donors accessible to the quenching. However, as we tried to recover the decay time(s) from the I(t) curves by using the one- (for the system with no mercury ions) and two- (for the system in the presence of mercury ions) component exponential model, we found the goodness-of-fit for the fits as judged by χR2 (~ 2.0) were not close to utility. We suppose the reason is that the silica particle-based FRET system is a rather complicated one in which one donor (or acceptor) may interact with multiple acceptors (or donors), and the separation distance between the donors and the acceptors is not a single value. Thus, we calculated the average decay time < t > by averaging t over the intensity decay of the donor using a numerical integration software according to Equation 2, and used them as the apparent decay time (lifetime) for the donor before and after addition of mercury ions. The average decay time for the system in the absence and presence of the mercury ions have been calculated as 9.09 ns and 7.37 ns respectively. The decrease in fluorescence emission decay time of the donor provides additional evidence that in this system FRET process was turned on by mercury ion, since FRET process usually leads to a fluorescence lifetime decrease of the donor directly induced by energy transfer from the donor to the acceptor [17].

[Eq. (2)]

where I (t) is the time-dependent intensity for the donor; t1 and t2 are the lower and upper limit of the integration respectively.

Calculation of the Förster critical radii (R0) and determination of experimental energy transfer efficiency

The Förster critical distance R0 is the characteristic distance, at which the efficiency of energy transfer is 50%. The magnitude of R0 is dependent on the spectral properties of the donor and the acceptor molecules. If the wavelength λ is expressed in nanometers, then J(λ) is in units of M-1cm-1nm4 and the Forster distance, R0 in angstroms (Å), is expressed as follows [Eq. (3)]:

[Eq. (3)]

Κ2 is the orientation factor for the emission and absorption dipoles and its value depends on their relative orientation, n is the refractive index of the medium and ΦD is the quantum yield of the donor. J(λ) is the overlap integral of the fluorescence emission spectrum of the donor and the absorption spectrum of the acceptor (Figure S7) [Eq. (4)].

[Eq. (4)]

FD(λ) is the fluorescence intensity of the donor in the absence of acceptor normalized so that ; εA(λ) is molar extinction coefficient of the acceptor, λ is wavelength. In current experimental conditions, the J(λ) was calculated to be 5.65×1013 M-1cm-1nm4. The Förster distance (R0) has been calculated assuming random orientation of the donor and acceptor molecules taking Κ2= 2/3, n =1.54 (silica), and ΦD = 0.68.

For the silica nanoparticle-based FRET system in current experimental situation, by using a commercial software Origin 7.0 as the integral tool, we calculated R0 =25Å (2.5 nm). Energy transfer will be effective within the range R0 ± 50% R0 [17].

In addition, we estimated the experimental energy transfer efficiency EE of the current particle-based FRET system using the equation EE= 1- I/I0, where I0 is the fluorescence intensity for the donor with no Hg2+ and I is the intensity for the donor at which the change in fluorescence spectrum stopped during the titration of Hg2+. The calculated EE is about 0.27, which is a lower value compared to that for some small molecular FRET systems [18, 24]. Unlike those FRET-based small molecular chemosensor in which donor and acceptor are connected by covalent bonds, the two fluorophores in the nanoparticles of the current work were segregated, with the donor dispersed in the nanoparticle core and the acceptor connected on the surface of the particle. The distance between the donor and acceptor moieties within the nanoparticles is not a fixed value and can’t be accurately determined, and one donor can interact with multiple acceptors resided in close proximity simultaneously. According to the Förster nonradiative energy transfer theory, the energy transfer efficiency E, expressed by Equation 5, depends on the Förster critical radius R0, the average number of acceptors (ring-opened rhodamine B/Hg2+ complex) interacting with the donor n, and the distance (r) between the donor (NBD) and the acceptor. The energy transfer is effective over distances in the R0 = ± 50% R0 range.

[Eq. (5)]

The Förster critical radius R0 for this silica nanoparticle-based FRET system was calculated as 2.5 nm. The measured average diameter for the nanoparticle is 55 nm, and in the NBD moieties (donor) reside in the particle core while the probe/Hg2+ complexes (acceptor) reside in the particle surface. Since the upper limit of the effective energy transfer distance for the sample is around 3.8 nm (R0 + 1/2R0), the outer sphere of about 3.8 nm thick, which represents about 28 % of the overall volume of the silica core of the nanoparticle system, could serve as the effective space [28, 29-30] in which the probe/Hg2+ complexes can quench the fluorescence emission of NBD; on the other hand, those donors residing in the non-effective space (the inner core of the particles, averagely represents about 72 % of the overall volume of the silica particle) could not be quenched by the acceptors. Considering the fact that the experimental energy transfer efficiency is close to the effective volume for the system, we suppose the lower experimental energy transfer efficiency for this silica nanoparticle-based FRET system is due to the larger percentage of donors that were not accessible to the quenching.

9