Pavitra V. Kumarab,
Beena G. Singh*a,
Anand Ballalc,
Vimal K. Jaind,
Michio Iwaokae and
K. Indira Priyadarsini*ab
aRadiation & Photochemistry Division, Bhabha Atomic Research Centre, Trombay, Mumbai – 400 085, India. E-mail: beenam@barc.gov.in; kindira@barc.gov.in
bHomi Bhabha National Institute, Anushaktinagar, Mumbai – 400 094, India
cMolecular Biology Division, Bhabha Atomic Research Centre, Trombay, Mumbai – 400 085, India
dChemistry Division, Bhabha Atomic Research Centre, Trombay, Mumbai – 400 085, India
eDepartment of Chemistry, School of Science, Tokai University, Hiratsuka-shi, Kanagawa 259-1292, Japan
First published on 13th July 2016
Selenium, like other chalcogens, shows a high affinity for gold nanoparticles (GNP). The binding interactions between the selenium centre and GNP can greatly influence its important physico-chemical properties, including electron transfer ability. To know if the chemical structure has any influence on these properties, in the present paper, binding and electron transfer reactions of two simple water soluble organoselenium compounds (SeC), i.e. a linear compound, (bis(2-ethanol)selenide) (SeEOH) and a cyclic compound, DL-trans-3,4-dihydroxy-1-selenolane (DHS) with GNP have been investigated. The binding with GNP of four different sizes (5–58 nm) was characterized by UV-visible spectroscopy, dynamic light scattering (DLS), zeta (ζ) potential, transmission electron microscopy (TEM) and surface enhanced Raman spectroscopy (SERS). Although both the compounds bind GNP through selenium atoms, they differ in orientation on the GNP surface. In DHS, only the selenium atom interacts, while in SeEOH, along with selenium atoms, alkyl groups also interact with the GNP surface. Stronger Se–GNP interaction and an increase in the electrophilicity of DHS as compared to SeEOH, was confirmed by their relative electron transfer reactivity with the ABTS˙− radical. Pulse radiolysis studies suggested that both the compounds on reaction with the hydroxyl (˙OH) radical produced similar selenium centered dimer radical cations (Se∴Se
)+, either in the presence or in absence of GNP, but significantly increased the yield of selenoxide in the presence of GNP, which is known to influence their antioxidant ability. Thus our results confirm that GNP can be used to modulate the electron transfer ability of selenium compounds.
Fig. 1 shows the representative absorption spectral changes for binding of SeEOH to GNP2, where addition of different concentration of SeEOH (5 μM–0.1 mM) to a fixed concentration of GNP2 (7 nM) resulted in decrease in the absorbance at 522 nm with monotonic evolution of a new band at 650–680 nm. Beyond 100 μM SeEOH, no further change in absorbance was observed. Earlier we reported that DHS causes similar spectral shifts on binding to GNP. The differential spectral shift, i.e. change in the absorption maximum, as a result of binding to GNP and the FWHM (full width at half maximum) was found to be lower for SeEOH as compared to that for DHS indicating higher polydispersity in the latter case. The binding constant (K) between GNP and SeC was calculated by using modified Benesi–Hildebrand equation (eqn (1)):25
![]() | (1) |
The double reciprocal plot of ΔA as a function of [SeC]0 gave a straight line and the ratio of the intercept and slope corresponds to the product of number of binding sites (n) and binding constant (K) (inset Fig. 1). Estimating the value of n according to the method described in ref. 26, the K values were calculated and are listed in Table 1.26 From the results, it can be inferred that under identical condition, the K value for SeEOH is higher than that for DHS by an order of magnitude.
S. no | Number of binding sites (n) | Binding constant (K, M−1) × 10−2 | Hydrodynamic diameter (nm) | |||
---|---|---|---|---|---|---|
SeEOH–GNP | DHS–GNP | GNP | SeEOH–GNP | DHS–GNP | ||
GNP1 | 1.9 × 104 | 5.2 ± 0.1 | 2.38 ± 0.09 | — | — | — |
GNP2 | 5.2 × 105 | 5.9 ± 0.2 | 0.97 ± 0.02 | 15 ± 3 | 26 ± 5 | 34 ± 4 |
GNP3 | 2.4 × 106 | 3.8 ± 0.1 | 0.38 ± 0.05 | 25 ± 3 | 69 ± 5 | 79 ± 6 |
GNP4 | 3.0 × 107 | 0.40 ± 0.02 | 0.09 ± 0.01 | 58 ± 5 | 140 ± 10 | 195 ± 8 |
![]() | ||
Fig. 2 Plot of correlation function G1(q,t) as a function of time for 7 nM GNP2 in absence (a) and in presence of (b) 0.1 mM SeEOH and (c) 0.1 mM DHS. |
Fitting the data to a monomodal distribution, the hydrodynamic diameter of GNP2 was estimated to be (15 ± 3) nm, which in the presence of 0.1 mM SeEOH and DHS increased to (26 ± 5) and (34 ± 4) nm, respectively. Similarly, hydrodynamic diameter of GNP3, GNP4 and their SeC-composites were estimated and the values are listed in Table 1.
The observed hydrodynamic diameter of SeC bound to GNP increases with increasing SeC concentration. At any given concentration, the size of SeEOH–GNP composites was lower than that for DHS–GNP composites. Since the hydrodynamic diameter obtained from DLS measurements is not a true representation of the actual size of the nanoparticles, TEM analysis was carried out. The TEM images revealed the average size of GNP1, 2, and 3 to be (5 ± 2), (12 ± 3), and (21 ± 4) nm, respectively. Fig. 3 shows the TEM images and size distribution of GNP1 and GNP3 in absence and presence of SeEOH and DHS. Addition of 1 mM SeEOH to the GNP caused aggregation without changing the shape of the GNP and the sizes were (5 ± 3), (17 ± 5) and (26 ± 5) nm for GNP1, GNP2 and GNP3, respectively. In the presence of 1 mM DHS, the size of GNP 1, 2 and 3 was found to be (7 ± 3), (20 ± 6) and (27 ± 8) nm, respectively. TEM images did not suggest any particle growth due to the binding of these compounds with GNP within a range of standard deviation, but indicate decrease in the interparticle distance leading to aggregation.
GNP sample | ζ-potential (mV) | ||
---|---|---|---|
GNP | SeEOH–GNP | DHS–GNP | |
GNP1 | −46.1 ± 0.5 | −30.1 ± 0.5 | −39.1 ± 0.8 |
GNP2 | −39.5 ± 2.9 | −27.6 ± 0.8 | −31.6 ± 1.5 |
GNP3 | −38.2 ± 1.7 | −24.2 ± 1.6 | −28.6 ± 1.2 |
GNP4 | −34.5 ± 0.9 | −21.6 ± 2.4 | −25.3 ± 0.4 |
Raman (SEOH),27 cm−1 | Raman peak (SeEOH), cm−1 | SERS (SeEOH–GNP) | ||
---|---|---|---|---|
Theory | Experimental | Peak position (cm−1) | Assignment | |
510.9 | 466 | 449 | S/Se-deformation | |
529.9 | 536 | 562 | 551 | C–Se–C stretching |
698.1 | 771 | 668 | 672 | C–Se–C stretching |
1015.2 | 984 | 1003 | 999 | CH2 bending (rocking) |
1059.9 | 1039, 1057 | 1055 | 1023, 1047 | CH2 bending (scissor) |
1127.8 | 1132 | 1124 | 1112 | CH2 bending (rocking) |
1198.4 | 1192 | 1186 | 1170 | CH2 bending (rocking) |
1298.8 | 1263 | 1276 | 1251 | CH2 bending (twisting) |
1332.6 | 1366 | 1321 | 1322 | CH2 bending (wagging) |
1502.0 | 1440 | 1467 | 1455 | CH2 bending (twisting) |
1558.0 | 1568 | 1563 | 1557 | CH2 bending (wagging) |
![]() | ||
Fig. 4 (a) Raman spectrum of pure SeEOH. SERS spectra of 0.1 mM SeEOH in presence of 7 nM (b) GNP1, (c) GNP, (d) GNP and (e) GNP4. |
However, in the case of SeEOH–GNP, along with these peaks (i.e. at 551 cm−1 and 1047 cm−1) several other peaks were also observed indicating interaction of alkyl side chain with GNP surface. Due to such interaction, it is expected that in SeEOH–GNP composites, the alkyl chain attached to selenium atom may remain parallel to the GNP surface. This kind of alkyl chain interaction appears to be dominating for larger GNP probably due to the fact that the surface curvature decreases with increase in GNP size. This assumption was further supported by the shift in peak maximum of SeC–GNP composites relative to SeC Raman peaks.
As listed in Table 4 the shift in peak maxima for C–Se–C stretching is more for DHS–GNP composites (636 cm−1) as compared to SeEOH–GNP composites (551 cm−1), whereas reverse trend is observed for CH2 bending peaks (DHS – 1234 cm−1 and SeEOH – 1023 cm−1), indicating role of alkyl chain in binding of SeEOH with GNP.
Compounds | Shift in peak position in absence and presence of GNP | |||
---|---|---|---|---|
GNP2 | GNP3 | GNP4 | ||
SeEOH | C–Se–C stretching (551 cm−1) | 9.7 | 9.7 | 11.8 |
CH2 bending (1023 cm−1) | — | 9 | 32.1 | |
DHS | C–Se–C stretching (636 cm−1) | 20 | 16 | 18 |
CH2 bending (1234 cm−1) | 1.8 | 1.8 | 0.6 |
For these studies the concentration of ABTS˙− radical, GNP and SeEOH were fixed at 30 μM, 7 nM and 1 mM respectively. ABTS˙− radical in the absence of SeC or GNP showed a very slow decay with a first order rate constant of (3.7 ± 0.3) × 10−6 s−1. On addition of 1 mM SeEOH or 7 nM GNP(1–4) only marginal changes in the decay of ABTS˙− were observed (Table 5). However on addition of 1 mM SeEOH–GNP1/2/3/4 to ABTS˙− radical solution, the absorbance at 820 nm decreased with simultaneous increase in the absorbance at 340 nm (due to formation of ABTS2−) (Fig. S5 in ESI†). Also, the rate constant values were found to increase with decrease in GNP size in a linear fashion (inset of Fig. 5). The observed rate constants (k) for the decay of ABTS˙− radical in the presence of SeEOH–GNP composite is listed in Table 5. Comparing these values with the earlier reported results for DHS–GNP composites, SeEOH–GNP composites appear to have low reactivity. However as reported earlier, since DHS itself showed significant reactivity with ABTS˙− radical as compared to SeEOH, the relative increase in the reactivity with ABTS˙− radical can be considered to be significantly higher for SeEOH than that for DHS. No attempt has been made to evaluate the bimolecular rate constant with GNP bound SeEOH or DHS.
S. no | (kABTS˙−+GNP), s−1 | (kABTS˙−+DHS), s−1 | (kABTS˙−+SeEOH), s−1 |
---|---|---|---|
In absence of GNP | 3.7 ± 0.3 × 10−6 | 0.97 ± 0.05 × 10−3 | 7.96 ± 0.08 × 10−5 |
GNP1 | 2.1 ± 0.1 × 10−4 | 1.08 ± 0.07 × 10−2 | 5.39 ± 0.08 × 10−3 |
GNP2 | 1.1 ± 0.1 × 10−4 | 3.52 ± 0.05 × 10−3 | 2.12 ± 0.05 × 10−3 |
GNP3 | 0.95 ± 0.05 × 10−4 | 2.94 ± 0.09 × 10−3 | 1.85 ± 0.09 × 10−3 |
GNP4 | 0.67 ± 0.03 × 10−4 | 2.39 ± 0.08 × 10−3 | 1.37 ± 0.07 × 10−3 |
As reported earlier, the (Se∴Se
)+ radical undergoes radical–radical reactions to produce selenoxide and more the stability of the radical, larger will be the radiation chemical yield (G-value) of the selenoxide. It has been well established that the selenoxide quantitatively reacts with dithiothretol (DTT) to give oxidised dithiothretol (DTTox).30 Therefore the radiolytic yield of selenoxide was estimated by monitoring the amount of DTTox by the above reaction using high performance liquid chromatography (HPLC) in the absence and presence of GNP1 (Fig. S6 in ESI†) at different absorbed doses (80 Gy to 120 Gy). The amount of selenoxide estimated was found to increase linearly with the absorbed dose and the slope of this plot corresponds to the G-value of the selenoxide.
As shown in the inset of Fig. S6,† it can be observed that G-value of selenoxide increases for both the compounds in the presence of GNP. However the extent of increase is higher for SeEOH as compared to DHS. Recently, Sicard-Roselli, et al. have demonstrated that the yield of ˙OH radical (G-value) increases in the presence of GNP which depends on the dose-rate, size and concentration of the nanoparticles.31 Therefore the yield of ˙OH radical in presence of GNP1 was estimated by monitoring fluorescence from 7-hydroxy coumarin (7-HOC) formed due to the reaction of ˙OH radical with coumarin.31 The yield of ˙OH radical in the absence and in presence of GNP1 were (0.6 ± 0.05) and (0.86 ± 0.08) μmol J−1, respectively. Since the yield of ˙OH radical increases in the presence of GNP, the amount of selenoxide formed was normalized with respect to the yield of ˙OH radical. After normalization, the amount of selenoxide was found to be (39.6 ± 6.0)% and (51.7 ± 4.0)% of the total ˙OH radical for SeEOH and SeEOH–GNP composites. Under similar experimental condition and normalisation, the yield of the selenoxide of DHS was estimated to be (51.0 ± 7.0)% and (56.0 ± 5.0)% of the total ˙OH radical in the absence and presence of GNP1 respectively. This clearly shows that, the selenoxide yield increases in SeEOH–GNP than that for DHS–GNP composites.
The differential mode of binding in SeEOH and DHS towards GNP was further confirmed by SERS studies. A large enhancement in C–Se–C stretching and CH2 bending modes was observed for both compounds, indicating involvement of Se atom in binding with GNP. The increase in strength of Se–GNP interaction leads to weakening of Se–C bond which is reflected by the shift of C–Se–C stretching peak towards lower wave number. For a given size, the shift in C–Se–C peak is more for DHS–GNP composites (20 cm−1 for GNP2) than that of SeEOH–GNP composites (9.7 cm−1 for GNP2) (Table 4). While the reverse trend is observed for CH2 bending peaks. This difference is negligible at lower GNP size but becomes significant at higher GNP size (GNP3 and GNP4). These results suggest that in SeEOH, the alkyl chain plays an important role in its binding and the orientation of SeEOH when bound to GNP surface. In DHS, such interactions are not feasible due to structural rigidity, therefore major interaction with GNP takes place via selenium atom.
Such binding is expected to differentially modulate the redox properties and rate of electron transfer reaction of SeEOH and DHS. This was assessed by monitoring the rate of electron transfer reaction between SeEOH in both free and GNP bound form with ABTS˙− radical (Fig. 5). The observed rate constants indicated that the reducing ability of SeEOH/DHS increased in presence of GNP and the reactivity increased with decrease in the GNP particle size (Table 5). The increase in the reducing ability of the GNP–SeC composites can be due to increase in the stability of the selenium centred radical formed on the GNP surface. Further, pulse radiolysis studies have shown that the (Se∴Se
)+ radical formed on reaction with ˙OH radical is more stable in SeEOH–GNP composites as compared to DHS–GNP composites (Fig. 6(A and B)). ˙OH radical reacts with SeEOH/DHS to form one-electron oxidised product which decays by two major pathways: (i) deprotonation to form α-C centred radical and (ii) association with parent SeEOH molecule to form (
Se∴Se
)+ radical.33 The former pathway is assisted by overlapping of σ orbital of α C–H bond with the p-orbital of Se atom. As indicated by SERS studies interaction of alkyl chain of SeEOH will lead to bending of CH2 moiety towards GNP surface and in turn will lower overlapping between σ orbital of α-CH2 and p-orbital of Se atom and deactivate the deprotonation pathway. Thus the binding of SeEOH with GNP will lead to preferential formation of (
Se∴Se
)+ radical which on disproportionation and sequential hydrolysis forms selenoxide.33 This was confirmed by the fact that there is increase in selenoxide formation in presence of GNP and the increase is much more pronounced in case of SeEOH than that for DHS. Formation of selenoxide is very important as it can regenerate parent SeC molecule in the presence of thiols. In absence of GNP, the selenium centred radical cation in SeEOH undergoes radiolytic degradation via deprotonation forming formaldehyde as one of the product. This pathway competes with dimer radical formation and results in low radiation chemical yield of its corresponding selenoxide. The deprotonation pathway is suppressed on binding of SeEOH with GNP which leads to increase in the yield of selenoxide. However, in DHS such degradation pathways are not observed, therefore no difference in radiation chemical yield of its selenoxide is observed in absence or presence of GNP. This differential increase in selenoxide is expected to influence their antioxidant activity significantly on binding to GNP surface. Thus our result demonstrates that small structural variations in selenium compounds can drastically change their redox properties that may have significance in their overall biological activity. Also such conjugation of selenium compound with GNP will result in the decrease of the radiosensitization effect of GNP and also decrease the toxicity of SeEOH due to suppression of oxidative degradation.
GNP of different sizes (GNP1–5 nm, GNP2–15 nm, GNP3–25 nm and GNP4–58 nm) were prepared by modified Turkevich method and separated from unreacted reactants by centrifugation as reported in our earlier work.23,36 Concentration of gold atoms in the GNP samples was estimated using a continuum source flame atomic absorption spectrometer (Contra AA 300, Analytik Jena, Germany), as reported earlier.23 The yield of the GNP was estimated to be 95 ± 5, 43 ± 5, 32 ± 3 and 30 ± 3% for GNP1, GNP2, GNP3 and GNP4 respectively. Absorption spectra were recorded on JASCO V-630 UV-visible spectrophotometer. For this, 7 nM GNP was incubated with varying concentration of SeEOH (0–1 mM) for 1 minute at pH 7 and corresponding absorption spectra were recorded.
ABTS˙− radicals was prepared by oxidation of aqueous 1 mM ABTS2− using 0.5 mM K2S2O8. The concentration of ABTS˙− radical was estimated by using its absorbance at 415 nm (ε415 nm = 3.6 × 104 M−1 cm−l).28 The kinetics of the reaction of ABTS˙− radical was studied by following the decrease in absorbance of ABTS˙− radical at 820 nm as a function of time, to avoid any interference from GNP.
The effect of binding of SeC with GNP on the particle size was studied by DLS technique.37 The measurements were done on Malvern-4800 Zetasizer instrument with He–Ne laser (632.8 nm) at room temperature and the scattered light was collected at an angle of 130 degree. The dynamic information of the particle can be derived from auto correlation function G1(q,t) where q is wave vector and t is time in seconds. The auto correlation data were fitted to single exponential decay to give translational diffusion constant, from which the hydrodynamic size was calculated using Stoke–Einstein equation.
TEM images were taken on Libra-120 plus TEM (Carl Zeiss, Germany) operated at 120 kV as the accelerating voltage. The GNP samples were prepared by room temperature evaporation of 10 μL GNP solution on the carbon coated copper grids.
The ζ-potential measurements were performed on Zetasizer Nano (Malvern Instruments, Ltd.). A sample cell with two gold electrodes is used to measure the ζ-potential using Henry equation.38 To estimate the change in ζ-potential due to binding of SeC with GNP, the concentration of GNP was fixed at 2.5 nM and that of SeEOH or DHS was varied from 50 μM to 1 mM. A voltage of 200 V was applied to induce electrophoresis of GNP solution taken in sample cell with gold electrodes.
Raman spectra were recorded using a micro-Raman spectrometer (SEKI Technotron) using excitation at 532 nm (power ∼20 mW at sample position) and a 10× objective lens (Olympus). The scattered light was collected by the same objective lens and passed through an edge filter (LPO3-532RU-25) to filter Rayleigh and anti-Stokes scattered signals. The surface enhanced Raman spectra were recorded at room temperature using the 660 nm line from a 50 mW DPSS laser (Ignis 660-500) M s−1 Laser Quantum Ltd. England. The sample solutions were exposed to laser in a standard 1 × 1 cm cuvette for 200 s and the Raman scattered light was collected at 90° scattering geometry and detected using a CCD (Synapse, Horiba JobinYvon) based monochromator (Triax550, Horiba JobinYvon, France) together with an edge filter, covering a spectral range of 200–1700 cm−1. Calibration of the spectrometer was carried out using benzene: CCl4 mixture (1:
1 v/v) and indene.
Pulse radiolysis studies were performed on 7 MeV linac electron accelerator with 100 ns pulse width coupled with absorption detector.39 Average dose of 9 Gy per pulse as estimated by thiocyanate dosimeter (aerated aqueous solution of 10 mM KSCN) was used for all experiments.40 Radiolysis of water at neutral pH generates both oxidizing radical ˙OH radical and reducing radical like hydrated electron (eaq−), along with H atom, and hydronium ion (H3O+). The reaction with ˙OH radical was carried out in N2O saturated solution, where eaq− generated during water radiolysis is quantitatively converted to ˙OH radical (G˙OH = 0.6 μmol J−1).41
Steady state radiolysis: or this N2O saturated aqueous solution containing 1 mM SeC and SeC–GNP composite containing 7 nM GNP1 was radiolyzed under steady state condition by γ-radiation using 60Co source (dose rate = 13.4 Gy min−1). The sample was radiolyzed at different doses and the maximum dose was set in such a way that not more that 10% of primary radical with respect to the SeC were generated. After radiolysis the samples were subjected to HPLC analysis to detect and quantify the selenoxide formed. For this 49 μL of the radiolyzed sample was incubated with 2 μL of DTT (50 mM) for 5 minutes and 20 μL of this reaction mixture was injected to Prontosil 120-5-C18 reverse phase column and eluted using 5:
95 acetonitrile
:
water (v/v) containing 0.1% TFA under isocratic method. The DTTox formed was detected using absorption detector at 240 nm and quantified by using standard DTTox calibration plot. HPLC chromatogram showed SeEOH, DHS, DTT and DTTox peaks at 4.3, 5.4, 8.3 and 14.4 min, respectively. The yield of ˙OH radical was calculated by monitoring the fluorescence of 7-HOC formed due to the reaction of ˙OH radical with coumarin. The fluorescence measurements were done on Syngene hybrid micro plate reader with excitation and emission wavelength as 326 nm and 455 nm respectively.31
Quantum chemical calculations: the geometry of SeEOH was optimized in vacuo using Density Functional Theory (DFT) at B3LYP (Becke non local model and Lee–Yang–Parr non local correlation functional)/6-31+G(d,p) level adopting the GAMESS suite of programs on a PC-based LINUX cluster platform.42 The resulted geometry was further optimized in water using solvent density model (SMD). The optimized structures were verified as global minima structure by performing the frequency calculation. Raman spectra of optimized geometry was calculated at B3LYP/6-31+G(d,p) level and visualization of different Raman modes was done using MacMolPlt v7.6.1.
Footnote |
† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra15106j |
This journal is © The Royal Society of Chemistry 2016 |