Sodium Lauryl Sulfate-Conjugated Cationic Gemini-Surfactant-Capped Gold Nanoparticles as Model System for Biomolecule Recognition

Abstract

Surfactant-based nanostructures are promising materials for designing novel colorimetric biosensors based on aggregation/disaggregation phenomena. In this work, a colorimetric sensor based on the plasmonic shift of surfactant-capped gold nanoparticles via the disaggregation mechanism was developed. To perform this, the optimum SDS concentration was firstly determined in order to form Au@16-s-16/SDS complex aggregates with a well-defined SPR band in the blue region. Once the optimal SDS concentration for Au@16-s-16 aggregation was established, the sensing method depended on the nature of the electrostatic charge of the biopolymer studied where both the strength of the biopolymer/SDS and biopolymer/Au@16-s-16 interactions and the cationic gold nanoparticles play a key role in the disaggregation processes. As a result, an instantaneous color change from blue to red was gradually observed with increasing biopolymer concentrations. The response of the sensor was immediate, avoiding problems derived from time lapse, and highly dependent on the order of addition of the reagents, with a detection limit in the nanomolar and picomolar range for DNA and Lysozyme sensing, respectively. This behavior can be correlated with the formation of different highly stabilized Au@16-s-16/biopolymer/SDS complexes, in which the particular biopolymer conformation enhances the distance between Au@16-s-16 nanoparticles among the complexes.

1. Introduction

Accurate and reliable biopolymer determination is of immense importance not only for its role in the pharmaceutical industry but also for medical applications [1]. For instance, in the case of DNA, its successful detection can have an important impact on many important areas of research such as invasive-species research, medical diagnostics, drug development, or environmental health [2]. Additionally, DNA detection makes up an important tool in identifying the presence of genetic diseases such as cancer [3]. Likewise, protein detection has been shown to be useful for clinical diagnosis, treatment, and biological research [4]. In particular, Lysozyme is a cationic bacteriolytic protein and important biomarker whose abnormal concentration in serum and urine can be related to the presence of many diseases such as leukemia, meningitis, renal infection, and even the presence of cancer cells [5,6]. Importantly, in the case of lysozyme, the production of an excess of this protein, detectable in urine and other biological fluids such as blood serum, was shown to be a good indicator of the presence of monocytic and myelomonocytic leukemia (both subtypes of acute myeloid leukemia) in the 1960s [7,8]. Moreover, it was postulated that some alterations in renal function are associated with these types of leukemia, which would cause a decrease in the reabsorption of lysozyme in the kidney, leading to a disease called lysozymuria [9]. Specifically, acute lymphoblastic leukemia, which is currently considered the main type of childhood leukemia, causes a sharp decrease in serum lysozyme concentration with respect to control cases [10,11]. In view of this great need, therefore, multiple biopolymer measurement methods have been developed such as colorimetry [12], Raman spectroscopy [13], fluorimetry [14], mass spectrometry [15], chromatography [16], gel electrophoresis [17], lab-on-chip methods [18], and electrochemical sensing [19]. For the conventional methods, such as DNA detection using fluorescent labels in combination with amplification by polymerase chain reaction (PCR), several major drawbacks still remain to be overcome. The main reason for this disadvantage is related to the need to use complex algorithms and expensive instrumentation [20]. Particularly in the case of the detection of proteins such as Lysozyme, in addition to the abovementioned limitations, high instrument dependence is combined with low sensitivity and time lapse due to the long period required for sample incubation [21]. Moreover, commercial kits for detecting serum Lysozyme concentration involve the use of expensive antibodies requiring preservation (limited to 2–8 °C) and qualified personnel to manipulate and analyze the samples [7,22]. Therefore, it is clear that further simplification in terms of the time, cost, instrumentation, and complexity of sample preparation is needed, with all of this needed to effectively take these implementations from the laboratory to the production line. Among these sensing modalities, colorimetric sensors based on nanoparticles (NPs) provide several great advantages in comparison with other sensing methods, such as the high capability of surface modification of nanoparticles and simple biorecognition [23]. Furthermore, since physicochemical properties including charge, hydrogen-bonding ability, hydrophobicity/hydrophilicity, and surface topology can easily be modulated on the surface of NPs, nanosystems provide a versatile scaffold for biosensor output [24]. Specifically, AuNPs can be readily fabricated in sizes comparable to proteins, facilitating high-affinity interactions [25]. In a recent work, anisotropic gold nanostructures were used in an electrochemical biosensor to detect cardiac tiopronin I, with a wide range of detection from 0.06 to 100 ng/mL with high selectivity [26]. In another work, a colorimetric sensor array based on walnut-like Au@MnO₂ nanoparticles and MnO₂ nanostars was developed for detecting alkaloids. The sensor is based on the etching of two kinds of nanomaterials by choline action [27]. The design and working principle of the colorimetric methods is based on significant changes in NPs’ optical light absorption due to biopolymer recognition, and related changes in the position and intensity of the nanoparticles’ SPR bands [23]. However, the major drawback of such colorimetric biosensors is related to the time needed for biopolymer recognition, since the mechanism of action of the majority of AuNPs-based colorimetric sensors is established in biopolymer-induced nanoparticle aggregation [28,29,30,31,32,33]. In fact, aggregation events are frequently time-dependent, with aggregation kinetics in the range of hours or even days, clearly limiting their practical application [34,35,36,37,38,39,40]. Our strategy relies on in situ reversible disaggregation processes induced by biopolymers starting from a previously aggregated colloidal system. Our initial biosensing studies involved biopolymer identification employing negatively charged Au@citrate nanoparticles linked to a monomeric cationic gemini surfactant, forming aggregated complexes as biopolymer receptors [41]. Our efforts have concurrently been directed towards enhancing the sensitivity of the method. For this reason, in this report, we used arrays of cationic gold nanoparticles covered with gemini surfactants featuring an SDS single-chain surfactant on the surface in order to obtain aggregated complexes.

The aim of this work was to develop a new method to detect large biomolecules with a well-defined global charge. To perform this, we selected DNA and Lysozyme as model systems of highly negative and positive charge biomolecules, respectively. The sensing strategy and detection mechanism of large biomolecules described in this work are illustrated in Scheme 1. At first, to carry out the biosensor construction, it was necessary to find out the SDS concentration at which gold colloid aggregation was maxima. When SDS concentrations are below the CMC, the SDS is in its monomer form and the favorable electrostatic and hydrophobic interactions among cationic Au@16-s-16 nanoparticles and the anionic surfactants induce the formation of Au@16-s-16/SDS aggregates. Then, for nanocomplex disaggregation processes induced by biomolecule addition, the global charge of them and their affinity for both SDS surfactant and cationic gold nanoparticles plays a key role, in such a way that when the biomolecules have a well-defined positive charge, as in the case of Lysozyme, a competitive binding for the SDS that forms part of the Au@16-s-16/SDS aggregates is given, promoting Au@16-s-16 disaggregation. Thus, sensor construction for positively charged biomolecules is based on the addition method, A, in which increasing concentrations of the positive biopolymer are added to the previously formed Au@16-s-16/SDS aggregates, promoting the gradual disaggregation of the complexes and a color change in the solution that passes from blue to red. Note that Au@16-s-16/Lysozyme/SDS complexes are also formed due to the high affinity of Au@16-s-16 nanoparticles with the biopolymer (see Scheme 1). On the other hand, when well-defined negative-charge biopolymers such as DNA are sensed, the interaction biopolymer/SDS is almost negligible and, thus, the competitive binding for the surfactant is not feasible. Thus, in this case, the addition method B is operative. In this way, the biopolymer is firstly added to the nanoparticles followed by the SDS surfactant at a fixed concentration that guarantees the maxima Au@16-s-16/SDS aggregation. As a result, the favorable interaction DNA/Au@16-s-16 causes gradual Au@16-s-16/SDS complex disaggregation with increasing biopolymer concentration. Consequently, the color of the solution gradually passes from blue to red. In the specific case of DNA, a well-defined Au@16-3-16/DNA/SDS network is formed in which gold nanoparticles are well separated from each other, guaranteeing the gold colloid disaggregation. Thus, the detection strategies involving Lysozyme and DNA are not the same and depend on the global charge of the biomolecule studied. Precisely, one of the advantages of this work compared with the previous one based on Au@citrate/12-s-12 complexes is the nature of the surfactant used to form the initial aggregates [41]. Note that the negative charge of SDS makes DNA/SDS interaction not feasible and, consequently, the method of detection changes depending on the global charge of the studied biopolymer. Moreover, two kinds of surfactants play a key role in the biosensor: (i) the positively charged 16-s-16 gemini surfactant that integrates the Au@16-s-16 nanoparticle and (ii) the negative charge SDS monomer. This results in the formation of more stable Au@16-s-16/SDS aggregates induced by the highly favored electrostatic and hydrophobic interactions among the surfactants. This characteristic of the new rearrangement allows a more gradual disaggregation process, expanding the range of concentrations in which the sensor response is linear as well as its sensitivity. On the other hand, another advantage of this work in comparison with the use of nanosystems based on surfactants is related to the simplicity of the sensor. In a previous work, an array of surfactant-stabilized gold nanoparticles that used CTAB or SDS as a monomeric surfactant was employed to distinguish different proteins. The results showed that higher protein concentrations in the solution were required to induce nanosystem aggregation under higher CTAB concentrations [42]. However, in this work, the sensor response is based on the use of fixed gold nanoparticle and SDS concentrations, which contributes to simplifying both the data analysis and sensor construction. Different methods based on colorimetric approaches and electrochemical methods have been developed to detect different biomolecules such as DNA or Lysozyme [43,44,45,46]. The main advantages of this work with respect to the previous ones are: (i) the rapid response of the sensor in the order of seconds which highly reduces the time cost for biomolecule determination, especially in the case of Lysozyme; (ii) the simplicity of the method that does not require qualified personnel for its determination and requires less complex protocols; (iii) the versatility of the method that permits detecting both positive- and negative-charge biomolecules by changing the addition order of the reactants; and (iv) the stability of the nanocomplexes which contributes to gaining reproducibility and storage stability for commercial purposes. Thus, the novel method is sensitive and fast compared with other colorimetric methods reported in the literature. For instance, a polymer–aptamer detection probe based on the AuNPs crosslinking strategy has served to detect Lysozyme through the UV-visible technique with a detection limit of 4.4 nM [47]. In another recent work, this protein was detected in human urine based on the use of Au@citrate of 15 nm and colorimetric analysis based on CIEL*a*b* with a detection limit of 4 µM [48]. In the case of DNA, Au@citrate nanoparticles were used to detect DNA based on SPR displacement with an LOD of 1 µM [49]. In another work, a more complex nanosystem based on oligonucleotide-modified AuNPs was used to detect genomic DNA with an LOD of 600 pM [50]. The greater physicochemical interaction between Au@16-s-16/SDS and biopolymers permits their recognition from the molar to the nanomolar or picomolar range of concentration for DNA and Lysozyme, respectively, with Au@16-s-16 SPR band modification from blue to red at about 200 nm. Taken together, these studies demonstrate that through the tuning of the AuNPs receptor structure and charge, highly effective and sensitive array-based sensors for biopolymers can be produced. However, the mechanism of molecular recognition developed in this work should be individually designed and adapted taking into account the global charge of the biomolecule and the media used in the quantification to be able to expand the pool of available analytes.

2. Experimental Section

2.1. Materials

All commercial chemicals mentioned were of Anal. R. Grade. The biopolymers used were Lysozyme (from chicken egg white) and calf thymus DNA. Both reactants were purchased from Sigma-Aldrich-Merck KGaA (Darmstadt, Germany). Donor equine serum was purchased from Cultex (16SH30074.03, CULTEK, Hanover, Germany). Hydrogen tetrachloroaurate (III) trihydrate, sodium cacodylate, and 3-aminopropyltriethoxilane (APTES) were also purchased from Sigma-Aldrich-Merck KGaA (Darmstadt, Germany). Sodium borohydride (NaBH₄) was purchased from Panreac Química S.L.U (Barcelona, Spain). No protein contamination was evaluated measuring the absorbance ratio of DNA stock solutions at 260 nm and 280 nm (A₂₆₀/A₂₈₀ = 1.87) [51]. The average number of DNA base pairs (mean bp) was tested in a previous work by using electrophoresis and ethidium bromide as biomarker. As a result, the DNA mean size in base pairs was higher than 10,000 [52]. DNA polynucleotide concentrations, given in phosphate groups, were determined spectrophotometrically from the molar absorptivity (6600 M⁻¹ cm⁻¹ at 260 nm) [53]. The temperature of the samples was maintained at 298.0 ± 0.1 K in all the experiments, and the water used for preparing the samples had a conductivity of less than 10⁻⁶ Sm⁻¹. The working pH for the experiments in water and in horse serum media was 6.4 and 7.4, respectively. Total concentrations of DNA and Lysozyme biopolymers, gold nanoparticles covered with 16-s-16 gemini surfactant, and the SDS surfactant in a working solution will now be referred to as C_DNA, C_Lysozyme, C_Au@16-s-16 (s = 3 or 6), and C_SDS, respectively.

2.2. Synthesis and Characterization of Cationic 16-3-16 and 16-6-16 Gemini Surfactants

To synthesize the gemini surfactant compounds, 0.05 mol of α,ω-dibromoalkane and 0.12 mol of N,N-dimethylalkylamine was mixed in 100 mL of acetonitrile as the solvent. Then, the resulting solution was stirred under reflux for 20 h. Upon cooling, a white solid was recovered with filtration. The obtained samples were recrystallized from ethyl acetate. This process was repeated up to five times and, finally, the products were dried under a vacuum [54]. The gemini-obtained were characterized using the mass spectrometry technique [55], nuclear magnetic resonance spectroscopy, and elemental analysis. Critical micelle concentrations (CMC) of the surfactants were measured by using the surface tension technique. As a result, values of 2.4 × 10⁻⁵ M and 3.2 × 10⁻⁵ M for 16-3-16 and 16-6-16, respectively, were obtained at 298 K. The obtained results are in accordance with the values reported by Zana et al. [56]. For more details of gemini surfactant characterization, see the SI of this paper.

2.3. Synthesis of Au@16-3-16 and Au@16-6-16 Gemini-Surfactant-Capped Gold Nanoparticles

Gold nanoparticles functionalized with gemini surfactants were prepared following a similar process developed by our group [57]. A modification of the Turkevich method was used for which Au³⁺ ions were reduced to Au⁰ ions using NaBH₄ as a reducing agent, obtaining stable nanoparticles [58]. Au@16-s-16 nanoparticles were synthesized using NaBH₄ (100 μL, 0.4 M) as a reductant for hydrogen tetrachloroaurate (III) hydrate (HAuCl₄·4H₂O) (390 μL, 23 mM) and the appropiate gemini surfactant concentration as a stabilizing agent (30 mL of 16-3-16 or 16-6-16 gemini surfactant 1 · 10⁻⁴ M aqueous solutions). A yellow solution that was stirred for 5 min was obtained at first when the surfactant was added to the gold. Then, a freshly prepared borohydride solution was added drop by drop to the gold–surfactant mixture. The sample was stirred moderately for 15 min in darkness and, finally, a reddish-color solution was obtained. TEM measurements and the ImageJ sofware program were used to characterize the size and morphology of the synthesized nanoparticles. For this, a sample set of over 200 nanoparticles was used. As a result, we obtained monodisperse Au@16-3-16 and Au@16-6-16 gold nanoparticles whose mean size was found to be 3.8 ± 0.8 nm and 4.1 ± 0.9, respectively (see Figure S1). The total concentration of the colloidal stock solution was taken from the average nanoparticle size obtained from the TEM (see Section 2.4.3) and considering the reduction process was fully accomplished. As a result of this, C_Au@16-s-16 was 1.74 × 10⁻⁷ M and 1.38 × 10⁻⁷ M for Au@16-3-16 and Au@16-6-16, respectively.

2.4. Methods

2.4.1. UV/Vis Spectroscopy

A CARY 500 SCAN UV-vis-NIR (Ultraviolet/Visible/Near Infrared) UV-vis spectrophotometer (Varian, Markham, ON, Canada) was used to measure absorbance spectra, by using a standard quartz cell of 1 cm path length. The temperature of the measurements was fixed at 298.2 K and the wavelength range explored for measuring the spectra was 400–800 nm. The wavelength accuracy and spectral bandwidth were ±0.3 and 1 nm, respectively. The Au@16-s-16/SDS samples prepared to study the aggregation process were incubated for 15 min to ensure the stabilization of the system after recording the absorbance spectra. In contrast, the disaggregation experiments of the complexes in the presence of biomolecules were carried out in situ. This is because we proved that the solutions were stabilized quickly. Absorbance titration experiments were carried out at a fixed colloidal gold concentration of C_Au@16-3-16 = 1.74 × 10⁻⁸ M and C_Au@16-6-16 = 1.38 × 10⁻⁸ M for all the experiments. The samples were repeated three times and the relative error was calculated to be < 5%. Moreover, it was checked that the samples were stable for at least 24 h of their preparation.

To measure biopolymer/16-s-16 equilibrium binding constants, a fixed C_16-s-16 was used, and C_biopolymer varied from 5.0 × 10⁻⁸ M to 8.0 × 10⁻⁴ M and from 5.0 × 10⁻⁹ M to 1.0 × 10⁻⁴ M for DNA and lysozyme, respectively. Absorbance data at 525 nm and 516 nm were then analyzed in accordance with the Hildebrand–Benesi model for Au@16-3-16/biopolymer- and Au@16-6-16/biopolymer-based systems, respectively [59,60] (see Section 3.2 for more details).

2.4.2. Deconvolution Procedure

The deconvolution of experimental absorbance spectra was carried out with the help of Fityk deconvolution software (version 0.9) [61]. The absorbance spectra of distinct nanosystems were always fitted to two Gaussian functions. The deconvolution model was based in the use of two absorbance bands: one of them centered at around 526–548 nm and 515–551 nm for Au@16-3-16/SDS and Au@16-6-16/SDS systems, respectively, that was assigned to the nonaggregated nanoparticles, and another Au@16-s-16 absorbance band quite far above 540 nm that was associated with aggregated nanoparticles. The second band can be assigned to particles with different size distributions depending on the aggregation state of the analyzed sample. Moreover, a variable linear function was included to ensure the best fit. The functionality of the added function was to correct the possible effect of residual gold salts remaining in the solution as well as the effect of light scattering phenomena in each sample [62,63].

2.4.3. TEM Measurements

Isolated gemini surfactant nanoparticles and Au@16-s-16/SDS complexes were placed on a copper grid coated with a carbon film for TEM examinations. The samples were dispersed in water as solvent and a total of 10 μL of nanoparticles or the complex solution was dropped onto the copper grid. The samples were then air-dried at room temperature for a period of 2 h. The size of gold nanoparticles was analyzed in Figure S1. EDS measurements were employed to confirm the presence of gold in the Au@16-s-16/SDS complexes. Figure S2 shows EDS spectra in which the presence of K and L lines associated with the gold element demonstrated the presence of gold in the nanocomplex. Moreover, the intensity (counts) versus position (nm) profiles for Au@16-s-16/SDS complexes provided values of d = 0.24 nm in both nanosystems (see Figure S3). TEM analysis was performed in a high-resolution TEM-TALOS F200S electron microscope (FEI Company, Hillsboro, OR, USA) equipped with an energy-dispersive X-ray spectrometer and working at 200 kV. ImageJ 1.52a software was used to analyze the obtained TEM images.

2.4.4. Zeta Potential Measurements

The zeta potential charges of the free nanoparticles and Au@16-s-16/SDS complexes were measured in water after 15 min of sample preparation and stabilization. A Zetasizer Nano ZS from Malvern Instrument Ltd. (Worcestershire, UK), which employs a laser Doppler velocimeter (LDV) to quantify the velocity of the particles based on their electrophoretic mobility, was used to measure the Zeta potential (ζ) values. The type of cell used was a DTS1060 polycarbonate capillary cell thermostatized to 298 K. The number of sample repetitions was at least six in each case. To prepare the samples, C_Au@16-s-16 concentrations were fixed at 1.74 × 10⁻⁷ M and 1.38 × 10⁻⁷ M for Au@16-3-16 and Au@16-6-16, respectively, while C_SDS was varied from 1.0 × 10⁻⁶ M to 1.0 × 10⁻³ M.

2.4.5. Atomic Force Microscopy Measurements (AFM)

The AFM micrographs were obtained with a Molecular Imaging Picoscan 2500 (Agilent Technologies, Las Rozas, Madrid, Spain). Images were registered in air and in tapping mode using scan speeds of 0.5 Hz. To perform this, silicon cantilevers (Model Pointprobe, Nanoworld Neuchâtel, Switzerland) were used; the resonance frequency was 240 kHz and the nominal force constant was 42 N/m. In order to guarantee the correct complex formation, in all cases, the total biopolymer concentrations were fixed at 5.0 × 10⁻⁵ M and 1.0 × 10⁻⁵ M for DNA and Lysozyme, respectively. These concentrations were selected in accordance with results obtained from TEM microscopy and UV-visible spectrophotometric titrations. To prepare samples for AFM visualization, a freshly cleaved mica surface was firstly modified with 0.1% (v/v) APTES in water solution. An incubation time of 20 min was used to modify the mica surface with APTES. This process was followed by washing of the surface with ultrapure water, which was then dried with air. Next, a total of 100 μL of each sample was dropped onto the APTES-modified mica surface. After 60 min of adsorption, the sample was washed with doubly distilled water and then air-dried. Subsequently, the AFM images were acquired and flattened to remove the background slope [64].

3. Results and Discussion

3.1. Au@16-s-16 Aggregation Induced by Anionic SDS Surfactant: Optimization of Biosensor Configuration

As previously mentioned, contrary to what occurs in aggregation processes, disaggregation processes occur instantaneously under the appropriate conditions. Consequently, for the development of an efficient response in a biosensor, it could be relevant to find a system configuration permitting the study of colloidal gold disaggregation phenomena. Thus, the first step of this work was to determine the optimum SDS concentration to induce Au@16-s-16 aggregation in the absence of any added biomolecule. For this purpose, it is crucial to find a well-defined SPR band of the Au@16-s-16/SDS aggregated complex located in the blue region around 600–700 nm. Firstly, it allows the incorporation of an SDS anionic surfactant to cationic Au@16-s-16 gold nanoparticles, and the formation of the complex is fundamentally promoted by attractive electrostatic interaction. At concentrations below the CMC of the anionic surfactant [65], SDS monomer interactions promote gold nanoparticle aggregation, leading to a color change in the solution from red to blue, visible to the naked eye (see Figures S4 and S5). Moreover, when we represent the maximum wavelength of the SPR band for the aggregation of the Au@16-s-16/SDS systems as a function of C_SDS concentration, it is possible to distinguish the optimum aggregation of Au@16-s-16 nanoparticles in the maximum observed for Figures S4 and S5, that is, 651 nm and 650 nm for Au@16-3-16/SDS and Au@16-6-16/SDS systems, respectively, at a C_SDS concentration of 30 μM in both cases. In addition, Figure 1A and Figure S6A denote a great shift in the maximum wavelength of the SPR band of 125 nm and 135 nm for Au@16-13-16 and Au@16-6-16, respectively, where the maximum passes from 526 nm (in the absence of SDS) to 651 nm for Au@16-13-16, and from 515 nm to 650 nm in the case of the analogous s = 6 derivative. As previously mentioned, this behavior can be explained considering the favorable electrostatic interaction between the anionic SDS surfactant and the positively charged Au@16-s-16 nanoparticle. It is important to note that the positive charge of the particles emerges from the adsorption of the 16-s-16 cationic gemini surfactants onto the colloid’s surface. Note that this process occurs when gold nanoparticles are synthesized [57]. It is significant that the aggregation process is progressive and time-dependent as to when a purple color is observed (see Figures S4 and S5) and a time lapse of approximately 15 min is required for Au@16-s-16/SDS complex stabilization working below the CMC. Furthermore, when we analyzed the absorbance results in depth, we noted that changes in the maximum SPR bands for both systems upon aggregation were accompanied by a shift in the absorbance intensities at the maximum wavelength (650 nm and 651 nm for s = 3 and s = 6 nanosystems, respectively) to higher values (see Figure 1C and Figure S6C), together with a concurrent decrease in the absorbance intensities at 526 nm and 515 nm for the s = 3 and s = 6 systems, respectively, whose wavelengths were selected to represent the behavior of the nonaggregated systems. However, starting from C_SDS = 30 μM above the maximum aggregation concentration, the trend in the absorbance spectra with C_SDS concentrations was opposite to that previously described (see Figure 1B and Figure S6B). In this sense, the absorbance intensity clearly decreased at the maximum wavelength (in blue) and increased at the minimum wavelength (in red). Therefore, two zones can be clearly distinguished in the graph of the absorbance data versus C_SDS concentrations: a change zone and a stabilization zone that begin near the CMC value (see Figure 1D and Figure S6D). Furthermore, the final state of the system at the higher SDS concentration was similar to the Au@16-s-16 isolated system; namely, when the surfactant self-assembled into micelles over CMC, no color change was observed with respect to the control. According to the bibliography, a valid explanation for this behavior would be that the formation of SDS micelles allows the Au@16-s-16 nanoparticles to remain separated from each other through steric impediment even upon interaction with the surfactant [66,67].

However, this model of interaction is highly conditioned by the nanoparticle/micelle size ratio. In our case, the mean diameter of the Au@16-s-16 nanoparticles was found to be 3.8 nm and 4.1 nm for the s = 3 and s = 6 derivative, respectively. Furthermore, the published values for the diameter of an SDS anionic micelle are between 3.5 and 4.0 nm, very similar to gold nanoparticles in solution [68,69]. Accordingly, it is reasonable to assume that for C_SDS concentrations over CMC, the interaction of the negatively charged micelles with positively charged gold nanoparticle surfaces acts as a steric impediment among nanoparticles, avoiding colloidal aggregation. Additionally, at concentrations below but close to CMC, the formation of surfactant aggregates on Au@16-s-16 nanoparticles could be the reason for the observed gradual color change, becoming purple in solutions near CMC. In light of this, since the formation of the micelles is not yet fully completed, it seems that the size of the aggregates can in part prevent the aggregation phenomena (see Figures S4 and S5). At concentrations over CMC, the surfactant–gold solutions returned to the red color exhibited by isolated nanoparticles, and no aggregation was observed. Note that these results are in good agreement with the model proposed by Kazakova et al., who studied the interactions of citrate-capped gold nanoparticles with an SDS surfactant [67].

The color changes observed, and thus the associated changes in the maximum wavelength of Au@16-s-16/SDS systems, can be used to give an approximation of the CMC of SDS surfactants. Therefore, a graphical representation of the SPR maximum wavelength (λ_MAX) with C_SDS near and above CMC allowed us to distinguish two very distinct branches that can be assimilated in two straight lines (see Figure 2). CMC was then calculated through the intersection of those lines. The implementation of this methodology for CMC determination through λ_MAX measurements gave very close values for SDS by using both Au@16-s-16 systems: a value of 7.2 × 10⁻³ M and 7.7 × 10⁻³ M for s = 3 and s = 6 derivatives, respectively. Note that these CMC values in water are in good agreement with those reported by Khan et al. [70] who found a CMC value of 8.0 × 10⁻³ M.

It is important to note that a better approximation was obtained for CMC determination employing Au@16-s-16 λ_MAX measurements than that based on the use of Au@citrate of 10 nm core size, where a CMC value of 7.5 × 10⁻⁴ M in the presence of 0.01 M of NaCl was found [67], in comparison with data previously reported by Williams et al. of 1.46 × 10⁻³ M under identical experimental conditions [71]. This fact can be explained taking into account two fundamental differences with previous similar methodology: the smaller diameter of Au@16-s-16 nanoparticles compared to Au@citrate and the cationic character of the gemini-surfactant-based nanosystem. Therefore, one of the applications of this colorimetric system based on cationic gold nanoparticles is the CMC determination of distinct anionic surfactants. Finally, the SPR experimental spectra of Au@16-s-16/SDS systems were analyzed in depth using a simple deconvolution procedure in order to obtain the isolated contribution of each individual band. For this purpose, the experimental spectra were divided into two contributions: the band corresponding to the nonaggregated and the aggregated system with λ_1,max and λ_2,max as characteristic parameters, respectively, as well as its corresponding area of peaks (see Section 2.4.2). The data in

Table 1. Deconvolution parameters for the experimental spectra obtained by Au@16-3-16/SDS and Au@16-6-16/SDS systems.

Au@16-3-16/SDS System			Au@16-6-16/SDS System
CSDS(M)	λ1, max/nm (Area1)	λ2, max/nm (Area2)	CSDS/M	λ1, max/nm (Area1)	λ2, max/nm (Area2)
0	526	---	0	515	---
1.0 × 10−6	527	---	1.0 × 10−6	517	---
5.0 × 10−6	528	---	5.0 × 10−6	518	---
1.0 × 10−5	528	---	1.0 × 10−5	519	---
2.0 × 10−5	547 (2.03)	646 (9.02)	2.0 × 10−5	550 (1.83)	648 (10.81)
3.0 × 10−5	548 (1.94)	651 (10.17)	3.0 × 10−5	551 (1.82)	650 (10.84)
4.0 × 10−5	547 (2.03)	647 (9.01)	4.0 × 10−5	534 (3.39)	605 (6.39)
5.0 × 10−5	540 (2.28)	615 (7.06)	5.0 × 10−5	524 (1.12)	571 (5.09)
6.0 × 10−5	533 (2.51)	589 (5.56)	6.0 × 10−5	524 (1.08)	569 (5.21)
7.5 × 10−5 (*)	531	---	7.5 × 10−5	528	---

(*) Data specification for the first concentration at which the deconvolution process is not necessary for each system.

show that the nonaggregated system is represented by a narrow peak, λ_1,max, which varies between 526 and 548 nm and 515 and 551 nm for Au@16-3-16/SDS and Au@16-6-16/SDS systems, respectively. Consequently, this peak corresponds not only to the isolated gold nanoparticle but also to the nonaggregated complexes formed between Au@16-s-16 and SDS in monomeric and micelle forms.

In addition, the broader red-shifted peak, λ_2,max, was only due to the assembly of Au@16-s-16 in the presence of monomeric SDS, since from C_SDS = 75 µM, only a single peak was visualized and the deconvolution procedure was not needed. On the basis of the results given in Table 1, the magnitude of the aggregation–disaggregation processes measured using the maximum changes registered in both peaks was very similar in both Au@16-s-16 systems. However, these processes were only slightly greater in the case of Au@16-6-16, which is in accordance with the more precise CMC value obtained for this system. To verify or not the aggregation/disaggregation processes of gold nanoparticles detected with the absorbance spectroscopy technique as a consequence of anionic SDS surfactant addition, TEM experiments were carried out (see Figure 3).

To verify the aggregation/disaggregation processes of gold nanoparticles detected with the absorbance spectroscopy technique as a consequence of anionic SDS surfactant addition, TEM and DLS experiments were carried out. Two different aggregation states for the Au@16-s-16/SDS systems can be observed as a function of the SDS concentration. When the SDS concentration was far below the value of the CMC of SDS in water, large-size aggregates could be seen (see Figure 3A–F). Hence, the interaction of monomeric anionic SDS surfactant with highly cationic gold nanoparticles covered with 16-3-16 and 16-6-16 gemini surfactant resulted in colloid aggregation, where favorable electrostatic and hydrophobic interactions probably played a key role in this process. Note that the size of the aggregates was very similar in the two nanosystems explored, with values that varied between 200 and 800 nm, revealing that the nature of the spacer length was negligible. However, for solutions prepared at SDS concentrations very near or above the CMC, the disaggregation process took place in both Au@16-s-16/SDS systems (see Figure 3G–I). A close examination of Figure 3I reveals that SDS micelles could probably act as an electrostatic barrier for Au@16-s-16 aggregation, showing well-spaced colloids. Note that this assumption is in accordance with the model proposed by Kazakova et al. and Kuong et al. [66,67] in that when the surfactant concentration was above CMC, the formation of the micelles led to the separation of distinct nanoparticles and the solution turned red. In addition, as the nanoparticle/SDS micelle size ratio was close to 1 in both cases (1.01 for s = 3 and 1.09 for s = 6 derivative), steric hindrance between the nanoparticles and SDS micelles could reinforce the disaggregation phenomena. Both aggregation/disaggregation behaviors described using the TEM measurement were in good agreement with the results obtained using DLS and spectrophotometric titration, showing a corresponding increase/decrease in both the hydrodynamic size of the complexes and the spectra maximum wavelength with varying SDS concentrations. The results obtained from DLS are given in

Table 2. Values of the hydrodynamic diameters of different Au@16-s-16 nanoparticles and Au@16-s-16/SDS structures.

CSDS/M	Au@16-3-16/SDS	Au@16-6-16/SDS
0	(3.6 ± 0.5) nm	(3.9 ± 0.5) nm
5.0 × 10−5	(748 ± 11) nm	(400 ± 12) nm
1.0 × 10−3	(18 ± 4) nm	(15 ± 4) nm

and Figure S7, showing different steps of aggregation/disaggregation processes observed for Au@16-s-16/SDS nanosystems.

Considering the hydrodynamic sizes values in Table 2, it can be concluded that at C_SDS below the CMC, the diameters of the Au@16-3-16/SDS aggregates was greater than those of the analogous Au@16-6-16/SDS nanosystem. Furthermore, at C_SDS above the CMC, the size of the Au@16-s-16 structures was not exactly the same as that observed in the absence of the monomeric surfactant. This result seems to support the idea that SDS micelles play a key role in Au@16-s-16 disaggregation in the absence of biomolecules (see Scheme 1). Despite the fact that the aggregation/disaggregation processes have been verified using absorbance spectroscopy, DLS and TEM microscopy techniques, the possibility that electrostatic forces may mediate such SDS/Au@16-s-16 interactions still needs to be confirmed. In order to verify the importance of electrostatic Au@16-s-16/SDS interactions, zeta potential experiments were performed; the results are given in Figure 4.

As shown in Figure 4A,B, for both systems, the highly positive charge of the free gold nanoparticle strongly diminished when a small quantity of SDS surfactant was added to the solution, in such a way that the charge switched from 50 mV to 3.1 mV and from 34.3 mV to 1.7 mV at 2.5 × 10⁻⁵ M for the Au@16-3-16/SDS and Au@16-6-16/SDS systems, respectively. The charge was then inverted to a small global negative charge of −17.5 mV and −14 mV at 5.0 × 10⁻⁵ M, and remained below −20 mV up to 6.5 × 10⁻⁵ M. Note also that these results are in good agreement with the data in Table 1, in which the deconvolution of the measured spectra was needed for both systems from 2.0 × 10⁻⁵ M to 6.0 × 10⁻⁵ M due to the overlapping spectra of the free and aggregated colloid in the solution. As is known, low values of zeta potential may be associated with colloid aggregation/flocculation phenomena that occur due to the action of van der Waals attractive forces among the nanoparticles [72]. Hence, it is clear from the zeta potential analysis that the driving forces for particle aggregation are mainly electrostatic in nature. Moreover, the observed sharp decrease in zeta potential at C_SDS values below the CMC is in accordance with the idea that the coating of SDS is responsible for the Au@16-s-16 aggregation at concentrations well below the CMC. This fact contributes to the destabilization of the colloidal suspension by eliminating nanoparticle repulsions.

Once the changes in the nanoparticle aggregation state as a function of C_SDS concentrations were analyzed in depth using UV-vis, TEM, DLS and zeta potential techniques, the optimization of the biosensor configuration was tackled. The sensor was prepared in two consecutive steps: (i) the formation of aggregated Au@16-s-16/SDS complexes as a starting point, followed by (ii) the disaggregation phenomena of the formed complexes, induced by the nanoparticle/biopolymer interaction which occurred gradually as a function of C_biopolymer. As previously discussed, at concentrations below the CMC for SDS, the direct interaction between the monomeric anionic surfactant and the gold nanoparticle induced the aggregation of the system, leading to a color change from red to blue easily detected by the naked eye. Thus, taking into account the C_SDS concentration needed for the maximum colloid aggregation (C_SDS = 30 µM), the formation of Au@16-s-16/SDS complexes was firstly accomplished. To this end, a mixture of gold nanoparticles and SDS with final concentrations of C_SDS = 30 µM and nanoparticle concentrations of C_Au@16-3-16 = 1.74 × 10⁻⁸ M or C_Au@16-6-16 = 1.38 × 10⁻⁸ M, respectively, was first prepared, giving blue colloidal solutions. The colloid disaggregation phenomenon was then induced by adding increasing amounts of biopolymer to the aggregated Au@16-s-16/SDS complexes, with the changes in the aggregation state of the nanosystem being mediated by the strong interaction between the biopolymer and gold nanoparticles. As a result, the color of the solution gradually changed from blue to red. These changes were accompanied by quantitative changes in the maximum SPR band as a function of C_biopolymer, leading to biopolymer sample quantification. The method suitability was tested using DNA and Lysozyme as examples of model systems for biomolecule recognition, working in a concentration range from 1.0 nM to 0.55 mM and from 1.0 pM to 0.1 mM for DNA and Lysozyme biopolymers, respectively. Two possible addition orders were analyzed for each nanosystem. In method A, the surfactant was first added to the nanoparticle solution, followed by the biopolymer. In method B, the addition order was the opposite:

(A) Au@16-s-16 + SDS + Biopolymer;

(B) Au@16-s-16 + Biopolymer + SDS.

As a result, no significant changes in the color or SPR band position were obtained for DNA detection in method A or lysozyme detection in method B, independent of the biopolymer concentration. Moreover, the gold nanosystem underwent aggregation features in all samples. Therefore, we considered discarding these methods in the explored nanosystems. However, considerable changes were registered for the opposite method in each system, namely for method B for DNA and method A for Lysozyme. The spectrophotometric results are given in Figure 5, Figure 6, and Figure S8.

From these results, it can be deduced that DNA can protect gold nanoparticles against the SDS aggregation, with this protective effect being greater with increasing C_biopolymer. These results are in line with those for the interaction of Au@citrate nanoparticles with DNA in the presence of low NaCl electrolyte concentrations, in which such favorable DNA/Au@citrate interactions protect the nanosystem from salt-induced aggregation [57].

However, the most striking feature of this work is the way in which Lysozyme was able to induce Au@16-s-16/SDS disaggregation following addition method A, where Au@16-s-16 and SDS were previously mixed together. This effect could be due to two possible causes: (i) the interaction between the Lysozyme and SDS, with SDS sequestration in the biopolymer structure hindering the aggregation of the nanosystem, or (ii) the direct Lysozyme/Au@16-s-16 interaction that induces SDS liberation to the bulk solution, including the possibility of induced biopolymer conformational changes. To analyze these possibilities in depth and shed light on the sensing mechanism for DNA and Lysozyme, in the next section, we evaluate both the binding interactions between the species and the structural features of these interactions.

3.2. Binding Interactions and Sensing Mechanism for DNA and Lysozyme Detection

The anionic surfactant SDS interacts with cationic Au@16-s-16 gold nanoparticles but not with DNA [73]. Thus, to avoid SDS-induced nanoparticle aggregation, it is necessary to add the biopolymer to the mixture before adding the anionic surfactant. Hence, optimal colorimetric changes were obtained with addition method B. In contrast to this, in the case of Lysozyme, the situation is completely different: SDS forms a stable complex with lysozyme without causing gross conformational changes in the enzyme molecule [74]. The great strength of the binding is controlled by both the hydrophobic regions and positive charges of the protein, and the hydrophobic tail and the negative charge of the detergent, with an association constant of 514 (M⁻¹) in direct SDS micelles [75,76]. The binding enthalpy determined from the ITC isotherm depends on the SDS concentration: at a low SDS concentration (<3 mM), it was −7.15 kJ/mol, indicating that binding was predominantly via the electrostatic interaction, while at high surfactant concentrations, the nature of binding was cooperative, with the equilibrium binding constant decreasing exponentially with increasing lysozyme concentration [77]. Thus, since Lysozyme is able to interact with SDS, a competitive binding between SDS and Lysozyme for gold nanoparticles could be responsible for the optimal colorimetric changes registered with addition method A. The equilibrium binding constants for biopolymer/Au@16-s-16 can be determined following the absorbance changes at the maximum SPR for Au@16-s-16 nanoparticles upon biopolymer addition using the Hildebrand–Benesi approach [59,60]. According to this approach, the binding of a nanoparticle (Au) to a free biopolymer site (S) to produce an occupied site (AuS) can be expressed by the following reaction:

$$Au+S\stackrel{K}{\leftrightarrow }Au$$

(1)

Denoting the optical absorbance of the nanoparticle in the presence of biopolymer or SDS, the optical response in its absence, and the molar extinction coefficient of the i-th species as A, A₀, and εi, respectively, it was shown that the concentrations of the gold nanoparticle/biopolymer complex and the free nanoparticle were: [AuS] = (A − A₀)/(ε_AuS − ε_Au) = ΔAbs/Δε and [Au] = C_Au − [AuS], respectively. Consequently, both the equilibrium constant, defined as K = [AuS]/([Au] × [S]), and the value of the difference in the extinction coefficients of the nanocomplex and the nanoparticle, ∆ε = ε_AuS − ε_Au, can be evaluated from titration data using iterative fits to an extension of the Hildebrand–Benesi equation [78]:

$$\left(\frac{\text{CAu}\times \text{CS}}{\Delta \mathrm{Abs}}+\frac{\Delta \mathrm{Abs}}{{}^{{\Delta \mathsf{\epsilon }}^2}}\right)=\frac{1}{\mathrm{K}\times \Delta \mathsf{\epsilon }}+\left(\frac{C_{\mathrm{Au}}+{\mathrm{C}}_{\mathrm{S}}}{\Delta \mathsf{\epsilon }}\right)$$

(2)

The values of the binding constants for DNA/Au@16-s-16, Lysozyme/Au@16-s-16 and Au@16-s-16/SDS interactions calculated from Equation (2) are collected in

Table 3. Equilibrium binding constants for the interaction among the reactants.

K (M−1)	Au@16-3-16	Au@16-6-16	SDS
DNA	(1.2 ± 0.5) × 105 (a)	(5.1 ± 0.5) × 104 (a)	− (b)
Lysozyme	(8.40 ± 0.04) × 106 (a)	(9.0 ± 0.5) × 105 (a)	514 (c)
SDS	(2.5 ± 0.2) × 104 (a)	(1.21 ± 0.13) × 104 (a)

(a) Fit to Equation (2) of the absorbance data at the maximum of SPR band. (b) Estimation from reference [73]. (c) Data obtained from reference [75].

; the curve fitting for distinct Au@16-s-16/biopolymer complexes is given in Figure S9.

From these results, it can be observed that gold nanoparticle/biopolymer binding was highly favored for both biopolymers and the associated binding free energy of the interaction was more negative than that corresponding to the SDS/biopolymer interaction. It is important to note that the strength of the interaction was higher in the case of Lysozyme. This fact is in line with colorimetric results, in which DNA biopolymer exerted an important protective effect on SDS-induced Au@16-s-16 aggregation (see Figure 5). However, it does not explain the behavior of the nanosystem based on Lysozyme, in which it was necessary to add the biopolymer to the previously mixed Au@16-s-16/SDS complex (addition method A) to effectively accomplish the disaggregation processes. Thus, the explanation of such behavior could be related to biopolymer conformational changes and must be explored with other structural techniques. In this sense, we used the highly sensitive AFM technique to explore the existence and nature of possible conformational changes induced in the biomolecules. In particular, we selected the system configuration in which we observed complete disaggregation for each nanosystem according to the results described in Figure 5 and Figure 6; taking into account the appropriate order of addition, the results are given in Figure 7 and Figure 8. Figure 7A–C show that when the SDS surfactant was added to the Au@16-3-16/DNA mixture using method B, the APTES-modified mica surface was covered with a large-scale two-dimensional x-y DNA network. In these structures, biopolymer aggregation was mediated by cross-linking features and some condensation events were evident from the existence of globular condensates along the biopolymer chains. Moreover, bright dots coincident with DNA chains could be observed; their height in the z-direction was compatible with the size of the isolated Au@16-3-16, taking into account the typical DNA height in the z-direction. The mean size of the dots from the analysis of more than 200 particles measured in the z-direction was (4.8 ± 1.1) nm. This value can be directly compared with the size of Au@16-3-16 measured with TEM (3.8 ± 0.8) nm and the DNA heights in the z-direction, which had a mean value of 1.2 nm. Given the DNA deformation induced by the force applied to the AFM tip, the DNA height in the z-direction was somewhat smaller than the theoretical values of 2 nm, in accordance with Bustamante’s hypothesis [79]. The similarities between the sizes of the dots coincident with DNA chains suggests that the complexes observed in the z-direction were formed as a consequence of the highly favored Au@16-3-16/DNA interaction, and the gold nanoparticle’s multiple binding sites were easily exposed to the biopolymer. The situation was quite different for the analogous 16-6-16-based nanosystem depicted in Figure 7D–F. That is, the presence of a large-scale DNA network in the x-y direction was replaced by the existence of large aggregates of about 20 nm and (10.3 ± 1.6) nm in the x-y and z-direction, respectively, in which some free DNA chains protruded outward. Moreover, some nanoparticles remained outside the nanocomplex. This fact suggests that the Au@16-6-16/DNA interaction and the distance between the nanoparticles along the DNA chains were not sufficient to maintain the stability of the DNA x-y network, and the biopolymer tended to collapse around Au@16-6-16-bound nanoparticles, forming the aggregates observed in the z-direction.

In the case of the Lysozyme-based nanosystem, favorable lysozyme/SDS binding influenced the nature of the nanostructures observed. Figure 8 shows that given the size of the free Lysozyme proteins, with a particle size of around 2.5 nm [41] and a mean size of the free nanoparticles measured by TEM of less than 5 nm, it is clear that both Au@16-s-16/SDS/Lysozyme nanocomplexes originated aggregates in the z-direction. The size of the particles in the z-direction was very similar among the distinct nanostructures, with an average size of (10.3 ± 1.6) nm and (12.5 ± 1.8) nm for 16-3-16 and 16-6-16-based nanosystems, respectively. However, larger aggregates were observed from the AFM analysis, with a size of about 2 µm in the x-y direction in the case of the Au@16-6-16/SDS/Lysozyme nanosystem; these were much larger compared to the analogous 16-3-16-based nanosystem.

To verify the size of the nanostructures observed using AFM technique, we performed DLS experiments in the presence of biomolecules. The results obtained are given in

Table 4. Values of the hydrodynamic diameters of different Au@16-s-16 nanosystems in the presence of fixed C_SDS = 5.0 × 10⁻⁵ M concentration and biomolecules.

CBiomolecule/M	Au@16-3-16/Biopolymer/SDS	Au@16-6-16/Biopolymer/SDS
0	(748 ± 11) nm	(400 ± 12) nm
CDNA = 1.0 × 10−6	d1 = (220 ± 10) nm, 5% d2 = (43 ± 5) nm, 95%	d1 = (106 ± 14) nm, 2% d2 = (33 ± 7) nm, 98%
CDNA = 5.0 × 10−4	d1 = (450 ± 6) nm, 96% d2 = (50 ± 6) nm, 4%	(11.7 ± 1.3) nm
CLysozyme = 5.0 × 10−8	(615 ± 21) nm	d1 = (342 ± 17) nm, 7% d2 = (79 ± 3) nm, 93%
CLysozyme = 1.0 × 10−4	(44 ± 3) nm	(712 ± 14) nm

and Figure S10. The hydrodynamic values observed in the table reveal the existence of large x-y aggregates for Au@16-3-16/DNA/SDS and Au@16-6-16/SDS/Lysozyme nanosystems, respectively. Moreover, the progress of the Au@16-s-16/SDS disaggregation phenomena with increasing biopolymer concentration could be observed when comparing the value of the hydrodynamic diameter in the absence of the biomolecule with those obtained at very low C_biopolymer concentrations, in which the formation of such structures were still in an early stage (see Table 4).

To summarize, distinct types of aggregates were observed for both DNA- and Lysozyme-based nanosystems: (i) a large-scale network in the x-y direction for the Au@16-3-16/DNA/SDS system without aggregation in the z-direction, (ii) small aggregates with an average size of 20 nm and 10.3 nm in the x-y and z directions, respectively, for the Au@16-6-16/DNA/SDS system, (iii) small aggregates in the z-direction of 10.3 nm in size for the Au@16-3-16/SDS/Lysozyme nanosystem, and (iv) large aggregates of 12.5 and 2 µm in the z and x-y directions, respectively, for the Au@16-6-16/SDS/Lysozyme nanosystem.

However, there were certain similarities among the aggregates depending on the type of gemini surfactant that integrated into the nanosystem. Specifically, nanosystems based on 16-3-16 gemini surfactant were smaller in size in the z-direction, and the distance between the neighboring Au@16-3-16 nanoparticles was greater than in the case of the homologous 16-6-16 nanosystem. In fact, this special spatial distribution of the nanoparticles among the complexes could contribute to the differences in the absorbance response pattern observed in Figure 5 and Figure 6, in which the amplitude of absorbance changes in the aggregate band was more evident for 16-3-16-based nanosystems in both biopolymers. On the other hand, focusing again on the type of surfactant used to functionalize the nanoparticle, it can also be noted that the 16-3-16 derivatives produced more stable Au@16-s-16/biopolymer complexes (see Table 3). Therefore, the strength of the binding Au@16-s-16/biopolymer, the nature of the aggregates, and the distance between the nanoparticles in the nanocomplexes could directly influence the aggregation state of the original nanosystem and, consequently, the sensitivity of the sensor. To correlate these aspects with the sensitivity of the sensor, the changes observed in the SPR experimental spectra of Au@16-s-16/SDS systems in the presence of biomolecules were analyzed in depth using deconvolution procedures (see Figure S11). The results are summarized in

Table 5. Deconvolution parameters obtained from Au@16-s-16 spectra in the presence of DNA and SDS following the addition method B (Au@16-s-16 + DNA + SDS). Gold nanoparticle and SDS concentrations were fixed in all experiments (C_Au@16-3-16 = 0.174 nM, C_Au@16-6-16 = 0.138 nM, and C_SDS = 30 µM).

Au@16-3-16/DNA			Au@16-6-16/DNA
CDNA(M)	λ1, max/nm (Area1)	λ2, max/nm (Area2)	CDNA/M	λ1, max/nm (Area1)	λ2, max/nm (Area2)
0	548 (1.94)	651 (10.17)	0	551 (1.82)	650 (10.84)
5.0 × 10−8	548 (1.68)	642 (11.08)	5.0 × 10−8	547 (1.87)	639 (10.01)
1.0 × 10−7	545 (1.65)	640 (10.73)	1.0 × 10−7	545 (1.70)	638 (10.15)
5.0 × 10−7	544 (1.66)	638 (10.71)	5.0 × 10−7	544 (1.73)	634 (9.51)
1.0 × 10−6	543 (2.02)	620 (9.17)	7.5 × 10−7	542 (1.71)	632 (9.93)
5.0 × 10−6 (*)	----	539	1.0 × 10−6	542 (1.73)	629 (9.31)
			5.0 × 10−6	----	538

(*) Data specification for the first concentration at which the deconvolution process is not necessary for each system.

and

Table 6. Deconvolution parameters obtained from Au@16-s-16 spectra in the presence of Lysozyme and SDS following the addition method A. Gold nanoparticle and SDS concentrations were fixed in all experiments (C_Au@16-3-16 = 0.174 nM, C_Au@16-6-16 = 0.138 nM, and C_SDS = 30 µM).

Au@16-3-16/Lysozyme			Au@16-6-16/Lysozyme
CLysozyme(M)	λ1, max/nm (Area1)	λ2, max/nm (Area2)	CLysozyme/M	λ1, max/nm (Area1)	λ2, max/nm (Area2)
0	548 (1.94)	651 (10.17)	0	551 (1.82)	650 (10.84)
1.0 × 10−9	545 (1.92)	640 (12.83)	1.0 × 10−9	547 (1.71)	637 (10.44)
5.0 × 10−9	545 (1.96)	639 (9.20)	5.0 × 10−9	543 (1.71)	629 (9.18)
1.0 × 10−8	545 (2.06)	637 (10.60)	7.5 × 10−9	541 (1.86)	628 (9.36)
5.0 × 10−8	544 (2.47)	631 (8.50)	1.0 × 10−8	542 (1.76)	627 (9.21)
1.0 × 10−7	543 (2.27)	623 (8.92)	5.0 × 10−8	541 (1.73)	625 (9.44)
5.0 × 10−7 (*)	----	553	1.0 × 10−7	541 (1.78)	624 (9.62)
			5.0 × 10−7	541 (1.68)	616 (8.47)
			1.0 × 10−6	539 (1.70)	614 (7.07)
			5.0 × 10−6	534 (1.84)	604 (7.21)
			1.0 × 10−5 (*)	----	586

(*) Data specification for the first concentration at which the deconvolution process is not necessary for each system.

.

Again, two characteristic bands were displayed when the deconvolution procedure was applied; one can be assigned to nonaggregated particles, designated λ₁, and the other to aggregated nanosystems, designated λ₂, which diminished as the C_biopolymer increased in each case. Taking into account the values of the peak area assigned to each wavelength λ₁ and λ₂, it seems evident that, at low biopolymer concentrations, the main contribution to the SPR band corresponded to the aggregated systems. Subsequently, the magnitude of the nonaggregated peak area gained importance until a certain concentration was reached, at which point a unique SPR band appeared.

As illustrated in Figure 9, the plots of ln (1/Δλ₂) vs. ln(C_biopolymer/Δλ₂) gave a linear calibration curve for all the nanosystems regardless of the surfactant used or the biopolymer detected.

The analytical method developed in water media was validated using horse serum media to assess the possible matrix effect in the sensor response. Figure 10 shows the plot of ln (1/Δλ₂) vs. ln(C_biopolymer/Δλ₂), giving a linear response in the same range of concentrations verified in water for all the studied nanosystems. These results support the selectivity and validity of the method due to the presence of multiple components in the serum media that can act as possible interferences in the measurement process.

However, looking at the values of the ordinate (y₀) and the slope (m) of the curves collected in

Table 7. Values for limit of detection (LOD), limit of quantitation (LOQ), and linear regression parameters (slope = m and y intercept = y_o and standard deviation = SD) for DNA and Lysozyme quantification based on colorimetric response obtained from data in Figure 9 and Figure 10.

System	m	yo	SD	LOD	LOQ
Au@16-3-16/DNA/SDS (water)	−0.519 ± 0.02	−5.80 ± 0.10	0.194	8 nM	24 nM
Au@16-6-16/DNA/SDS (water)	−0.498 ± 0.02	−0.570 ±0.10	0.198	9 nM	27 nM
Au@16-3-16/SDS/Lysozyme (water)	−0.44 ± 0.02	−5.77 ± 0.11	0.185	0.12 nM	0.37 nM
Au@16-616/SDS/Lysozyme (water)	−0.304 ± 0.017	−4.89 ± 0.08	0.109	8.5 pM	25.7 pM
Au@16-3-16/DNA/SDS (horse serum)	−0.966 ± 0.05	−10.4 ± 0.30	0.230	0.24 nM	0.73 nM
Au@16-6-16/DNA/SDS (horse serum)	−0.846 ± 0.04	−9.45 ± 0.30	0.240	0.50 nM	1.52 nM
Au@16-3-16/SDS/Lysozyme (horse serum)	−0.581 ± 0.03	−8.51 ± 0.25	0.140	3.1 pM	9.5 pM
Au@16-6-16/SDS/Lysozyme (horse serum)	−0.322 ± 0.015	−5.99 ± 0.14	0.143	3.5 pM	11 pM

, it is clear that the specific colorimetric response was distinct for each nanosystem and the type of media used, in such a way that given the slope values, the best sensitivity was obtained for the Au@16-3-16/DNA/SDS nanosystem (see Figure 9A and Figure 10A), and in general, when the 16-3-16 gemini surfactant was used. Moreover, the detection limit (LOD) and quantification limit (LOQ) values were calculated as LOD = 3.3 × σ/m and LOQ = 10 × σ/m, where σ is the standard deviation of the response [80].

Taking into account these values for each nanosystem, again, the colorimetric response suitability was presented in the following order: Au@16-3-16/SDS/Lysozyme > Au@16-6-16/SDS/Lysozyme > Au@16-3-16/DNA/SDS > Au@16-6-16/DNA/SDS. As a result, the Au@16-3-16-based nanosystem presented a better sensing response for each biopolymer studied here. Finally, a comment about the variation in the sensor response in the presence of horse serum seems to be pertinent. Note that some of the serum components were even biomolecules of high molecular weight such as gamma globulin, lactose dehydrogenase, or albumin. Additionally, there were other smaller biomolecules such as creatinine or bilirubin that could also interfere in the measurement. However, we demonstrated that if the concentration of serum is fixed in the measurement, the sensor response varies but retains its linearity in the same range of DNA and Lysozyme concentrations, supporting the validity of the method. Thus, correlating the colorimetric response with the thermodynamic and structural features analyzed at the beginning of this section, we can conclude that: (i) the greater the biopolymer–nanoparticle interaction is, the better the sensor response obtained; (ii) if the biopolymer/SDS interaction is negligible, it is better to add the biopolymer to the mixed Au|@16-s-16/SDS complex according to addition method A; (iii) the presence of larger aggregates in the z-direction is unfavorable for the sensitivity of the method; (iv) the higher the interparticle distance in the nanocomplex is, the better the sensor response and sensitivity obtained. Note that maximum nanoparticle separation was obtained in the specific case of the formation of cross-linking aggregates in the x-y direction.

4. Conclusions

Biopolymer sensing was controlled following Au@16-s-16/DNA/SDS or Au@16-s-16/Lysozyme/SDS complex disaggregation using the UV-visible spectroscopic technique. The disaggregation procedure was accompanied by changes in the SPR band and color changes from blue to red that were visible to the naked eye, with a limit of detection in the picomolar range for lysozyme sensing.

Thermodynamic and structural studies using UV-visible spectroscopy, DLS, and AFM techniques served to gain insight into better nanosystem configuration, in which factors such as favorable Au@16-s-16/biopolymer binding, the absence of z-aggregates, and the interparticle distance among nanoparticles in aggregates are key parameters to be optimized and controlled for a better sensor response. In this sense, a better linear response was observed for both Au@16-3-16/DNA/SDS and Au@16-3-6/SDS/Lysozyme nanosystems, highlighting, from a structural perspective, the formation of cross-linking aggregates that favor both the distance between nanoparticles in the complexes and the absence of z-aggregates. These facts demonstrated that controlling the structural and binding features among the components of the nanosensor is crucial for the effective sensing of biomolecules using colorimetric methods.

Finally, the colorimetric method developed was shown to be simpler, faster, and more sensitive compared to other previously reported colorimetric, fluorimetry, mass spectrometry, chromatography, gel electrophoresis, and electrochemical methods. Moreover, the new approach resolves problems associated with time lapse determination, especially in the case of Lysozyme, which usually requires a long incubation period for sample quantification.