TMB+-mediated etching of urchin-like gold nanostructures for colorimetric sensing

The morphology-dependent localized surface plasmon resonance of gold nanostructures has been widely utilized for designing sensors. One method relies on the color change of gold nanoparticles upon etching. In previous work, TMB2+ oxidized from 3,3′,5,5′-tetramethylbenzidine (TMB) was found to etch gold nanorods (AuNRs), leading to a spectrum of different colors. However, the preparation of TMB2+ needs the addition of a strong acid and other harsh conditions. Herein, a new colorimetric biosensing platform was developed using urchin-like gold nanoparticles (AuNUs). Compared with AuNRs, the etching of AuNUs can happen under mild conditions by TMB+ at pH 6, protecting enzymes and proteins from denaturation. The role of CTAB surfactant was dissected, and its bromide ions were found to be involved in the etching process. Based on these observations, a one-step colorimetric detection of H2O2 was realized by using horseradish peroxidase and H2O2 to oxidize TMB. Within 30 min, this system achieved a detection limit of 80 nM H2O2. This work offered fundamental insights into the etching of anisotropic gold nanostructures and optimized the etching conditions. These advancements hold promise for broader applications in biosensing and analytical chemistry.


Introduction
Gold nanoparticles (AuNPs) possess much higher extinction coefficients than organic dyes due to their localized surface plasmon resonance (LSPR), allowing visual observation at low nanomolar and even picomolar concentrations [1,2].The position of LSPR peaks is dependent on the size and morphology of AuNPs.Therefore, colorimetric detection can be made based on morphology changes of AuNPs, especially anisotropic AuNPs [3].Over the last few decades, anisotropic AuNP etching-based sensors were prevalent [4].For example, the etching of gold nanorods (AuNRs) can happen along the longitudinal direction, leading to a continuous color change [5][6][7].Interestingly, a high concentration of cetrimonium bromide (CTAB) appeared to be essential for etching AuNRs.In the presence of CTAB, the redox potential of AuBr 2− /Au 0 (0.93 V versus NHE) can be dramatically decreased by the formation of the AuBr 2− -(CTA) 2+ /Au (<0.2 V versus NHE) [8][9][10].The addition of hydrogen peroxide (H 2 O 2 ), acids, and O 2 can also help the oxidation or etching of AuNRs.However, a high CTAB concentration (usually over 50 mM) and extreme conditions (e.g.250 mM HCl or 45 °C) limited the applications of AuNRs [9,11,12].Therefore, we wish to explore other anisotropic gold nanostructures that might be etched more easily under mild conditions.
Urchin-like AuNPs (AuNUs) are important anisotropic gold nanostructures for a broad range of applications from biosensors to cancer therapy [13][14][15][16][17]. AuNUs can grow on AuNP seeds when additional Au 3+ ions are reduced by sodium citrate and hydroquinone [18,19].The tip areas on the AuNUs are highly reactive because of their high surface energy [20,21], and the sharp tips exhibit larger electric fields at their concavo-convex sites compared to neutral curvature areas [22].With these sharp tips, morphological changes of AuNUs can be triggered easily.For example, upon laser irradiation, AuNUs can melt into spherical AuNPs [23].
Enzyme-linked immunosorbent assay (ELISA) based detection systems have been widely used for the measurement of various kinds of disease biomarkers [24][25][26].3,3′,5,5′tetramethylbenzidine (TMB) is an important substrate that can be oxidized to TMB + (blue) or TMB 2+ (yellow) for colorimetric immunoassays.It was reported that TMB 2+ can efficiently etch AuNRs, which converts the color of TMB 2+ to a color change of AuNRs to increase the sensitivity of detection [5].However, as mentioned above, acids and heating were needed for this reaction to occur.
In this work, we studied TMB + -induced etching of AuNUs, and comparisons were made with AuNRs.In particular, we tried to understand the role of CTAB, and dissected its effect based on its surfactant part and its halide part.In the presence of a low concentration of CTA + and Br − ions, TMB + can efficiently etch the branches of AuNUs.As a result, the morphology change of AuNUs was accompanied by a vivid color variation.Based on these understandings, we used TMB + -induced etching of AuNUs to design a highly sensitive colorimetric biosensor for H 2 O 2 detection under ambient conditions.

TEM, ζ-potential measurements, and UV-vis spectroscopy
Transmission electron microscope (TEM) images were captured using a 100 kV Phillips CM10 TEM.ζ-potentials were assessed using dynamic light scattering (Zetasizer Nano 90, Malvern).In a standard trial, AuNUs-38 NPs were dispersed in 1 ml buffer with different pHs.The determination of ζ-potential values was conducted at 25 °C.UV-vis absorption spectroscopy measurements were conducted using an Agilent 8453 A UV-vis spectrometer.

Preparation of spherical AuNP seeds
Citrate-capped Au seeds were synthesized according to the literature [27,28].Briefly, the 100 ml 1 mM HAuCl 4 was heated and boiled for 30 s.Then, 10 ml of 38.8 mM sodium citrate was added quickly.The mixture solution changed from light yellow to wine red in 2 min.The 13 nm AuNPs were obtained after refluxing for another 20 min.The same protocol was used to prepare 38 nm Au seeds, except that the concentration of HAuCl 4 was doubled.

Preparation of AuNUs
First, 30 mM hydroquinone was freshly prepared with Milli-Q water and used the same day.For a typical synthesis, 1 ml HAuCl 4 was diluted with 180 ml H 2 O under vigorous stirring.Subsequently, 600 μl AuNP seeds, 3 ml 38.8 mM sodium citrate, and 10 ml 30 mM hydroquinone were added sequentially.The 13 nm and 38 nm spherical Au seeds were, respectively, used for generating AuNUs-13 and AuNUs-38 NPs.The solutions were incubated at room temperature for 30 min under stirring.In the end, the resulting AuNUs were washed with 5 mM pH 6 phosphate buffer at 4000 rpm for 8 min and stored at 4 °C for further use.The morphologies of AuNUs were characterized by TEM (Phillips CM10 100 kV) and UV-vis spectroscopy.

Preparation of AuNRs
AuNRs were prepared by a seed-mediated method using a binary surfactant system as reported by Murray [29].For seed preparation, 5 ml 0.5 mM HAuCl 4 was added into 5 ml 0.2 M CTAB solution in a 20 ml scintillation vial.Then, 0.6 ml of ice-cold fresh 0.01 M NaBH 4 was diluted to 1 ml with water and injected into the HAuCl 4 -CTAB mixture under rapid stirring (1200 rpm).After stirring for 2 min, the color of the solution changed from yellow to brown, and the seed solution was used after standing for 30 min at room temperature.To prepare a growth solution, 7.0 g CTAB and 1.234 g sodium oleate (NaOL) were dissolved in 250 ml of warm water (50 °C) in an Erlenmeyer flask.After the solution was cooled to 30 °C, 18 ml 4 mM AgNO 3 was added to the solution under stirring.The mixture was kept undisturbed at 30 °C for 15 min, and then 250 ml of 1 mM HAuCl 4 was added and stirred for another 90 min.1.5 ml HCl (37 wt% in water) was further added into the solution and stirred for 15 min.Afterward, 1.25 ml of 0.064 M ascorbic acid (AA) was added to the solution under vigorously stirred for 30 s. Finally, 0.4 ml seed solution was injected into the growth solution with stirring for 30 s.The growth solution was left undisturbed for 12 h.Finally, the AuNRs were centrifuged at 7000 rpm for 30 min to remove excess emulsifier and washed once with water.The final AuNRs were stored in 5 mM CTAB at 4 °C.
Preparation of TMB + and TMB 2+ 2 ml 0.5 mM TMB substrate in 5 mM pH 4 buffer solution was irradiated under UV light (∼370 nm) for 30 min to get blue TMB + .The final concentration of TMB + was determined by the absorbance of TMB + at 652 nm with an extinction coefficient ε of 3.9 × 10 4 M −1 cm −1 [30].TMB 2+ was prepared by mixing TMB + solution and 250 mM H 2 SO 4 with a 1:1 volume ratio.The final concentration of TMB 2+ was determined by the absorbance of TMB 2+ at 450 nm with ε of 5.9 × 10 4 M −1 cm −1 .

Etching of AuNUs/AuNRs by TMB +
In a typical etching experiment, 80.5 ml H 2 O, 7.5 ml 100 mM CTAC, 30 ml 100 mM pH 6 phosphate buffer, 20 ml AuNUs/ AuNRs and 3 μl 500 mM NaBr were added into microtubes in sequence.Then, 9 μl TMB + of various concentrations were pipetted into the microtubes, respectively.The final volume of samples was 150 μl.After vigorous stirring for 30 s, the samples were incubated at room temperature for 30 min before UV-vis absorbance measurements.
For the surfactant impact study, AuNUs were individually incubated with 0.1% Triton X-100, 0.1% Tween 20, 0.1% Tween 80%, and 0.1% CTAB.Subsequently, shifts in SPR peak positions (Δλ) were calculated based on the absorbance spectra.To investigate the impact of halides, NaBr in a standard trial was replaced with either NaCl or NaF.CTAC was employed in all trials exploring halide effects.In optimizing etching time and pH, the kinetics of etching were recorded using a TECAN SPARK microplate reader.
First, a 123 μl mixture solution was prepared with 1.5 μl 10 mM TMB substrate, 15 μl 100 mM pH 6 phosphate buffer, 7.5 μl 100 mM CTAC, and 99 μl H 2 O. Subsequently, 1 μl 0.1 mg ml −1 HRP was added into the mixture solution in a 96-well plate, followed by the addition of 20 μl AuNUs-13 NPs, 3 μl 500 mM NaBr, and 3 μl various concentrations of H 2 O 2 .The total volume in each well was 150 μl.The final concentrations of TMB substrate, phosphate buffer, and CTAC were 100 μm, 20 mm, and 5 mm, respectively.Then, the mixture solution was incubated at room temperature for 30 min.Finally, the UV-vis absorbance was monitored by a plate reader, and the color of the samples was recorded using a digital camera.

Results and discussion
TMB + -mediated etching of AuNUs First, we aimed to study the etching reaction of AuNUs by TMB + .The AuNUs were synthesized by a seed-mediated growth method.Spherical AuNPs, which were prepared by citrate reduction, were used as seeds.For AuNP seeds of around 38 nm, the resulting urchin-like products were called AuNUs-38.
AuNUs grown from 13 nm seeds were also made and named AuNUs-13.From the TEM image shown in figure 1(A), these AuNUs-38 had multiple sharp tips, and most of the AuNUs-38 were between 100 and 130 nm (figure 1(D)).Compared with the spherical AuNP seeds with a surface plasmon resonance (SPR) peak at 526 nm (figure S1), the SPR peak of AuNUs-38 showed a large red-shift, yielding a blue solution.These two sizes were chosen to study the effect of size.13 nm AuNPs are the most common, 38 nm has tripled the size, and even larger seeds would result in AuNUs that are easily precipitated.
These sharp edges in AuNUs have higher surface energy, and they might be more easily etched.TMB is a common chromogenic substrate and upon one-electron oxidation, the TMB + product has blue color.We hope to use TMB + to etch the AuNUs to amplify the color change since AuNUs have much higher extinction coefficients compared to TMB + .To avoid potential effects of other molecules, we produced TMB + by using UV irradiation (so no H 2 O 2 or HPR was added) [5].It needs to be noted that this method only yielded around 11% of TMB + , while the rest 89% were still unreacted TMB substrates (figure S2).
With AuNUs and TMB + prepared, TMB + -mediated etching of the AuNUs was then studied.In the presence of 5 mM CTAB, the sharp tips of AuNUs-38 were etched and rounded by the added TMB + (figure 1(B)).The blue color of the AuNUs-38 solution turned red in 30 min.Although the etching products were still not fully spherical, a significant blue shift of SPR peaks (from 691 to 603 nm) was observed (figure 1(C)).From the histogram in figure 1(D), the size of etched AuNUs-38 NPs was decreased from 110 to around 90 nm.

Comparison of etching of AuNUs and AuNRs
Since previous work mainly used AuNRs, the etching of AuNRs and AuNUs was then compared.These gold nanostructures were incubated with various TMB + concentrations at room temperature.Interestingly, the AuNRs were quite stable under the etching conditions for AuNUs (figures 2(B), (C)).Even after a 24 h incubation, no absorbance peak shift happened for AuNRs, while obvious color change occurred for the AuNUs (figure 2(D)).To etch the AuNRs, a high CTAB concentration (50 mM), strong acids, TMB 2+ , and a high temperature of 80 °C were all required (figure S3(B)).Even under such a harsh condition, the shift in the absorption spectra was small, and no obvious color change was observed in 1 h (figure 2(E)).We expected that the AuNRs could be more difficult to etch since they lacked sharp branches with a high surface energy (figure S3(A)).Therefore, while AuNUs could be etched using TMB + , the etching of AuNRs required TMB 2+ (figure 2(A)).TMB 2+ etched AuNRs along their longitudinal direction, leading to a continuous color change [4].

Effect of the size of AuNUs
We also studied AuNUs-13, which were smaller than AuNUs-38 in size, but with the same morphology.By comparing these two distinct nanoparticle sizes, we aimed to elucidate size-dependent effects on the etching mechanism and the rate of colorimetric responses.After 1 h incubation, color change happened for both AuNUs-38 and AuNUs-13 (figure 2(B)).Although AuNUs-13 with smaller sizes showed a smaller SPR peak shift, a much more obvious color change was obtained (figure 2(C)).For subsequent studies, we opted for the larger AuNUs for conducting quantitative spectroscopic measurements due to their larger SPR peak shifts.At the same time, the smaller AuNUs were used for colorimetric sensing purposes.

Effects of halides and surfactants
To achieve the best etching performance, the effects of reaction conditions, such as pH, etching time, surfactants, and halide ions were then investigated.The shift in SPR peak position (Δλ) was chosen to indicate the extent of etching of the AuNUs.We started by optimizing surfactants.The as-synthesized AuNUs were mainly covered by weakly adsorbed citrate groups, which can be easily displaced by stronger capping agents [31,32].Initially, cetrimonium bromide (CTAB) was used as a capping agent since CTA + has been well studied in AuNR etching [8].In addition, a few common surfactants were also evaluated including Triton X-100, Tween 20, and Tween 80. Interestingly, TMB + only etched the AuNUs in the presence of CTAB (figure 3(A)).We also noticed that the etching was faster in higher CTAB concentrations (figure S4).
CTAB contains Br − as its counterion, and Br − has a strong affinity to the gold surface [33].Halide counterions of the surfactant have been proved to be critical in the oxidation of CTAB-capped AuNRs [34,35].Halide adsorption was widely studied for shape-controlled AuNP synthesis [36].It is known the adsorbed concentrations of Br − and I − are much higher than that of Cl − ions on the Au surface [37,38], and the interaction strength of halides with gold has the trend of I − > Br − > Cl − .Thus, we expected that Br − might be important during the etching process.Interestingly, without CTA + , Br − alone failed to etch (figure 3(B)).Therefore, CTA + appeared to be required for AuNUs etching as well.
To study the role of halide counterions for deeper mechanistic understanding, we conducted more etching experiments with a mixture of cetrimonium chloride (CTAC) and three halides (F − , Cl − , and Br − ).I − was omitted here because a low concentration of iodide can cause serious etching of AuNUs without the help of CTA + groups (figure S5).With 5 mm CTAC and up to 50 mm F − or Cl − , no SPR peak shifts were observed after incubation with 4 μm TMB + for 30 min (figures 3(C), (D)).These trends are consistent with the etching of AuNR and Au nanostars [20,39].Therefore, Br − and CAT were both essential for etching the AuNUs.The importance of Br − was also confirmed by conducting the etching experiment in a mixture of 0.25 mm CTAB 0.75 mM CTAC, 0.5 mM CTAB 0.5 mm CTAC, and 0.75 mm CTAB 0.25 mm CTAC (figure S6).A larger Δλ was generated with a higher portion of CTAB, again indicating the involvement of Br − in the etching process.To achieve a faster etching speed or a larger SPR peak shift, a higher CTAB concentration was needed.However, a too-high CTAB concentration may generate many bubbles, which can cause problems for quantitative absorbance measurements.Also, CTAB has poor solubility at room temperature.Therefore, we need to achieve fast etching with the lowest possible surfactant concentration.This goal might be achieved by using a mixture of CTAC (with the Br − in CTAB replaced by Cl − ) and NaBr.To test this hypothesis, the CTAC concentration was fixed at 5 mm (10-folder lower than the typical 50 mm CTAB concentration), and larger SPR peak shifts occurred with more NaBr added (figure 3(E)).To avoid the aggregation of the AuNUs, 10 mm NaBr concentration was chosen for subsequent work (figure S7).With 10 mM NaBr, larger SPR peak shifts were observed with higher CTAC concentrations (figure 3(F)).When the CTAC concentration was higher than 2 mm, negligible improvements were obtained.This can be explained by that the critical micelle concentrations (CMC) of CTAB and CTAC are both around 1 mm at room temperature [40,41].Therefore, we chose to use 5 mm CTAC in this work.

Optimization of etching time and pH
After understanding the effects of surfactants and Br − , the effect of etching time and pH conditions were further investigated.Figure 4(A) shows that the etching was fast in the initial 20 min, after which the shift of the SPR peak was much slower.In 1 h, a nearly 90 nm shift was observed.For the control sample (TMB substrate), only a minor increase in Δλ (9 nm) after even 1 h.As a result, all the samples were incubated for 30 min to obtain a large and relatively stable difference in Δλ.
We then studied the effect of pH on the etching.An acidic environment was shown to facilitate AuNP etching in the presence of dissolved oxygen [9,42].However, many proteins are only stable in a narrow pH range near neutral.For example, horseradish peroxidase (HRP) loses its structural and conformational stability at pH below 4 [43].Thus, the etching of AuNUs-38 was studied between pH 4 and 8. From figure 4(B), at pH 4, the Δλ of the TMB control sample was very close to TMB + .The color difference of these two samples was indistinguishable with such close absorbance peak positions.This result suggested that the etching was mainly caused by the low pH, and the detection of TMB + was difficult at pH 4. The largest Δλ difference was observed at pH 6.When the pH was higher than 6, the etching activity of TMB + decreased dramatically.This can be explained by the low stability of TMB + at higher pH (figure 4(C)).Therefore, pH 6 was chosen as the optimal pH for further studies.In terms of the surface charge of AuNUs-38, it was all negatively charged from pH 4-8 (figure 4(D)), and thus charge should not be the reason for the difference in etching at different pH.

Visual detection of H 2 O 2
So far, we have gained a fundamental understanding of this etching reaction and optimized the etching conditions.Herein, using this reaction, we aimed to demonstrate a proof-ofconcept analytical application for the detection of H 2 O 2 .H 2 O 2 is an important by-product of many enzymatic reactions and has been used as a target molecule of many biosensors [44][45][46].For example, the oxidation of glucose by glucose oxidase (GOx) can produce H 2 O 2 , which is the actual target of electrochemical glucose meters.
To realize visual detection, AuNUs synthesized from smaller 13 nm AuNP seeds were used (figure 5(A)), since a more vivid color change was generated (figure 2(B)).The experimental setup was similar to the conventional colorimetric ELISA.Horseradish peroxidase (HRP) enzymes can catalyze H 2 O 2 to produce a reactive radical species (•OH).Then •OH was reacted with TMB substrate to generate TMB + (figure 5(B)).Thanks to the mild conditions for AuNUs etching, HRP-H 2 O 2 catalysis, and AuNUs etching can be realized in one step.After TMB + -mediated etching, an obvious blue shift in the SPR peak was observed (figure S8).The etching reaction also happened fast in just 20 min, and the control samples were very stable at pH 6 (figure S9).For H 2 O 2 sensing, only the samples with HRP displayed a color change (from blue to red).As shown in figure 5(C), with 400 nM or 600 nM H 2 O 2 , no obvious color change was observed.However, such low H 2 O 2 concentrations did lead to an impressive color change in the samples with the AuNUs.It is well known that H 2 O 2 can oxidize AuNRs in the presence of Br − under acid conditions at high temperatures [42].However, the concentration used for AuNRs oxidation is much higher than that used in our experiments.Over 1 mM H 2 O 2 is required to cause slight etching of AuNUs (figure S10).
This sensor demonstrates exceptional sensitivity, manifesting a substantial color change at H 2 O 2 concentrations higher than 400 nm.In this study's conditions, the limit of detection for H 2 O 2 was determined to be 80 nm (3σ/slope, inset of figure 5(D)).Notably, this represents one of the most sensitive colorimetric sensors for H 2 O 2 .For example, this sensor has approximately a 7-fold increase in sensitivity compared to a previous colorimetric detection method [47].

Conclusions
In summary, we reported rapid etching of AuNUs by TMB + under mild conditions.The reaction conditions including time, pH, and surface ligands were also optimized.All CTA + , TMB + , and Br − molecules were important in etching the AuNUs.Although the blue color of TMB + at low concentrations failed to be discerned with the naked eyes, nanomolar level TMB + can still cause a vivid color change of AuNUs by etching, thanks to the extremely high extinction coefficients of the AuNUs.With these observations and under optimized conditions, we developed a one-step colorimetric biosensing platform for H 2 O 2 detection by TMB + -mediated etching of AuNUs.This work has expanded the TMB + -mediated etching of gold nanostructures and is useful for improving plasmonic biosensors.

Figure 1 .
Figure 1.TEM micrographs of AuNUs-38 NPs (A) before and (B) after the addition of TMB + .(C) UV-vis spectra of AuNUs-38 before and after etching.Inset photos are the color of the corresponding samples.(D) The size distribution histograms of AuNUs-38 NPs before and after etching.

Figure 2 .
Figure 2. (A) Schematic illustration of the etchings of AuNUs and AuNRs by TMB + and TMB 2+ , respectively.(B) Color change of three types of gold nanostructures incubated with TMB + .(C) SPR peak shifts of AuNRs, AuNUs-38, and AuNUs-13 after 1 h incubation with various TMB + concentrations.In these etching experiments, 5 mm CTAC and 10 mM NaBr were used.Absorption spectra of AuNRs reacted with TMB 2+ in the presence of (D) CTAC, and (E) CTAB. 10 μm TMB 2+ was used in these experiments.

Figure 3 .
Figure 3. (A) SPR peak shifts of AuNUs-38 etched in the presence of 0.1% surfactants.0.1% CTAC and CTAB were 3.1 mm and 2.7 mm, respectively.(B) Normalized absorption spectra of AuNUs incubated with mixtures of various NaBr concentrations and 4 μm TMB + .The effects of (C) F − and (D) Cl − on the etching of AuNUs-38 in the presence of 5 mm CTAC and 4 μm TMB + .(E) SPR peak shifts of AuNUs-38 NPs as a function of NaBr concentration.In these experiments, 5 mm CTAC was used.(F) SPR peak shifts of AuNUs-38 etched in the presence of various concentrations of CTAC and fixed 10 mm NaBr.

Figure 4 .
Figure 4. (A) The etching kinetics of AuNUs-38, where SPR peak shifts were monitored.(B) The SPR peak shifts of AuNUs-38 incubated in different pH environments.Acetate buffers were used for pH 4-5; phosphate buffers were used for pH 6-8.(C) The stabilities of TMB + produced by UV light at different pHs.(D) Zeta-potentials of AuNUs-38 at different pH values.

Figure 5 .
Figure 5. (A) A TEM image of AuNUs synthesized from 13 nm spherical AuNP seeds.(B) Schematic illustration of etching of AuNUs induced by the peroxidation product of HRP-catalyzed TMB.(C) Color changes of the proposed method with the increase of H 2 O 2 concentration.(D) LSPR shifts of AuNUs-13 as a function of H 2 O 2 concentration.Inset: the response at a low concentration range.