Gold nanoparticles have long announced their birth with a dramatic change in colour, shifting from pale yellow to deep ruby red as a reaction unfolds. For decades, scientists have often used that transformation as a quick visual sign that the Turkevich synthesis—the classic citrate-based method for producing spherical gold nanoparticles—has worked. Now, researchers in Indonesia have turned the same familiar colour change into a detailed kinetic record, showing that an ordinary video camera can reveal multiple stages of nanoparticle formation that would otherwise require sophisticated laboratory instruments.
In a study published in the Journal of Nanoparticle Research, Chairul Ichsan and colleagues analysed a 601-second recording of tetrachloroauric acid being reduced by citrate. Instead of treating the video as a simple before-and-after demonstration, the team examined it frame by frame, extracting numerical information from the changing appearance of the solution. Every frame became a data point in a time-resolved experiment. The researchers converted the images into ten different colour representations, including linear RGB, the broadcast-oriented YUV, YIQ and Y′CbCr systems, and perceptual models such as CIELAB, CIELCh, Oklab and Oklch. They also calculated the structural similarity index, or SSIM, to track how closely each frame resembled a stable final-state reference.
The result was a visual “fingerprint” of the reaction. Rather than one continuous colour transition, the data revealed five distinguishable kinetic phases. The first involved the fading of the initial yellow colour as citrate began reducing the gold precursor. This was followed by a nucleation-linked collapse in lightness, as newly formed nanoscale gold clusters altered the way the solution absorbed and scattered light. A dark, pre-plasmonic latency period came next, during which the reaction continued but did not yet produce the characteristic red appearance associated with gold nanoparticle surface plasmons. The fourth stage was the rapid growth of the red colour, and the final phase represented maturation as the optical properties approached a steady state.
These stages matter because the colour of a gold nanoparticle suspension is not merely decorative. It is a consequence of how electrons on the nanoparticle surfaces interact with incoming light. Gold nanoparticles support a collective oscillation of their conduction electrons known as localized surface plasmon resonance. For relatively small, spherical particles, this resonance commonly falls in the visible region of the spectrum and produces the intense ruby-red colour familiar from colloidal gold. Changes in particle size, concentration, aggregation and local dielectric environment can shift or broaden that optical response. The video therefore captures a complex combination of chemical reduction, nucleation, particle growth and evolving light–matter interactions.
By fitting mathematical kinetic models to the colour trajectories, the researchers separated several characteristic times. The reduction process had a half-time of approximately 21 seconds, indicating that the gold precursor began changing rapidly. Nucleation occurred later, with a midpoint near 57 seconds. The red-growth stage was substantially slower on the overall reaction clock, reaching its half-time at roughly 240 seconds. The researchers found that the growth data were consistent with the Finke–Watzky model, which describes slow, continuous nucleation followed by rapid autocatalytic growth on existing particle surfaces. A ratio of approximately 47 between the autocatalytic and nucleation terms provided the study’s clearest kinetic signature: once suitable gold surfaces existed, further growth became much more efficient.
One of the most striking findings came from comparing colour information with SSIM. SSIM is widely used in image analysis to quantify structural changes by comparing luminance, contrast and local patterns between images. When the researchers calculated SSIM using luminance alone, the signal rose toward a plateau around the nucleation stage. It effectively indicated that the broad visual structure of the solution had changed, but it became largely insensitive to the later development of the red colour. Full-colour SSIM behaved differently, continuing to evolve during the growth stage. The researchers describe this as two clocks operating within one reaction: a luminance clock that reports the early structural or lightness transition, and a chromatic clock that continues measuring the plasmonic colour evolution.
The comparison across colour spaces also produced a practical result for laboratories that lack specialised imaging equipment. Hue angle was the most discriminating descriptor during red growth, making it particularly effective for following the development of the plasmonic signal. Yet inexpensive broadcast chrominance channels carried nearly equivalent information. That means researchers may not need advanced colour-science workflows to obtain useful kinetic measurements. Standard video data, if recorded consistently, can be transformed into channels that distinguish changes in hue and saturation more effectively than raw brightness alone. The approach could potentially make real-time monitoring more accessible in teaching laboratories, resource-limited research settings and automated chemical workflows.
The team also tested whether the analysis was secretly dependent on choosing a favourable patch of the image. In a 52-configuration sensitivity study, the region of interest was changed in area over a 2,590-fold range, moved across the imaged solution and altered in shape. Despite these substantial changes, the extracted reaction timings remained nearly unchanged. The midpoint of nucleation shifted by only 0.35 percent, while the growth half-time varied by 0.80 percent. According to the researchers, this indicates that the region of interest primarily controls the contrast and strength of the measurement, rather than determining the kinetics themselves. In other words, a smaller or differently shaped sampling area may make the signal noisier or weaker, but it does not fundamentally rewrite the reaction timeline.
The video-derived measurements were also compared with conventional ultraviolet–visible spectroscopy through characterisation of ten nominally identical synthesis batches. UV–visible spectroscopy remains a standard tool for studying nanoparticle formation because it directly records wavelength-dependent absorption and can identify the emergence and position of the surface-plasmon band. The video method does not replace that information in every context: it cannot independently determine particle size, shape distribution or concentration with the same specificity. Its strength is different. It provides an operando, non-contact record of the entire visible evolution without requiring a spectrometer, allowing researchers to connect the familiar visual sequence to quantitative models of reduction, nucleation and autocatalytic growth.
The study’s broader message is that common laboratory observations may contain far more information than scientists traditionally extract from them. A colour change that is usually judged by eye can become a kinetic probe when recorded under controlled conditions and analysed with suitable mathematical and image-processing tools. The authors’ workflow combines colour-space transformations, perceptual descriptors, SSIM analysis and established reaction models to show how chemical information can be recovered from a simple time-lapse recording. By distinguishing early lightness changes from later chromatic growth, the method offers a more complete view of Turkevich synthesis than a single endpoint measurement. The researchers have made the analysed video, processed colour and SSIM data, and analysis code available from the corresponding author on reasonable request, potentially giving other laboratories a foundation for adapting the approach to additional colloidal reactions.
Subject of Research: Quantitative video-based analysis of Turkevich gold-nanoparticle synthesis kinetics
Article Title: Two clocks for one reaction: resolving Turkevich gold-nanoparticle kinetics via operando colour-space and structural-similarity video analysis
Article References: Ichsan, C., Maretha, D.E., Rizkyantoro, B. et al. “Two clocks for one reaction: resolving Turkevich gold-nanoparticle kinetics via operando colour-space and structural-similarity video analysis.” Journal of Nanoparticle Research 28, 229 (2026).
Image Credits: AI Generated
DOI: 10.1007/s11051-026-06744-5
Keywords: Gold nanoparticles; Turkevich synthesis; colour-space analysis; structural similarity index; reaction kinetics; operando optical monitoring; surface plasmon resonance; video analysis
