Colloidal J-Aggregate Flakes

• Colloidal J-aggregate flakes are colloidally dispersed, flake-like assemblies of dye molecules that preserve a sharp J-band at 587 nm and support excitonic behavior.
• Their synthesis via TDBC/PDDA employs controlled electrostatic assembly, yielding stable flakes with tunable morphology and robust solvent performance.
• These flakes enable promising applications in excitonic nanophotonics and energy transport owing to anisotropic exciton dynamics and strong light-matter coupling.

Searching arXiv for papers on colloidal J-aggregate flakes and closely related J-aggregate flake/sheet physics. Colloidal J-aggregate flakes are flake-like colloidal assemblies of molecular J-aggregates that preserve the defining excitonic characteristics of J-aggregation while existing as dispersed, solution-processable mesoscale objects. In the most direct recent realization, they are synthesized from the cyanine dye TDBC and the cationic polyelectrolyte PDDA, yielding plate- or flake-like colloids that remain colloidally stable, retain the sharp J-band at $587\ \text{nm}$, and support room-temperature surface exciton polaritons (Estévez-Varela et al., 18 Jul 2025). More broadly, the term also connects to finite two-dimensional self-assembled J-aggregate sheets, interfacial two-dimensional J-aggregate domains, and other anisotropic excitonic morphologies whose optical and transport behavior is governed by delocalized Frenkel excitons, strong oscillator strength, and morphology-dependent anisotropy (Morales-Curiel et al., 2020).

1. Definition and scope

Colloidal J-aggregate flakes are best defined as colloidally dispersed, flake-like or platelet-like assemblies of dye molecules organized in the J-aggregate state. J-aggregates are supramolecular dye assemblies in which transition dipoles couple coherently, producing a narrow, intense, red-shifted excitonic resonance relative to the monomer (Estévez-Varela et al., 18 Jul 2025). In the TDBC/PDDA colloidal system, the aggregate maintains the classic TDBC J-band at $587\ \text{nm}$, while the monomer band remains at $520\ \text{nm}$ (Estévez-Varela et al., 18 Jul 2025).

The phrase should be distinguished from several related but not identical categories. A finite two-dimensional monolayer cyanine dye aggregate on a brickstone lattice is a sheet-like excitonic object and is the closest morphology to a finite self-assembled two-dimensional flake/domain in theoretical transport work, but it is explicitly an idealized finite 2D aggregate rather than a structurally heterogeneous colloidal dispersion (Morales-Curiel et al., 2020). Langmuir-Blodgett films of thiacyanine dyes contain two-dimensional J-aggregate domains in a supported monolayer, but they are not free colloidal flakes in bulk liquid (Hussain et al., 2011). Silica-coated nanotubular J-aggregates are colloidally dispersed anisotropic excitonic assemblies, yet they are tubular rather than flake-like (Qiao et al., 2015).

The experimentally established colloidal flake platform is based on TDBC and PDDA. These flakes introduce a stable colloidal excitonic building block that retains J-aggregate optical character and exhibits room-temperature surface exciton polariton behavior (Estévez-Varela et al., 18 Jul 2025). A broader interpretation includes any flake-like J-aggregate assembly whose optical response is governed by the same narrow Frenkel-exciton resonance, anisotropic packing, and collective excitonic dielectric function discussed for thin films, sheets, and hybrid excitonic nanostructures (Triolo et al., 2015, Kondorskiy et al., 2021, Lebedev et al., 27 Feb 2025).

2. Materials chemistry and colloidal assembly

The most explicit colloidal-flake synthesis uses TDBC, formally 5,6-dichloro-2-[[5,6-dichloro-1-ethyl-3-(4-sulfobutyl)-benzimidazol-2-ylidene]-propenyl]-1-ethyl-3-(4-sulfobutyl)-benzimidazolium hydroxide sodium salt, together with PDDA, poly(diallyldimethylammonium chloride) (Estévez-Varela et al., 18 Jul 2025). TDBC carries negatively charged sulfonate substituents, while PDDA is cationic. The assembly mechanism is electrostatic and hierarchical: NaCl screens electrostatic repulsion between TDBC molecules and favors J-aggregate formation, after which PDDA associates with the anionic J-aggregates, reverses surface charge, and stabilizes flake-like colloids (Estévez-Varela et al., 18 Jul 2025).

The reported synthesis is chemically specific. An aqueous TDBC solution at concentrations from $0.1$ to $5\ \text{mM}$ is prepared. Then $1\ \text{mL}$ of this TDBC solution is rapidly added, under bath sonication, to $5\ \text{mL}$ of an aqueous solution containing $7\ \mu\text{M}$ PDDA and $0.5\ \text{M}$ NaCl. Sonication is continued for $15\ \text{min}$. After assembly, the product is purified by centrifugation at $587\ \text{nm}$0 for $587\ \text{nm}$1, redispersed in $587\ \text{nm}$2 of $587\ \text{nm}$3 NaCl, washed twice, and finally redispersed in $587\ \text{nm}$4 water (Estévez-Varela et al., 18 Jul 2025). The final PDDA concentration is kept constant at $587\ \text{nm}$5, and the J-aggregate:PDDA molar ratio is tuned from 3 to 150 by varying the TDBC amount (Estévez-Varela et al., 18 Jul 2025).

Colloidal stability is confirmed by zeta potential. Bare TDBC J-aggregates show $587\ \text{nm}$6, whereas after PDDA association the zeta potential becomes strongly positive, around $587\ \text{nm}$7 across all J-aggregate:PDDA ratios studied (Estévez-Varela et al., 18 Jul 2025). Spectroscopically, all colloidal samples in water show the sharp J-band at $587\ \text{nm}$8 and no monomer band at $587\ \text{nm}$9, whereas pristine TDBC in water without PDDA exists as a mixture of monomer and aggregate (Estévez-Varela et al., 18 Jul 2025).

Solvent robustness is a defining chemical feature. In a $520\ \text{nm}$0 ethanol:water mixture, pristine TDBC dissociates completely to monomer with a dominant $520\ \text{nm}$1 band, while the colloidal flakes retain the $520\ \text{nm}$2 J-aggregate signature if the J-aggregate:PDDA ratio exceeds 15 (Estévez-Varela et al., 18 Jul 2025). At ratios below 3, both monomer and aggregate bands are present, indicating insufficient stabilization, and for ratios above 30 the flakes remain stable even in pure ethanol or methanol (Estévez-Varela et al., 18 Jul 2025).

A related but distinct stabilization strategy is inorganic encapsulation. In nanotubular C8S3 J-aggregates, a mild APTES/TEOS sol-gel process produces a continuous but porous amorphous silica nanoshell with helically wound ribbon-like topography while preserving the J-aggregate optical signatures (Qiao et al., 2015). This suggests a transferable route for improving photostability, mechanical robustness, and processability of fragile anisotropic J-aggregate colloids, although that work does not address flakes directly (Qiao et al., 2015).

3. Morphology, dimensions, and structural interpretation

The TDBC/PDDA product is a polydisperse population of flake-like colloids. TEM shows predominantly plate- or flake-like nanostructures with well-defined edges, and their lateral size ranges from about $520\ \text{nm}$3 to several micrometers for J-aggregate:PDDA ratios from 15 to 150 (Estévez-Varela et al., 18 Jul 2025). At ratio 3, flakes are scarce and less-defined morphologies coexist, consistent with incomplete spectroscopic stabilization (Estévez-Varela et al., 18 Jul 2025). AFM confirms the flake topology and gives thicknesses from approximately $520\ \text{nm}$4 to $520\ \text{nm}$5, with representative flakes around $520\ \text{nm}$6, $520\ \text{nm}$7, and $520\ \text{nm}$8 thick (Estévez-Varela et al., 18 Jul 2025).

For single-particle optical analysis, flakes are classified as small $520\ \text{nm}$9, medium $0.1$0, and large $0.1$1 (Estévez-Varela et al., 18 Jul 2025). Larger lateral dimensions generally correlate with greater thickness (Estévez-Varela et al., 18 Jul 2025). The internal molecular order is not crystallographically resolved: no electron diffraction, XRD, or related direct structural evidence is provided, so the flakes are not established as crystalline, layered-crystalline, or amorphous/polycrystalline (Estévez-Varela et al., 18 Jul 2025). What is directly supported is a flattened, anisotropic flake morphology with sufficient optical homogeneity to exhibit a bulk-like excitonic dielectric response (Estévez-Varela et al., 18 Jul 2025).

This uncertainty in microscopic packing is important. The optical data show preserved J-aggregate behavior, morphology-dependent scattering, and a room-temperature Reststrahlen-like response, but the exact molecular packing model inside the flakes is not resolved (Estévez-Varela et al., 18 Jul 2025). A plausible implication is that colloidal flake function can be established experimentally before a full structural solution is available, provided the excitonic signatures remain strong and reproducible.

The flake concept also has antecedents in theory and interfacial assemblies. A modeled 2D monolayer cyanine dye J-aggregate of 500 molecules arranged on a brickstone lattice is a finite 2D sheet-like molecular aggregate and the closest morphology in that study to a finite self-assembled 2D flake/domain (Morales-Curiel et al., 2020). Likewise, NK Langmuir-Blodgett films are described as containing domains where dye molecules are organized into a two-dimensional brick-like array, but microscopy does not resolve discrete flakes and the system remains a supported monolayer rather than a colloidal dispersion (Hussain et al., 2011).

4. Excitonic optical response

The defining optical hallmark of colloidal J-aggregate flakes is retention of the J-band. In the TDBC/PDDA colloids, the main extinction band remains centered at $0.1$2, strongly red-shifted relative to monomeric TDBC at $0.1$3 (Estévez-Varela et al., 18 Jul 2025). The absorption spectrum remains symmetric and centered at $0.1$4 independent of molar ratio, implying that the intrinsic excitonic transition energy of the J-aggregate subunits is unchanged by flake formation (Estévez-Varela et al., 18 Jul 2025).

At the same time, the colloidal architecture modifies the total optical response. The extinction band is broader than that of pristine J-aggregates in solution, and the full width at half maximum increases with J-aggregate:PDDA ratio (Estévez-Varela et al., 18 Jul 2025). Integrating-sphere measurements show that the broadening originates from a morphology-dependent scattering component. The scattering spectrum peaks at longer wavelengths than the absorption and redshifts as the ratio increases, so the total extinction broadens toward the red as larger flakes become more abundant (Estévez-Varela et al., 18 Jul 2025).

Photoluminescence retains narrow-emission J-aggregate character with a small Stokes shift of about $0.1$5–$0.1$6, depending on the J-aggregate:PDDA ratio, but the PL maximum progressively redshifts as the molar ratio increases (Estévez-Varela et al., 18 Jul 2025). Because the absorption peak itself does not move, the shift is not attributed to intrinsic changes in exciton energy. Instead, the proposed explanation is that the stronger and redder scattering contribution of larger flakes modifies the local density of optical states and biases radiative recombination toward longer wavelengths (Estévez-Varela et al., 18 Jul 2025). Time-resolved PL would be needed to test that interpretation explicitly (Estévez-Varela et al., 18 Jul 2025).

Several broader excitonic principles from related J-aggregate work help interpret flake optics. First-principles simulations of a push-pull molecular J-aggregate show that the J-band is a collective, supramolecular excitation that is longitudinal, delocalized along parallel chains, and only weakly perturbed when going from bulk to a monolayer sheet (Guerrini et al., 2018). This suggests that a thin flake can preserve the essential J-band provided the dominant in-plane chain-like packing motif is maintained. The same study also finds that the overall red shift results from competing coupling mechanisms, some giving red shift and others giving blue shift, deriving from both coupling between transition densities and renormalization of single-particle energy levels (Guerrini et al., 2018).

Other first-principles work further refines the microscopic picture. In a push-pull molecular J-aggregate, the bright excitation at the J-band onset is a mixed state combining intramolecular charge transfer and intermolecular charge transfer, while a lower-energy, very weak pure intermolecular charge-transfer excitation exists inside the J-band manifold (Guerrini et al., 2019). This suggests that in some flake-forming molecular systems, J-band spectroscopy alone may not distinguish a purely Frenkel picture from a mixed intra/intermolecular CT exciton.

A common misconception is that any strong, collective J-band should be described as plasmonic. A many-body perturbation theory analysis of a molecular J-aggregate found that the effect of aggregation is to lower the generalized plasmonicity index associated with the J-band relative to the isolated dye, indicating nonplasmonic character of the excitation (Guerrini et al., 2019). For colloidal J-aggregate flakes, the default interpretation of the optical resonance is therefore excitonic, not microscopic plasmonic, unless a separate induced-potential analysis supports a plasmonic classification.

5. Surface exciton polaritons and excitonic nanophotonics

The most distinctive photonic property established for colloidal J-aggregate flakes is support of room-temperature surface exciton polaritons. In single-flake angle-resolved reflectance measurements using visible Fourier imaging spectroscopy, a representative medium-sized flake shows a broad, high-intensity reflectance band that is invariant with angle of incidence and lies at shorter wavelength than the $0.1$7 excitonic absorption (Estévez-Varela et al., 18 Jul 2025). This angular independence indicates that the feature is not a geometric cavity resonance or diffraction artifact but reflects the intrinsic dielectric function of the material (Estévez-Varela et al., 18 Jul 2025).

The feature is interpreted as an excitonic Reststrahlen band. The underlying dielectric description is a Lorentz oscillator,

$0.1$8

or equivalently

$0.1$9

with $5\ \text{mM}$0 or $5\ \text{mM}$1 the background permittivity, $5\ \text{mM}$2 the oscillator strength, $5\ \text{mM}$3 the excitonic resonance, and $5\ \text{mM}$4 the damping rate (Estévez-Varela et al., 18 Jul 2025). Near resonance, $5\ \text{mM}$5 peaks because of absorption while $5\ \text{mM}$6 disperses rapidly and can become negative on one side of resonance, producing the excitonic analogue of a Reststrahlen band (Estévez-Varela et al., 18 Jul 2025).

Direct SEP evidence comes from Kretschmann attenuated total reflection measurements. For $5\ \text{mM}$7-polarized light, a distinct angle-dependent reflectance dip appears at wavelengths shorter than the excitonic absorption and only for incident angles greater than the cover-glass critical angle, reported as $5\ \text{mM}$8 (Estévez-Varela et al., 18 Jul 2025). At $5\ \text{mM}$9, $1\ \text{mL}$0 polarization exhibits a pronounced minimum beyond the critical angle, whereas $1\ \text{mL}$1 polarization does not show the same coupling signature (Estévez-Varela et al., 18 Jul 2025). This polarization selectivity is consistent with TM-like surface polaritons.

The p-polarized response near the critical angle also shows a sharp reflectance increase described as a Fano-type resonance, arising from anti-crossing between the SEP branch and the light line (Estévez-Varela et al., 18 Jul 2025). The novelty is not merely that TDBC can support excitonic negative-permittivity behavior, but that this is observed at the single-particle level in non-continuous colloidal flakes (Estévez-Varela et al., 18 Jul 2025).

Related theory and nanostructure studies place this result in a wider framework. Nanodisks carved from ultrathin J-aggregate thin films can support localized surface resonances because the real part of the dielectric function becomes sufficiently negative on the blue side of the J-band (Triolo et al., 2015). In a dielectric model,

$1\ \text{mL}$2

negative real permittivity becomes possible if

$1\ \text{mL}$3

with $1\ \text{mL}$4 (Triolo et al., 2015). This establishes a design rule for excitonic flakes: sufficiently strong oscillator strength and sufficiently narrow linewidth are jointly required to access surface-mode or localized-resonance regimes.

For thin platelet-like geometries, the same work shows that a $1\ \text{mL}$5-radius, $1\ \text{mL}$6-thick nanodisk can generate $1\ \text{mL}$7, while a 4 nm-gap disk dimer can exceed $1\ \text{mL}$8 (Triolo et al., 2015). These are not colloidal flake experiments, but they are the closest electromagnetic proxy to excitonic platelet behavior. A plausible implication is that colloidal J-aggregate flakes with comparable dielectric response should support edge-localized near fields and strong gap hot spots in coupled-flake configurations.

6. Exciton transport, anisotropy, and realistic illumination

Colloidal J-aggregate flakes are also relevant as exciton-transport media. A theoretical model of exciton dynamics in thin-film cyanine dye J-aggregates treats TC, TDBC, and U3 monolayers on a two-dimensional brickstone lattice and finds that transport is anisotropic and dependent on the initial exciton energy (Valleau et al., 2012). For TDBC with $1\ \text{mL}$9 static disorder and $5\ \text{mL}$0 dynamic disorder, the diffusion lengths estimated using an experimental lifetime of $5\ \text{mL}$1 are

$5\ \text{mL}$2

presented as upper-bound estimates (Valleau et al., 2012). The same work finds a ballistic regime for about $5\ \text{mL}$3–$5\ \text{mL}$4, followed by diffusion, and reports that transport is about 2–3 times faster along one in-plane axis than along the other (Valleau et al., 2012).

These results are not direct predictions for colloidal flakes, because the model assumes an idealized monolayer without explicit edge chemistry, domain mosaicity, or solvent heterogeneity (Valleau et al., 2012). Even so, they support a robust qualitative picture: if a flake preserves ordered in-plane J-aggregate packing, exciton motion should be both long-ranged and anisotropic.

The role of realistic illumination has been analyzed in a finite 2D self-assembled cyanine J-aggregate of 500 molecules arranged in a brickstone lattice (Morales-Curiel et al., 2020). The single-exciton Hamiltonian is

$5\ \text{mL}$5

with $5\ \text{mL}$6, $5\ \text{mL}$7, and $5\ \text{mL}$8 for the TC monolayer (Morales-Curiel et al., 2020). Open-system dynamics are described with a Lindblad master equation including dephasing, dissipation, and sink trapping (Morales-Curiel et al., 2020). The transport efficiency is defined as

$5\ \text{mL}$9

The key photophysical result is that at long distances incoherently delocalized excitation gives the highest transport efficiency, whereas at short distances localized excitation performs best (Morales-Curiel et al., 2020). At a sink distance of $7\ \mu\text{M}$0, incoherent excitation gives $7\ \mu\text{M}$1 higher efficiency than coherent delocalized excitation and $7\ \mu\text{M}$2 higher efficiency than localized excitation (Morales-Curiel et al., 2020). The mechanism is that incoherent light prepares a mixture of delocalized exciton eigenstates,

$7\ \mu\text{M}$3

which corresponds in the site basis to a strongly spatially extended excitation (Morales-Curiel et al., 2020).

For colloidal J-aggregate flakes, this is a significant correction to the common laser-based intuition. Under sunlight-like conditions, incoherent illumination does not merely wash out useful transport behavior; it can enhance long-range transport because the aggregate’s intrinsic exciton eigenstates are already delocalized (Morales-Curiel et al., 2020). The result is model-based and assumes perfect 2D order, no static energetic disorder, and no recombination losses during the transport window (Morales-Curiel et al., 2020), but it supports the use of flake-like J-aggregate assemblies as light-harvesting antennas and exciton-delivery media.

7. Anisotropy, hybrid nanostructures, and open questions

A central issue for colloidal J-aggregate flakes is anisotropy. In hybrid plasmon–exciton nanoparticles with molecular J-aggregate shells, treating the shell as an isotropic scalar dielectric can be qualitatively wrong (Kondorskiy et al., 2021). The relevant dielectric model is tensorial,

$7\ \mu\text{M}$4

with separate longitudinal and transverse oscillator sets (Kondorskiy et al., 2021). This is especially important for flat or layered aggregate morphologies, where the dominant dipole may lie parallel to the aggregate plane rather than normal to it (Kondorskiy et al., 2021).

For colloidal flakes, the likely consequence is that in-plane and out-of-plane optical responses should not be assumed equal. Spectral position, line shape, and the visibility of hybrid modes can all depend on whether the dominant excitonic dipoles are parallel or perpendicular to the flake plane (Kondorskiy et al., 2021). In practice, this means that polarization-resolved spectroscopy and orientation control should be central to future flake studies.

Hybridization with other excitonic media is another emerging direction. In a layered Ag/Al$7\ \mu\text{M}$5O$7\ \mu\text{M}$6/1L-WS$7\ \mu\text{M}$7/TDBC:PVA platform, TDBC J-aggregates with a Frenkel exciton resonance at $7\ \mu\text{M}$8 and linewidth $7\ \mu\text{M}$9 strongly couple to a surface plasmon polariton and, through that shared mode, hybridize with the WS$0.5\ \text{M}$0 Wannier-Mott exciton at $0.5\ \text{M}$1 (Morales et al., 13 Mar 2025). The coupling strength of the J-aggregate layer scales as $0.5\ \text{M}$2, where $0.5\ \text{M}$3 is the number of molecules participating in the mode volume (Morales et al., 13 Mar 2025). This suggests that flake area, thickness, and areal coverage should be decisive parameters in future flake-based polaritonic devices.

The main unresolved questions remain structural and mesoscopic. For the TDBC/PDDA colloidal flakes, the internal packing is incompletely resolved, size polydispersity is substantial, and scattering, reflectance strength, and photoluminescence all depend strongly on morphology (Estévez-Varela et al., 18 Jul 2025). Real colloidal flakes are also expected to differ from idealized monolayers because of edge disorder, finite-temperature structural motion, nonuniform dielectric environment, trap states, multiple domains, and interflake electronic decoupling (Morales-Curiel et al., 2020, Valleau et al., 2012). The present evidence therefore supports a strong qualitative principle: flake-like J-aggregate colloids can act as stable excitonic mesoscale objects with preserved J-band absorption, morphology-dependent scattering, room-temperature SEP behavior, and potentially efficient long-range exciton transport. Exact quantitative performance remains morphology- and environment-specific.

A concise way to state the current position is that colloidal J-aggregate flakes are now established as a real materials class in the form of TDBC/PDDA colloids (Estévez-Varela et al., 18 Jul 2025), while theory and related experiments show that their broader significance lies in combining finite two-dimensional excitonic geometry, anisotropic dielectric response, realistic-light exciton transport, and room-temperature excitonic nanophotonics (Morales-Curiel et al., 2020, Guerrini et al., 2018, Kondorskiy et al., 2021).