Abstract

   Efficient photocatalytic tumor therapy relies on deep tissue
   penetration via near-infrared (NIR) light absorption and potent
   oxidative stress induced by efficient photogenerated hole (h^+)
   migration. Herein, we report a supramolecular donor–acceptor (D–A)
   structure fabricated by integrating self-assembled
   tetra(4-carboxyphenyl) porphyrin (SA-TCPP) with amino-functionalized
   graphene quantum dots (GQDs-NH[2]). The D–A configuration significantly
   narrowed the band gap for extended NIR absorption and amplified the
   interfacial electric field 6.2 times. This facilitated efficient h^+
   migration with a 5.3-fold enhancement, generating highly oxidation
   holes. Under 700 nm irradiation, this system induced potent oxidative
   cell apoptosis, leading to 100% tumor elimination in vivo. This study
   presents a dual-promotion strategy involving NIR responsiveness and
   charge migration, advancing the efficiency of photocatalytic cancer
   therapy.

   Keywords: photocatalytic therapy, near-infrared response, charge
   migration, interfacial electric field, D−A structure
     __________________________________________________________________

   graphic file with name au5c00535_0010.jpg
     __________________________________________________________________

   graphic file with name au5c00535_0009.jpg

Introduction

   Despite the emergence of various clinical therapies, cancer remains a
   formidable threat to human health, often complicated by challenges such
   as drug resistance, low treatment response rate, and significant toxic
   side effects. , Photocatalytic therapy (PCT) has recently emerged as a
   promising strategy, offering the potential for efficient, safe, and
   precise treatment. PCT leverages photocatalysts that absorb light
   energy, triggering reactions that disrupt the local the
   oxidation–reduction (REDOX) balance within the tumor microenvironment
   and inducing irreversible cancer cell apoptosis. Specifically, upon
   light excitation, photogenerated holes (h^+) in the valence band and
   the electrons (e^–) in the conduction band migrate to the catalyst
   surface, generating reactive oxygen species (ROS) or directly oxidizing
   biomolecules to kill cancer cells. However, the efficacy of PCT is
   often hampered because most traditional photocatalysts are activated by
   ultraviolet–visible (UV–vis) light, which has limited tissue
   penetration depth (only a few millimeters). Near-infrared (NIR) light
   (700–1100 nm), conversely, penetrates biological tissues more
   effectively (up to ca. 1 cm) through a light transparent window, making
   it ideal for treating deeper tumors. Consequently, there is an urgent
   need for advanced photocatalysts that exhibit strong NIR absorption and
   facilitate efficient charge migration under NIR excitation. ,− While
   some NIR-responsive materials exist, they often rely on strategies like
   incorporating potentially toxic heavy metals or employing upconversion
   techniques, the latter of which can suffer from low overall light
   utilization efficiency. , Therefore, developing biocompatible
   photocatalysts directly excitable by NIR light is crucial for advancing
   safe and effective deep-tissue PCT.

   Organic semiconductor photocatalysts, including porphyrin,
   phthalocyanine, and indocyanine green, have garnered significant
   attention for biomedical photocatalysis due to their tunable spectral
   responses, good biocompatibility, − and facile structure modification.
   , Porphyrins, a class of chromophores with π-conjugated electrons, are
   particularly noteworthy for their broad light absorption, excellent
   photoelectric conversion capabilities, and high chemical stability.
   Among them, tetra(4-carboxyphenyl) porphyrin (TCPP) is an attractive
   precursor owing to its strong light-harvesting ability and good
   hydrophilicity. , Our previous work demonstrated that self-assembled
   TCPP (SA-TCPP), formed through hydrogen bonding, π–π stacking, and
   hydrophobic interaction, adopts a unique supramolecular structure that
   can selectively accumulate in tumor cells through size regulation.
   Under 600 nm irradiation, SA-TCPP can generate strong oxidative h^+ for
   rapid elimination of solid tumors. However, it exhibits poor
   photocatalytic activity under NIR irradiation due to its limited
   intrinsic NIR absorption and substantial charge carrier recombination.
   Thus, designing SA-TCPP derivatives with enhanced NIR-responsiveness
   and improved charge separation and migration efficiency is essential
   for effective tumor in-depth treatment.

   Several strategies have been developed to broaden the light absorption
   and enhance charge transfer in organic semiconductors, including
   constructing donor–acceptor (D–A) structure, introducing auxochromic
   groups, extending π-conjugate systems, and leveraging plasmon
   resonance. Among these, constructing D–A heterostructures is
   particularly effective, as it can establish a strong interfacial
   electric field (IEF), providing an intrinsic driving force for the
   separation and migration of photogenerated carriers. ,
   Amino-functionalized graphene quantum dots (GQDs-NH[2]), as auxochrome
   groups containing lone π-electrons, can effectively enhance light
   absorption through quantum confinement effects and π-plasmon resonance.
   , Inspired by the excellent electron transfer properties, which are
   known to facilitate charge separation, we hypothesized that modifying
   SA-TCPP with GQDs-NH[2] to construct a D–A heterostructure could not
   only expand light to the NIR region but also improve charge separation
   efficiency, thereby addressing current challenges in photocatalytic
   tumor therapy.

   Herein, we report the successful fabrication of a D–A supramolecular
   structure, designated SA-TCPP@GQDs, via electrostatic assembly of
   SA-TCPP and GQDs-NH[2]. This integration significantly enhanced
   intermolecular π–π conjugation and extended the intrinsic absorption
   band edge to the NIR region (up to 833 nm). Crucially, the D–A
   interaction established a strong interfacial electric field (IEF)
   within SA-TCPP@GQDs, markedly promoting the separation and migration
   efficiency of photogenerated carriers (a 5.3-fold increase for h^+
   migration). Once internalized by cancer cells and irradiated with
   near-infrared light (700 nm), SA-TCPP@GQDs generated a substantially
   enhanced flux of highly oxidizing holes. These holes directly induced
   potent oxidative stress, damaging vital intracellular components and
   triggering irreversible tumor cell apoptosis ([43]Scheme ). Notably,
   the efficient generation of oxidizing h^+ under NIR light offers
   potential advantages for treating hypoxic tumors, which are often
   resistant to therapies relying solely on oxygen-dependent ROS.
   Furthermore, SA-TCPP@GQDs demonstrated favorable tumor targeting
   capabilities, blood circulation stability, and biological metabolism in
   vivo, making it highly promising for safe and efficient PCT treatment.
   This study presents a novel approach for NIR-responsive cancer therapy
   by constructing the D–A system that enhanced light absorption and
   promoted charge separation simultaneously.

1. Schematic Illustration of Supramolecular SA-TCPP@GQDs Nanoparticles
Increased Photogenerated Holes under Near-Infrared (NIR) Light for Efficient
Photocatalytic Cancer Therapy.

   1
   [44]Open in a new tab

Results and Discussion

Fabrication and Characterization of Supramolecular SA-TCPP@GQDs with D–A
Structure

   Self-assembled tetra(4-carboxyphenyl) porphyrin (SA-TCPP) was prepared
   following the previously reported method. Subsequently,
   amino-functionalized graphene quantum dots (GQDs-NH[2]) were integrated
   into SA-TCPP through coassembly to yield the SA-TCPP@GQDs. Transmission
   electron microscopy (TEM) images showed that SA-TCPP@GQDs formed
   nanoparticles with an average size of 34.7 nm and a thickness of about
   50 nm ([45]Figure a, ). GQDs-NH[2] was uniformly distributed on the
   surface of SA-TCPP (). High-resolution transmission electron microscopy
   (HRTEM) images displayed distinct closely adjacent lattice fringes of
   SA-TCPP and GQDs-NH[2], corresponding to a d-spacing of 0.35 and 0.20
   nm, respectively ([46]Figure b). The former (0.35 nm) confirms π–π
   stacking in the porphyrin supramolecular structure, while the latter
   demonstrates the successful integration of GQDs-NH[2]. This close
   interfacial contact between the two components confirms the formation
   of an intimate nanoheterojunction.

1.

   [47]1
   [48]Open in a new tab

   Characterization of D–A supramolecular SA-TCPP@GQDs. (a) TEM images and
   (b) HRTEM images of SA-TCPP@GQDs (the mass ratio of GQDs-NH[2] to
   SA-TCPP is 5%). (c) XRD pattern of TCPP, SA-TCPP, GQDs-NH[2], and
   SA-TCPP@GQDs. (d) FT-IR spectra of SA-TCPP and SA-TCPP@GQDs. (e) Zeta
   potential measurements of SA-TCPP, GQDs-NH[2], and SA-TCPP@GQDs. (f)
   Frontier molecular orbital distribution of SA-TCPP@GQDs.

   The presence of high crystallinity and a well-defined interface within
   SA-TCPP@GQDs is expected to be conducive to establishing a strong
   interfacial electric field. In the X-ray diffraction (XRD) pattern
   ([49]Figure c), SA-TCPP shows a diffraction peak of about 21° due to
   π–π stacking of supramolecular structure, while the diffraction peak of
   SA-TCPP@GQDs shifts to higher angles, indicating that the introduction
   of GQDs-NH[2] enhances the conjugation interaction of the porphyrin
   assembly, thereby shortening the intermolecular π–π stacking distance.
   Raman spectra further reveal that the ratio of the two characteristic
   vibration peaks (I [1/2]) of SA-TCPP@GQDs is 0.87, which is slightly
   greater than that of SA-TCPP (0.83), suggesting that the π–π stacking
   degree of SA-TCPP@GQDs is higher than that of SA-TCPP ().

   A Fourier-transformed infrared (FT-IR) spectrum ([50]Figure d) shows
   that the vibration frequency of C–O in SA-TCPP@GQDs is lower than that
   of SA-TCPP, attributed to the electrostatic interaction between −COOH
   of SA-TCPP and −NH[2] of GQDs-NH[2]. Furthermore, SA-TCPP@GQDs have a
   zeta potential of +13.6 mV, in contrast to −68.3 mV for SA-TCPP,
   following electrostatic adsorption by amine-rich GQDs-NH[2] ([51]Figure
   e). This positive zeta potential is beneficial for cell membrane
   combination, thereby improving intracellular uptake efficiency.
   Besides, SA-TCPP exhibits good hydrophilicity due to the four carboxyl
   groups at the end of TCPP, and the hydrophilicity is further enhanced
   after modification by GQDs-NH[2] ().

   Moreover, the theoretical calculations of the frontier molecular
   orbital distribution showed that the highest occupied molecular orbital
   (HOMO) and lowest unoccupied molecular orbital (LUMO) of SA-TCPP@GQDs
   are located on the GQDs-NH[2] moiety and SA-TCPP moiety, respectively
   ([52]Figure f). This distribution directly confirms the characteristics
   of the donor–acceptor (D–A) structure. Collectively, the above results
   show that supramolecular SA-TCPP@GQDs have a D–A structure and enhanced
   conjugation effect, which is expected to expand the range of light
   absorption and promote charge transfer transition.

D–A Structure Broadens NIR Absorption and Promotes Charge Migration

   The strong electrostatic interaction between the carboxyl and amino
   groups promotes the π–π conjugation and decreases the π–π* transition
   energy, thus contributing to the redshift of SA-TCPP@GQDs. As shown in
   [53]Figure a, the typical Q-bands (500–700 nm) in SA-TCPP and
   SA-TCPP@GQDs show a redshift compared with the porphyrin monomer
   (TCPP), indicating the formation of J-type aggregates in the
   supramolecular systems. Notably, the absorption redshift of
   SA-TCPP@GQDs is even larger, and the absorption peak in Q-bands
   decreases, proving that the electrostatic interaction further promoted
   the intermolecular conjugated interaction. As a result, the intrinsic
   absorption band edge of SA-TCPP@GQDs reaches the NIR region of 833 nm
   ([54]Figure b).

2.

   [55]2
   [56]Open in a new tab

   D–A charge transfer transition enhances the NIR absorption. (a) UV–vis
   absorption spectra of TCPP, SA-TCPP, SA-TCPP@GQDs, and
   SA-TCPP@GQDs-NH[2]. (b) Diffuse reflection spectroscopy of samples and
   solar energy utilization. (c) Calculated electronic density of states
   (DOS). (d) Energy band structure diagrams for GQDs-NH[2] and SA-TCPP.

   UV–vis diffuse reflection spectroscopy (DRS) reveals a new absorption
   peak at ∼700 nm, indicating a charge transfer (CT) transition between
   SA-TCPP and GQDs-NH[2] (). The theoretical model for the calculation of
   the electronic density of states (DOS) () was further carried out. As
   shown in [57]Figure c, SA-TCPP has semiconductor properties where the
   Fermi level is positioned in the middle of the gap without any electron
   distribution around the Fermi level, and the incorporation of
   GQDs-NH[2]introduces positively correlated DOS near the Fermi level.
   This suggests the formation of transition energy states during the CT
   transitions in SA-TCPP@GQDs, which are composed of localized or
   isolated occupied states. Such states not only narrow the band gap but
   also facilitate electron diffusion via a polaron-hopping mechanism. ,

   Based on the electrochemical Mott–Schottky curve and band gap analysis
   (), the band structure diagram is obtained and shown in [58]Figure d.
   The valence band position of GQDs-NH[2] is higher than that of SA-TCPP,
   while the conduction band position is lower than that of SA-TCPP. Their
   staggered energy level structure would greatly facilitate
   photogenerated carrier migration between materials. These findings
   underscore the impact of the D–A structure and electrostatic
   interactions on enhancing light absorption and promoting charge
   separation and migration, positioning SA-TCPP@GQDs as a promising
   candidate for NIR-responsive photocatalysts.

   The charge transfer mechanism between GQDs-NH[2] and SA-TCPP was
   further investigated using Kelvin probe force microscopy (KPFM) in a
   N[2] atmosphere to measure the contact potential difference (CPD). The
   work functions (WFs) of SA-TCPP and GQDs-NH[2] are calculated as 4.53
   and 4.59 eV, and the relative Fermi levels (FLs) of the two are
   determined and illustrated in . The FL of SA-TCPP was significantly
   higher than that of GQDs-NH[2] before contact. When the two are in
   close contact due to electrostatic interaction, as shown in [59]Figure
   a, the intrinsic electrons spontaneously transfer from the GQDs-NH[2]
   to SA-TCPP until the Fermi levels of the two are in equilibrium and the
   energy bands upward and downward, respectively. As a result, the charge
   transfer pathway at the interface was successfully established. In this
   case, an interfacial electric field (IEF) from the SA-TCPP toward the
   GQDs-NH[2] was generated, which could be a powerful driving force to
   promote the separation and migration of photogenerated carriers. On
   that account, the IEF intensity was quantitatively determined by the
   model developed by Kanata et al. (), and the IEF of SA-TCPP@GQDs is
   about 6.2 times higher than that of SA-TCPP ([60]Figure b). The
   internal CT transition behavior of SA-TCPP@GQDs not only increases the
   NIR absorption but also induces the generation of IEF, which is
   expected to promote charge separation and migration.

3.

   [61]3
   [62]Open in a new tab

   Efficient charge transfer facilitated by the interfacial electric field
   (IEF) in D–A structure. (a) Schematic of charge transfer route and IEF
   direction in SA-TCPP@GQDs. (b) IEF intensity for SA-TCPP and
   SA-TCPP@GQDs. (c) Transient photovoltage (TPV) spectra (excitation
   pulse at 355 nm and 50 mJ). (d) Electrochemical impedance spectroscopy
   (EIS) Nyquist plots. (e) Transient absorption spectra of SA-TCPP@GQDs
   at different time delays and SA-TCPP (femtosecond laser excitation at
   370 nm). (f) Time profiles of normalized fitted decay kinetics of
   transient absorption at 465 nm. (g) Time-resolved fluorescence decay
   spectra (ex@430 nm). (h) Surface photovoltage spectra (SPV). (i)
   Photogenerated holes of SA-TCPP and SA-TCPP@GQDs detected with in situ
   CLSM.

   Furthermore, the separation and migration of photogenerated carriers
   facilitated by the IEF are systematically investigated. First, KPFM was
   used to determine the surface potential. The surface potential of
   SA-TCPP@GQDs (ΔE = 48 mV) is about 1.9 times that of SA-TCPP (ΔE = 25
   mV) (), which proves the photogenerated carrier separation efficiency
   of SA-TCPP@GQDs driven by IEF. Photoelectron chemical measurement was
   used to assess the charge behavior. The photocurrent of SA-TCPP@GQDs
   was much stronger than that of SA-TCPP under light (), indicating that
   the IEF can promote carrier separation effectively. Transient
   photovoltage (TPV) spectroscopy showed that SA-TCPP@GQDs exhibit high
   first peak photovoltage and long second peak photovoltage ([63]Figure
   c), demonstrating that IEF can significantly improve the charge
   separation efficiency.

   Electrochemical impedance spectroscopy (EIS) was measured to evaluate
   the resistance of charge transfer ([64]Figure d). Obviously,
   SA-TCPP@GQDs has a smaller arc diameter and a lower fitted charge
   transfer resistance (R [2]) than SA-TCPP (), indicating that the IEF is
   conducive to decreasing the charge transfer resistance and thus
   promoting the charge transfer kinetics. The femtosecond transient
   absorption spectra (TAS) were measured at 370 nm excitation. As shown
   in [65]Figure e, the positive signal of SA-TCPP@GQDs at 430–720 nm is
   wider and stronger than that of SA-TCPP, which comes from the excited
   state absorption (ESA) of the singlet excited state. Furthermore, the
   decay kinetics of photogenerated carriers were revealed by fitting the
   time curve of the transient absorption detected at 465 nm ([66]Figure
   f). Compared with SA-TCPP (τ[1] = 135.73 ps), the captured electron
   lifetime (τ[1]) in SA-TCPP@GQDs decreases significantly to 12.96 ps,
   which indicates that the interfacial electron transfer is faster in D–A
   structure. At the same time, due to the longer exciton relaxation
   process, the decay lifetime of SA-TCPP@GQDs τ[2] (330.90 ps) is longer
   than that of SA-TCPP (68.42 ps), which means that photogenerated
   carriers have more opportunities to participate in the photocatalytic
   reaction.

   The photoluminescence (PL) spectrum showed that the fluorescence
   emission peak of SA-TCPP@GQDs at 625–750 nm was significantly reduced
   compared to that of SA-TCPP (), indicating that IEF effectively
   inhibits the recombination of h^+ and e^–, thus improving the
   utilization rate of photogenerated carriers. Moreover, the
   time-resolved fluorescence decay (TR-PL) spectra measured at the
   steady-state emission peak show a longer lifetime (τ) of photogenerated
   carriers in SA-TCPP@GQDs ([67]Figure g), further implying that the
   long-range transport of photogenerated carriers is significantly
   enhanced, driven by IEF. Surface photovoltage (SPV) was used to further
   evaluate the charge transfer behavior. As shown in [68]Figure h,
   positive SPV signals of SA-TCPP@GQDs and SA-TCPP are displayed at
   500–800 nm, indicating that the photogenerated holes accumulate on the
   catalyst surface as dominant carriers. It is worth noting that the SPV
   response of SA-TCPP@GQDs is significantly enhanced, indicating a higher
   carrier migration efficiency.

   To map the surface reaction of photogenerated h^+ and e^–, the
   previously reported fluorescence detection method of resorufin was
   used, , and the h^+ formation was observed and quantified by confocal
   laser fluorescence microscopy (CLSM). As illustrated in [69]Figure i, a
   weak fluorescence signal of h^+ exists in SA-TCPP (inset, left), while
   a stronger signal appears in SA-TCPP@GQDs (inset, right). To compare
   the concentration of holes, fluorescence quantity and fluorescence
   intensity are used to define the number and strength of fluorescence
   signals per unit area. Specifically, the relative fluorescence quantity
   and intensity of SA-TCPP@GQDs are 3.1 and 5.3 times stronger than those
   of SA-TCPP, which proves that the giant IEF greatly promotes effective
   migration of photogenerated h^+.

Enhanced Photogenerated Holes for Efficient Photocatalytic Tumor Therapy in
Vitro

   Inspired by the above properties of NIR response and efficient charge
   migration, we investigated the cellular uptake behavior of SA-TCPP@GQDs
   and its therapeutic potential for PCT. Based on the unique
   multiwavelength excited self-fluorescence properties of SA-TCPP@GQDs
   (), the cellular uptake of nanoparticles can be detected and observed
   without additional fluorescence modifications. Flow cytometry analysis
   revealed concentration-dependent uptake in HeLa cells with an
   endocytosis efficiency of 97.6% at 12.5 μg/mL ([70]Figure a). Notably,
   SA-TCPP@GQDs exhibited higher cellular uptake than SA-TCPP under
   identical conditions (), attributed to their positively charged surface
   ([71]Figure e), which enhances electrostatic interaction with the
   negatively charged cell membrane. In contrast, uptake in normal cells
   (293T, MCF-10A, and L02) remained below 20% for both SA-TCPP and
   SA-TCPP@GQDs, as further confirmed by confocal microscopy ().
   Strikingly, >95% uptake efficiency was observed in various cancer cells
   (HeLa, MCF-7, and HepG2, [72]Figure b), highlighting their exceptional
   selectivity for tumor cells and promising potential for targeted
   therapy.

4.

   [73]4
   [74]Open in a new tab

   Tumor cell endocytosis of SA-TCPP and SA-TAPP@GQDs nanoparticles. (a)
   Cellular uptake rate of HeLa cells treated with different
   concentrations of SA-TCPP and SA-TCPP@GQDs for 12 h. (b) Comparison of
   cellular uptake rate of SA-TCPP and SA-TCPP@GQDs (12.5 μg/mL) in
   cancerous or normal cells. (c) Cellular uptake pathway study of
   SA-TCPP@GQDs (12.5 μg/mL). (d) CLSM images of HeLa cells incubated with
   SA-TCPP or SA-TCPP@GQDs. The cell nucleus and lysosome were stained by
   Hoechst 33258 (10 μg/mL) and 100 nM LysoTracker@Red before CSLM
   imaging, respectively. Scale bar: 20 μm.

   Next, the cellular uptake pathway of the SA-TCPP@GQDs was studied.
   Prior to incubation with SA-TCPP@GQDs, HeLa cells were pretreated with
   different endocytosis inhibitors, including chlorpromazine and sucrose
   (clathrin-mediated endocytosis inhibitor), rottlerin (micropinocytosis
   inhibitor), nystatin (caveolae-mediated endocytosis inhibitor), and
   methyl-β-cyclodextrin (Me-β-CD, clathrin-independent endocytosis
   inhibitor), or incubated at 4 °C. As shown in [75]Figure c, after the
   above inhibitor treatment, sucrose can slightly inhibit endocytosis,
   while Me-β-CD is significantly inhibited with the cellular uptake rate
   reduced to 16.3%. Furthermore, the cellular uptake of SA-TCPP@GQDs was
   significantly inhibited when the cells were treated at 4 °C. Therefore,
   clathrin-independent and energy-dependent endocytosis was confirmed for
   the uptake of SA-TCPP@GQDs.

   Furthermore, the intracellular localization of the nanoparticles was
   monitored by using confocal laser scanning microscopy (CLSM). After
   treating HeLa cells with SA-TCPP and SA-TCPP@GQDs for 12 h, the
   accumulation of fluorescent nanoparticles in the cytosol was observed
   ([76]Figure d). Moreover, it was shown that SA-TCPP@GQDs could more
   obviously enter the lysosomes by staining of the cell nucleus and
   endolysosomes before CLSM imaging (it appears as color overlap).
   Cytotoxicity experiments under dark conditions showed that HeLa cells
   treated with 400 μg/mL SA-TCPP and SA-TCPP@GQDs still had a survival
   rate of more than 85% (). This demonstrates the low intrinsic toxicity
   of both nanomaterials in the absence of light, even after cellular
   internalization and trafficking, including potential accumulation
   within the lysosomes. Therefore, the efficient cellular uptake observed
   for SA-TCPP@GQDs, combined with its excellent biocompatibility under
   dark conditions, provides a solid foundation for its subsequent use in
   light-triggered photocatalytic tumor therapy.

   To explore the PCT potential of SA-TCPP@GQDs, the intracellular
   generation of ROS was monitored using the fluorescent probe DCFH-DA. As
   shown in [77]Figure a, DPBS, SA-TCPP, and SA-TCPP@GQDs treated cells
   showed no obvious green fluorescence signal under dark conditions,
   indicating minimal baseline ROS. After irradiation at 700 nm for 15
   min, HeLa cells treated with SA-TCPP showed enhanced fluorescence to a
   certain extent, while SA-TCPP@GQDs showed more obvious ROS green
   fluorescence. Flow cytometry was employed for the quantitative analysis
   ([78]Figure b). The results confirmed that the fluorescence intensity
   of SA-TCPP@GQDs-treated cells was 2.6 times higher than that of
   SA-TCPP, clearly indicating superior ROS generation by SA-TCPP@GQDs
   under NIR irradiation. Furthermore, active species capture experiments
   were performed to identify the primary active species. As shown in
   [79]Figure c, by comparing the apoptosis rate of HeLa cells treated
   with SA-TCPP@GQDs and then adding different trapping agents under
   lighting, it was found that sodium azide (NaN[3]) did not significantly
   reduce the killing effect after capture of ^1O[2], proving that this
   process is different from the photodynamic therapy (PDT) dependent on
   O[2]. In addition, ^1O[2], •OH, •O[2] ^–, and h^+ were captured by
   sodium azide (NaN[3]), tert-butanol (t-BuOH), p-benzoquinone (p-BQ),
   and formic acid (KI), respectively, and the cell viability rate of HeLa
   cells decreased gradually under light. The above results showed that
   holes were the main reactive oxygen species, and the contribution
   sequence of these active species to photocatalytic cell killing was h^+
   > •O[2] ^– > •OH > ^1O[2]. Furthermore, the above contribution order
   was also proven in the capture experiments of bisphenol A (BPA)
   degradation (). Additionally, the signals of the above active species
   were detected by ESR and showed varying degrees of signal enhancement
   over time ().

5.

   [80]5
   [81]Open in a new tab

   ROS levels enhanced under NIR light for effectively killing cancer
   cells. (a) Fluorescence images of intracellular ROS assay using 10 μM
   DCFH-DA after HeLa cells treated with DPBS, SA-TCPP, and SA-TCPP@GQDs
   for 12 h. Scale bar: 50 μm. The concentration of photocatalyst was 12.5
   μg/mL. The illumination conditions were 700 nm, 70 mW/cm^2, and 15 min.
   (b) Normalized fluorescence intensity of HeLa cells with the above
   treatments. n = 3, ***p < 0.001 (TTEST). (c) Active species capture
   experiment in the SA-TCPP@GQDs (12.5 μg/mL) photocatalytic killing of
   tumor cells. (d) Cytotoxicity of HeLa cells was treated by DPBS,
   SA-TCPP, and SA-TCPP@GQDs with different wavelength irradiation. (e)
   Comparison of growth inhibition of different cancer cell lines treated
   with SA-TCPP@GQDs. All cell lines were treated with 12.5 μg/mL
   SA-TCPP@GQDs for 12 h before viability was assayed.

   On this basis, the photocatalytic therapy of tumor cells was further
   explored. Under 700 nm irradiation (30 min, 70 mW/cm^2), HeLa cells
   treated with SA-TCPP@GQDs showed concentration-dependent apoptosis (),
   while SA-TCPP treatment had no obvious inhibition of cell growth. To
   explore the light response window of the two materials, HeLa cells
   treated with 12.5 μg/mL SA-TCPP@GQDs were irradiated with wavelengths
   of 600, 650, 700, and 750 nm, respectively, and the cell survival rate
   was analyzed. As shown in [82]Figure d, SA-TCPP showed certain
   inhibition of HeLa cell growth at 600 and 650 nm but had no ideal
   inhibition effect starting from 700 nm. The LC[50] value of
   SA-TCPP-treated HeLa cells (600 nm) was 12.5 μg/mL, while it was 6.25
   μg/mL for SA-TCPP@GQD-treated (700 nm) cells, demonstrating the
   low-dose and NIR-responsive photocatalytic properties. Notably, the
   cell survival rate of SA-TCPP@GQDs under the above light source
   conditions was lower than 40%, demonstrating excellent NIR
   light-responsive cell killing ability. Furthermore, under NIR-light,
   SA-TCPP@GQDs significantly enhanced photogenerated h^+ and achieved
   strong oxidative killing of a variety of tumor cells ([83]Figure e). In
   addition, the PCT efficiency of SA-TCPP@GQDs is significantly higher
   than those of other organic supramolecular materials (), which fully
   proves its high efficiency and universality, highlighting its great
   potential for biological applications.

   Furthermore, proteomic analyses have been conducted to investigate how
   photogenerated h^+ affects the cellular redox balance. HeLa cells were
   incubated with DPBS, SA-TCPP, and SA-TCPP@GQDs for 12 h, irradiated for
   20 min, and centrifuged for proteomic analysis. A total of 7,011
   proteins were quantified in the three groups of samples (, ). Principal
   component analysis (PCA) revealed significant differences among the
   three data groups, indicating distinct changes in protein expression
   under different drug treatments (). Initially, we compared the data of
   the SA-TCPP group and the SA-TCPP@GQDs group with the DPBS group,
   individually. Based on fold change greater than 2 and P-values of t
   test of less than 0.05, 786 proteins were upregulated, and 653 proteins
   were downregulated in the SA-TCPP group. In the SA-TCPP@GQDs group,
   1,203 proteins were upregulated, and 604 proteins were downregulated (,
   ).

   Subsequently, we conducted signal pathway enrichment analysis for
   differential proteins and found that the upregulated proteins in the
   SA-TCPP group were enriched in signaling pathways, such as the
   tricarboxylic acid (TCA) cycle, reactive oxygen species, mitochondrial
   ATP transmembrane transport, and mitochondrial ATP synthesis
   ([84]Figure a, ). The DNA repair and DNA damage response pathways were
   also significantly enriched. These results indicate that cells respond
   to reactive oxygen species, altering basic metabolic processes,
   affecting DNA synthesis, and intensifying DNA damage. Downregulated
   proteins were enriched in the processes of cell division and the cell
   cycle as well as in the negative regulation of apoptosis and the
   positive regulation of cell adhesion ([85]Figure a, ).

6.

   [86]6
   [87]Open in a new tab

   Proteomic analysis of HeLa cells treated with nanoparticles and
   photocatalytic therapy. Signaling pathway enrichment analysis of
   differential proteins in (a) SA-TCPP and (b) SA-TCPP@GQDs. (c)
   Schematic diagram of the photogenerated hole (h^+)-induced apoptosis of
   tumor cells.

   The signaling pathway enrichment results in the SA-TCPP@GQDs group were
   largely similar to those in the SA-TCPP group. Protein expression was
   upregulated in the TCA cycle, reactive oxygen species, and DNA damage
   response pathways and downregulated in the cell-matrix adhesion and
   negative regulation of apoptosis pathways ([88]Figure b, ). Notably,
   these consistent pathways had stronger enrichment effects in the
   SA-TCPP@GQD group. Additionally, fatty acid degradation was
   upregulated, suggesting that abnormal cell metabolism following
   SA-TCPP@GQD treatment involves not only glucose metabolism but also
   lipid metabolism. Meanwhile, regulation of lysosome size was
   downregulated in the SA-TCPP@GQDs group, suggesting that lysosome
   activity was abnormal and the apoptosis process may increase,
   demonstrating the superiority of the enhanced the NIR absorption and
   charge migration of the D–A structure.

   By further analyzing the function of the proteins involved in the
   signaling pathways (), we summarized the tumor cell effects induced by
   ROS. As illustrated in [89]Figure c, photogenerated h^+ causes
   irreversible cell damage through a variety of pathways, as described
   below: (1) The abnormal expression of enzymes in glucose and lipid
   metabolism disrupts cell metabolism and indirectly affects
   mitochondrial metabolism. This leads to enhanced respiratory chain and
   electron transport activities, increasing ATP and NAD^+ production. (2)
   Elevated ROS levels exacerbate endoplasmic reticulum (ER) stress and
   increase Ca^2+ release, potentially inhibiting DFFA’s effect on DNA
   fragmentation and thereby enhancing the DNA fragmentation process. (3)
   Some DNA damage response proteins, such as PARP1, were significantly
   enhanced, suggesting the occurrence of DNA damage. (4) Elevated ROS
   levels also inhibit certain tumor-regulated signaling pathways, such as
   the NF-κB pathway, which promotes cell proliferation, thus weakening
   the processes of cell proliferation and division. (5) Apoptosis is the
   most significant phenotype observed, with apoptosis-promoting protein
   AIFM1 being upregulated and antiapoptotic proteins BID, TRIAP1, and
   AVEN being downregulated. Notably, the antioxidant enzyme SOD1 is also
   downregulated. In summary, the excessive increase in the intracellular
   photogenerated h^+ level induces a strong intracellular stress
   response, disrupting the redox balance and leading to irreversible cell
   apoptosis.

Near-Infrared Light-Driven Complete Elimination of Solid Tumors in Vivo

   Encouraged by the high performance of SA-TCPP@GQDs to inactivate cancer
   cells, we next evaluated their efficacy for photocatalytic solid tumor
   elimination in vivo. To track the metabolic process of nanoparticles,
   mice were intravenously injected with DPBS, SA-TCPP, and SA-TCPP@GQDs
   dispersions, and the real-time distribution was observed by a small
   animal imaging system. As shown in [90]Figure a, SA-TCPP and
   SA-TCPP@GQDs exhibit time-dependent fluorescence enhancement in mice,
   and the fluorescence signals throughout the whole body appear around 9
   h and gradually fade with metabolism 12 h later. It is worth mentioning
   that porphyrin supramolecular nanoparticles can maintain stability in a
   variety of physiological solutions for days ( and ), which ensures
   their real-world applicability. Besides, the nanoparticles detected in
   mouse urine still maintain the original nanoparticle morphology (),
   demonstrating the blood circulation stability of supramolecular
   nanoparticles in complex physiological environments.

7.

   [91]7
   [92]Open in a new tab

   Targeted enrichment and photocatalytic elimination for solid tumors in
   vivo. (a) Fluorescence images of metabolic processes at different times
   (from left to right) after intravenous administration of DPBS, SA-TCPP,
   and SA-TCPP@GQDs. (b) Distribution of nanoparticles in various organs
   and tumors after HeLa tumor-bearing mice were treated with DPBS,
   SA-TCPP, and SA-TCPP@GQDs (0.5 mg/kg) for 6 h. (c) Schematic diagram of
   photocatalytic therapy process. (d) Images of mice with solid tumor or
   tumor removal after photocatalytic treatment. (e) Tumor volume of HeLa
   tumor-bearing mice treated with DPBS, SA-TCPP, and SA-TCPP@GQDs
   nanoparticles. Scale bar: 1 cm. The results were presented as mean ± SD
   (n = 6). (f) Mouse body weight changes during the treatment.

   To assess the tumor-targeted efficiency of the nanoparticles, a HeLa
   tumor bearing mouse xenograft model was established by subcutaneously
   injecting HeLa cells into the armpit of female BALB/C nude mice.
   Tumor-bearing mice were intravenously administered DPBS, SA-TCPP, or
   SA-TCPP@GQDs for subsequent fluorescence imaging analysis.
   Nanoparticles typically accumulate in solid tumors via the enhanced
   permeability and retention (EPR) effect. In this study, the
   tumor-targeting ability of SA-TCPP@GQDs primarily arises from passive
   targeting, driven by the size-dependent accumulation of supramolecular
   porphyrin-based nanoparticles. The nanoparticles exhibit time-dependent
   tumor enrichment, with notably enhanced imaging observed at 6 h
   postinjection (). As shown in [93]Figure b, after injection of
   SA-TCPP@GQDs and SA-TCPP nanoparticles into mice, significant
   fluorescence signals were generated at the tumor site. Moreover,
   relatively enhanced fluorescence signals were also observed in the
   lungs of mice in the SA-TCPP@GQD group, which can be attributed to the
   efficiency improvement brought about by the positive charge property.
   Furthermore, fluorescence monitoring of various organs further
   confirmed the targeted enrichment of the tumor (). Therefore,
   SA-TCPP@GQDs is considered as a photocatalytic therapeutic agent with
   easy biometabolism, mechanical stability, and tumor targeting
   potential.

   Considering the above metabolism and tumor enrichment time, we selected
   6 h after intravenous injection as the appropriate time for PCT.
   [94]Figure c illustrates the process of photocatalytic tumor therapy.
   To this end, the mice were treated with DPBS, SA-TCPP, and SA-TCPP@GQDs
   and irradiated under 700 nm NIR light. The changes of the tumor site
   were observed, as shown in [95]Figure d. The solid tumor was convex
   with a size of about 100 mm^3, and it turned white after treatment with
   SA-TCPP@GQDs and irradiation. The temperature change curves () and
   thermal imaging of mice ([96]Figure S27) during the treatment showed
   that the treatment was mainly focused on photocatalytic therapy rather
   than photothermal therapy caused by local thermal effects. After 2
   days, a clear black scab appeared at the radiation site, and the skin
   became flat with the elimination of the solid tumor. Moreover,
   nanoparticles did not spread to other organs after irradiation and
   continued to stay in the shrunken tumor site ().

   The in vivo antitumor efficacy of SA-TCPP@GQDs nanoparticles was
   subsequently studied by intravenous injection of different
   nanoparticles to HeLa tumor-bearing mice, including DPBS, SA-TCPP, and
   SA-TCPP@GQDs at a dosage of 0.5 mg kg^–1. Irradiation was performed 6 h
   after injection (), and tumor changes were monitored every 2 days. As
   shown in [97]Figure e, in the control group injected with DPBS, the
   tumor volume continued to grow substantially. Despite a certain
   inhibition after injection of SA-TCPP, the tumor volume also gradually
   increased over time due to the repeated growth of cancer cells without
   complete elimination. Fortunately, after being treated by SA-TCPP@GQDs,
   the solid tumor no longer showed growth after a significant reduction
   in volume, and the skin gradually returned to normal as the scab fell
   off. There was no significant change in the body weight of mice in the
   three groups ([98]Figure f), which proved that drug treatment and
   photocatalytic therapy had no obvious interference with the
   physiological state of mice and had good biocompatibility.

   Hematoxylin and eosin (H&E) staining were used for histological
   analysis of tumor sections, as shown in [99]Figure g, and the cell
   density of SA-TCPP@GQD treatment was significantly reduced compared
   with DPBS and SA-TCPP treatment. Furthermore, the use of SA-TCPP and
   SA-TCPP@GQDs nanoparticles did not cause significant hepatocellular
   injury, as revealed by liver function analysis (). These results
   highlight the significant advantages of SA-TCPP@GQDs in terms of their
   photocatalytic efficiency and biocompatibility. Compared with other
   porphyrin-based phototherapeutic materials (), SA-TCPP@GQDs demonstrate
   outstanding performance in NIR light-driven photocatalytic cancer
   therapy. Their metal-free composition avoids the risk of long-term
   biological accumulation and toxicity, while the spontaneously generated
   photogenerated holes serve as the primary active species, effectively
   overcoming the challenges posed by the deeply hypoxic tumor
   microenvironment. These combined features enable more efficient tumor
   ablation and underscore the strong potential of SA-TCPP@GQDs for future
   applications in photocatalytic cancer treatment.

Conclusion

   In summary, we successfully developed a supramolecular nanostructure,
   SA-TCPP@GQDs, featuring a well-defined D–A structure. Strong electronic
   coupling between the SA-TCPP and GQDs-NH[2] components effectively
   reduced the band gap and expanded the light absorption to the NIR
   region. Crucially, the D–A structure in SA-TCPP@GQDs generated a
   6.2-fold enhancement in the interfacial electric field, markedly
   facilitating efficient separation and migration of the photogenerated
   carriers. Consequently, once taken up by tumor cells, SA-TCPP@GQDs
   nanoparticles generated an enhanced flux of potent photogenerated holes
   (h^+) under NIR irradiation, inducing severe oxidative stress and
   leading to rapid cell apoptosis. Furthermore, in vivo studies
   demonstrated the excellent tumor-targeting capability of SA-TCPP@GQDs
   and confirmed their high therapeutic efficacy, achieving complete
   elimination of solid tumors. This work presents a novel and effective
   supramolecular D–A strategy for engineering photocatalytic materials
   with both robust NIR responsiveness and superior charge carrier
   dynamics, holding significant promise for advancing targeted and highly
   effective photocatalytic cancer therapy.

Experimental Section

Preparation of Self-Assembled TCPP (SA-TCPP)

   A 0.35 g portion of TCPP was dissolved with 15 mL of 0.1 M KOH, and the
   mixture was stirred at 60 °C until completely dissolved. Afterward, 0.1
   M HCl was added dropwise until the pH was weakly acidic (pH 6), and
   there was no significant solid precipitation. The mixture was stirred
   at room temperature for 8 h, then centrifuged and washed with deionized
   water twice to remove additional K^+ and Cl^–. SA-TCCP was collected
   and dried under a vacuum at 60 °C.

Preparation of SA-TCPP@GQDs

   The commercial GQDs-NH[2] (1 mg/mL) was added to the SA-TCCP assembly
   solution under stirring and high-power ultrasound for 3 h. Finally, the
   dark purple solid was collected by filtration with a 0.22 μM filter
   membrane and dried at 60 °C to obtain SA-TCPP@GQDs. In this work, the
   mass ratio of GQDs-NH[2] to SA-TCPP of 5% was used due to the optimal
   photocatalytic performance.

Characterization

   Microscopic morphology was observed with a transmission electron
   microscope (TEM, Hitachi HT 7700) and a high-resolution transmission
   electron microscope (HRTEM, JEOL JEM-2100F). The atomic force
   microscopy study was carried out using Cypher VRS with a Kelvin probe
   (HQ NSC18/Pt). Zeta potential and particle size were measured with a
   nanoparticle analyzer (Horiba SZ-100). X-ray diffraction (XRD) patterns
   were measured on a Rigaku (Smart-Lab) X-ray diffractometer with Cu Kα
   (λ = 1.54 Å). Fourier transform infrared (FT-IR) spectra were carried
   out using a Bruker V70 FTIR spectrometer. The ultraviolet–visible
   diffuse reflectance spectrum (DRS) was obtained by a U-3010
   spectrometer (Hitachi, Japan). Transient photovoltage (TPV) spectra
   were obtained on a photovoltage spectrometer (CEL-TPV1000) with an
   excitation pulse at 355 nm. Femtosecond transient absorption (TAS)
   spectra were collected by an amplifier laser system (35 fs, 1 kHz, 800
   nm, Spitfire Ace, Spectra Physics), two fiber optic spectrometers
   (AcaSpec-ULS2048CL-EVO, AvaSpec-NIR256-1.7-HSC-EVO), and a mid-infrared
   spectroscopy instrument (FPAS-0144) with an excitation wavelength of
   370 nm. Time-resolved photoluminescence spectra were collected on a
   fluorescence spectrometer (Edinburgh FLSP920) with an excitation
   wavelength of 430 nm. Room-temperature photoluminescence (PL) spectra
   were recorded on a fluorescence spectrometer (Edinburgh F900) with an
   excitation wavelength of 413 nm. Kelvin probe force microscopy (KPFM,
   SPM-9700) was utilized to ascertain the surface potential of the
   materials. The surface photovoltage (SPV) spectra were obtained on a
   surface photovoltage spectrometer (Zolix RAOB05–1), a home-built
   instrument as previously reported. An electrochemical workstation
   (Chenhua CHI660E) was used for photoelectrochemical measurements.
   Electron spin resonance (ESR) spectra (JEOL FA-200) were used to
   identify the main active species. Confocal laser scanning microscopy
   (CLSM) images were obtained from a laser scanning confocal microscope
   (Zeiss LSM 700). Multiskan SkyHigh (A51119770DPC) was used to analyze
   the absorbance at 491 nm of the MTT assay in cell experiments. Flow
   cytometric analysis was conducted using a CytoFLEX flow cytometry
   system (Beckman Coulter, USA). Fluorescence imaging and distribution
   study were performed on an IVIS small animal imaging system
   (PerkinElmer, USA).

Electrochemical Measurements

   The photoelectrochemical measurements were performed on an
   electrochemical workstation (Chenhua CHI660E). A 300 W Xe lamp with
   cutoff filter (λ ≥ 420 nm) was used as the light source. The Pt plate
   was used as a counter electrode, saturated Ag/AgCl as a reference
   electrode, and working electrode as a sample coated on ITO glass,
   respectively. 0.1 M Na[2]SO[4] was used as the electrolyte solution. A
   positive bias voltage of 0.2 V was provided in the electrochemical
   test.

Supplementary Material

   [100]au5c00535_si_001.pdf^ (2.4MB, pdf)
   [101]au5c00535_si_002.xlsx^ (1.4MB, xlsx)

Acknowledgments