Curcumin-coated anodized titanium surfaces for enhanced in vitro cytocompatibility and early osteogenic activity

Abstract

Objective

To characterize a carrier-free curcumin coating applied by dip-coating to acid-etched and anodized Ti-6Al-4 V (Cur-Ti) and to compare its in vitro cytocompatibility and early osteogenic activity, measured as total alkaline phosphatase (ALP) activity, with those of a sol-gel hydroxyapatite coating (HA-Ti) prepared on the same substrate and of uncoated anodized titanium.

Methods

Grade 5 titanium discs were acid-etched, anodized to form a TiO₂ nanotube layer, and coated with curcumin by dip-coating from a dimethyl sulfoxide solution or with HA by a sol-gel process. Coatings were characterized by FESEM, XRD, FTIR, confocal Raman, and UV-Vis reflectance spectroscopy. Cytocompatibility was assessed by the MTT assay in L929 fibroblasts and MG-63 osteoblast-like cells at 24 and 48 h according to ISO 10993-5, and total ALP activity was measured in MG-63 cells cultured on uncoated Ti, HA-Ti, and Cur-Ti for 7 and 14 days (n = 6 discs per group per time point). Coating thickness, curcumin loading and release, and ALP normalized to protein or DNA were not measured.

Results

Characterization confirmed a continuous, conformal, amorphous curcumin layer with intact functional groups on the anodized substrate. Relative to HA-Ti, MTT viability on Cur-Ti was higher by 13.1% and 10.6% in L929 cells and by 13.4% and 11.9% in MG-63 cells at 24 and 48 h, respectively (all p ≤ 0.003; Cohen’s d 2.2 to 6.2), and both coatings exceeded the 70% viability threshold of ISO 10993-5. Total ALP activity on Cur-Ti was 164.62 ± 3.71 U/L at day 7 and 229.82 ± 2.51 U/L at day 14, which was 70.2% and 114.9% above uncoated Ti and 24.1% and 28.8% above HA-Ti (Tukey HSD, p < 0.0001).

Conclusion

A carrier-free curcumin coating on anodized Ti-6Al-4 V supported higher cytocompatibility and higher total early osteogenic (ALP) activity than a conventional sol-gel HA coating on the same substrate. Because ALP was not normalized per cell and cell-free MTT controls were not included, these findings indicate an early osteogenic signal at the well level rather than enhanced differentiation per cell. Curcumin loading and release, antibacterial activity, wettability, coating adhesion, and corrosion behavior were not evaluated and are identified as priorities for future work before any multifunctional claim can be made.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s11671-026-05032-4.

Introduction

Titanium and its alloys are the most widely used materials for endosseous dental and orthopedic implants because of their biocompatibility, corrosion resistance, and mechanical properties that approximate those of cortical bone [1, 2]. Pristine titanium surfaces are bioinert, however, and surface modification is used to promote the biological interactions with host tissue that govern osseointegration [3, 4]. Among established strategies, hydroxyapatite (HA; Ca₁₀(PO₄)₆(OH)₂) coatings have been used extensively because of their chemical similarity to the mineral phase of bone and their capacity to promote osteoblastic activity at the implant-tissue interface [5]. Depending on their composition and deposition process, HA coatings can nevertheless be limited by poor adhesion under mechanical loading, delamination, phase instability during plasma-spray deposition, and brittle fracture, which can compromise long-term clinical performance [6–9].

These limitations have stimulated interest in bioactive coatings derived from naturally occurring compounds, particularly polyphenols. Curcumin (diferuloylmethane), the principal bioactive constituent of Curcuma longa (turmeric), has garnered considerable attention for its pleiotropic pharmacological properties, including potent anti-inflammatory, antioxidant, antimicrobial, and osteoinductive activities [10–12]. At the molecular level, curcumin mediates its osteogenic effects through several well-characterized signaling cascades: inhibition of the nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB) pathway, which relieves the suppression of osteoblastic differentiation by pro-inflammatory cytokines [13, 14]; upregulation of runt-related transcription factor 2 (RUNX2) and alkaline phosphatase (ALP) expression [15, 16]; activation of bone morphogenetic protein-2 (BMP-2)/Smad signaling [16, 17]; and context-dependent modulation of the Wnt/β-catenin pathway [18, 19].

Critical pharmacological limitations of native curcumin are its poor aqueous solubility, rapid metabolic degradation, and limited bioavailability [20]. Local delivery from an implant surface circumvents the constraints of systemic administration, and nanoparticulate or nanoporous carriers have been used to improve the aqueous dispersibility of curcumin, protect it against degradation, and sustain its release from biomaterial surfaces [21, 22]. Electrochemical anodization of titanium produces self-organized titanium dioxide (TiO₂) nanotube arrays that provide a high-surface-area, nanoporous architecture known to enhance protein adsorption, cell adhesion, and osteogenic differentiation [23–25], and these nanotubular structures also serve as reservoirs for localized drug delivery [26, 27]. Subsequent application of a bioactive coating by dip-coating is a cost-effective and scalable method of surface functionalization [28, 29]; the reproducibility of a dip-coated layer and the resulting coating dimensions must, however, be established for each formulation.

Curcumin has been combined with titanium implant surfaces in several configurations. In a systematic review of in vitro studies of curcumin-functionalized titanium, all included studies reported enhanced osteogenic activity, but half used HA-coated rather than uncoated titanium as the comparator and most originated from a single research group, which limits the generalizability of the evidence [30]. In the HA-carrier approach, curcumin is loaded onto HA-coated Ti-6Al-4 V, alone or together with vitamin K2, vitamin C, or epigallocatechin gallate, and is released from the ceramic layer [31–33]. In carrier-mediated approaches, curcumin is immobilized on TiO₂ nanoarrays through polydopamine and cyclodextrin polymer layers [34], on copper-bearing titanium through polydopamine [35], or within polyacrylonitrile nanofibers on titanium discs [36]. Direct loading of curcumin onto anodized TiO₂ nanotubes on commercially pure titanium has recently been reported to improve the adhesion, proliferation, and osteogenic differentiation of bone marrow mesenchymal stem cells [37]. Carrier-free curcumin has also been drop-cast onto anodized TiO₂ nanotubes on Ti-6Al-4 V, where it reduced Escherichia coli and Staphylococcus aureus populations by up to 43% and 38%, respectively, within 24 h and supported the adhesion, proliferation, and osteoblastic differentiation of mesenchymal stem cells; that study used drop-casting rather than dip-coating, was designed primarily as an antibacterial evaluation, and included neither an HA comparator nor the ISO 10993-5 fibroblast line [38].

The present study was designed to characterize a carrier-free curcumin coating applied by single-step dip-coating to acid-etched and anodized Ti-6Al-4 V (Cur-Ti) using FESEM, XRD, FTIR, confocal Raman microscopy, and UV-Vis reflectance spectroscopy, and to compare its in vitro cytocompatibility (MTT assay in L929 fibroblasts and MG-63 osteoblast-like cells) and early osteogenic activity (ALP activity of MG-63 cells) with those of a sol-gel HA coating prepared on the identical substrate and of uncoated anodized titanium. The contribution of the work lies in the combination of three features rather than in any one of them: (i) curcumin was applied without an HA, polymer, or cyclodextrin intermediate layer, whereas the only previous carrier-free curcumin coating on anodized Ti-6Al-4 V was drop-cast and evaluated primarily for antibacterial activity [38], and the only direct loading of curcumin onto TiO₂ nanotubes with an osteogenic endpoint used commercially pure titanium [37]; (ii) the curcumin coating was benchmarked head-to-head against a sol-gel HA coating on the same substrate, a comparison that is not available from the HA-carrier systems, in which HA is the vehicle for curcumin rather than the comparator [31–33]; and (iii) cytocompatibility was assessed in both the ISO 10993-5 reference fibroblast line and an osteoblast-like line, with exact p-values, confidence intervals, and effect sizes, by a group independent of that responsible for most of the published evidence [30]. The hypothesis was that the curcumin coating would provide cytocompatibility and early osteogenic (ALP) activity at least equivalent to those of the HA coating. Curcumin loading and release, antibacterial activity, wettability, coating adhesion, and corrosion behavior were outside the scope of this first-stage in vitro study and are identified explicitly as untested in Sect.  4.4.

Materials and methods

Sample preparation

Titanium substrate preparation and anodization

Grade 5 titanium alloy (Ti-6Al-4 V) discs (10 mm in diameter, 2 mm in thickness) were cleaned sequentially by ultrasonic treatment in acetone, absolute ethanol, and deionized water (15 min each) and dried under ambient conditions. The cleaned discs were etched for 30 s at room temperature in a freshly prepared 1:1 (v/v) mixture of 6% hydrofluoric acid (HF) and 12% hydrochloric acid (HCl) to remove the native oxide layer, homogenize surface reactivity, and produce a uniformly micro-roughened surface; they were then rinsed immediately with deionized water and dried. The etched discs were anodized electrochemically to form a TiO₂ nanotube layer, following the protocol of Crenn et al. [23]. Anodization was performed at room temperature in a two-electrode configuration, with the titanium disc as the anode and a platinum foil cathode, in an electrolyte of 50 wt% glycerol, 49.5 wt% distilled water, and 0.5 wt% HF to which glacial acetic acid was added to a final concentration of 1 M, because acetic acid yields mechanically more robust nanotubes without altering their morphology [23]. A constant potential of 20 V was applied for 1 h, a condition reported to produce self-organized TiO₂ nanotubes on Ti-6Al-4 V with an inner diameter of approximately 70 nm and a length of approximately 240 nm while increasing surface hydrophilicity [23]. After anodization, the discs were rinsed thoroughly with deionized water and dried under ambient conditions. The discs were not annealed and were used in the as-anodized state; the oxide layer therefore remained amorphous, which is consistent with the absence of anatase or rutile reflections in the XRD pattern (Fig. 2). All groups shared this common substrate: the HA-Ti and Cur-Ti specimens received their respective coatings, whereas the uncoated control consisted of identically cleaned, acid-etched, and anodized discs from the same anodization batch without further coating, so that the groups differed only in the surface coating applied.

Fig. 2.

X-ray diffraction patterns of the hydroxyapatite-coated titanium disc (HA-Ti, top), the curcumin-coated anodized titanium disc (Cur-Ti, middle), and crystalline curcumin powder as a reference (bottom), recorded with Cu Kα radiation (λ = 1.5406 Å) over a 2θ range of 10° to 90°. Intensities are normalized and vertically offset for clarity

Hydroxyapatite coating via sol-gel technique

HA coatings were deposited on titanium discs using the sol-gel technique. The sol was prepared by dissolving calcium and phosphorus precursors (calcium nitrate tetrahydrate and triethyl phosphite) in an ethanol-based solution at a Ca/P molar ratio of 1.67, and the mixture was stirred continuously for 24 h to obtain a homogeneous sol. The surface-prepared titanium discs (Sect.  2.1.1) were then coated by dip-coating, with the process repeated for three to five layers to ensure uniform coverage. Each deposited layer was dried at 60–80 °C for 10–15 min before the subsequent coat was applied. Following the final coating, the discs underwent sintering at 500–700 °C for one to two hours in air to facilitate crystallization of the amorphous calcium phosphate film into a stable hydroxyapatite phase [5, 7].

Curcumin coating via dip-coating

A curcumin coating solution was prepared by dissolving curcumin (Sigma-Aldrich, St. Louis, MO, USA) in dimethyl sulfoxide (DMSO) at 5 mg/mL (approximately 13.6 mM) with magnetic stirring for 30 min at room temperature, protected from light. Curcumin is freely soluble in DMSO (reported solubility approximately 25 mg/mL), and a clear, homogeneous yellow-orange solution with no visible particulate was obtained; no suspension was formed, and a single solution batch was used for all Cur-Ti specimens. The acid-etched and anodized discs (Sect.  2.1.1) were coated by dip-coating [28, 29]. Each disc was held vertically with clean forceps, immersed fully in the curcumin solution for 60 s, and withdrawn manually at a slow, steady rate of approximately 5 mm per second to allow a uniform liquid film to drain. Three dip cycles were applied, with the disc air-dried for 10 min between cycles, to build a continuous coating on the nanotextured surface. After the final cycle the discs were dried at room temperature (approximately 25 °C) in the dark in a laminar-flow hood for 24 h; all discs were coated in a single session by the same operator. Solvent evaporation left a continuous curcumin layer on the titanium surface, the continuity of which was confirmed by FESEM (Fig. 1) and the amorphous character of which was indicated by XRD (Fig. 2). The specimens were designated Cur-Ti (curcumin-coated anodized titanium) and the HA-coated specimens HA-Ti. Coating reproducibility across specimens was controlled procedurally (identical solution batch, immersion, withdrawal, and drying conditions, and visual inspection of every disc for a uniform, continuous yellow-orange layer before use); between-batch reproducibility was not assessed. The mass of curcumin deposited per disc and its release into the culture medium were not quantified, so the actual curcumin exposure experienced by the cells is unknown; this is discussed as a limitation in Sect.  4.5.

Fig. 1.

FESEM micrograph (4000×; horizontal field width 104 μm; 15 kV; scale bar 20 μm) of the Cur-Ti surface showing continuous, conformal coverage of the acid-etched and anodized Ti-6Al-4 V substrate by the curcumin coating. Individual nanotubes and features below approximately 100 nm are not resolved at this magnification

Physicochemical characterization

Surface morphological analysis (FESEM)

Surface topography and coating morphology were assessed by field-emission scanning electron microscopy (FESEM; FEI Nova NanoSEM 450, FEI Company, Hillsboro, OR, USA). Discs were mounted on aluminum stubs with conductive carbon adhesive tape and examined in high-vacuum mode. Images were acquired in secondary electron mode with the Everhart-Thornley detector (ETD) at an accelerating voltage of 15 kV, spot size 10, and a working distance of 16.7 mm; these settings, together with the magnification, horizontal field width, and scale bar, are recorded in the instrument data bar embedded in the micrograph. Several fields across the central region of each disc were surveyed at low magnification (500 to 1000×) to confirm uniform coverage before a representative field was recorded. The micrograph reported (Fig. 1) was acquired at 4000× magnification (horizontal field width 104 μm; scale bar 20 μm), which resolves the micrometer-scale topography and the continuity of the coating but not individual nanotubes or features below approximately 100 nm; no higher-magnification image was recorded. Coating thickness, particle size distribution, and quantitative surface coverage were not measured.

Crystalline phase analysis (XRD)

The crystallographic phase composition of the coated specimens was analyzed by X-ray diffraction (XRD; X’Pert Pro, PANalytical, Almelo, The Netherlands) using Cu Kα radiation (λ = 1.5406 Å) over a 2θ range of 10° to 90° with a step size of 0.02°. Diffraction patterns were compared with the Joint Committee on Powder Diffraction Standards (JCPDS) reference files 44-1294 for α-phase titanium and 09-0432 for hydroxyapatite, and crystalline curcumin powder was recorded as a reference for the curcumin coating [39].

Chemical analysis (FTIR spectroscopy)

Functional group identification was performed by Fourier-transform infrared spectroscopy (Nicolet iS50; Thermo Fisher Scientific, Waltham, MA, USA) in the spectral range of 4000–400 cm⁻¹. Characteristic absorption bands associated with the curcumin molecular structure, including mixed C = C/C = O stretching of the conjugated β-diketone moiety, aromatic ring vibrations, and phenolic C–O bonds, were specifically assessed to confirm chemical integrity following deposition [40]. For the hydroxyapatite-coated titanium discs fabricated via the sol-gel technique, FTIR analysis was performed to verify the successful formation of the HA phase. The presence of characteristic phosphate (PO₄³⁻) absorption bands at approximately 560–600 cm⁻¹ and 1000–1100 cm⁻¹, corresponding to the ν₄ and ν₃ vibrational modes respectively, along with hydroxyl (OH⁻) stretching bands in the region of 3300–3600 cm⁻¹, was used to confirm the crystalline hydroxyapatite structure [5].

Molecular characterization (confocal Raman microscopy)

Confocal Raman microscopy (Alpha300 RA; WITec GmbH, Ulm, Germany) was used to confirm the molecular identity and distribution of curcumin on the anodized titanium surface. Spectra were excited with a 532 nm diode-pumped solid-state laser at a power below 5 mW at the sample surface to prevent thermal decomposition of curcumin. Spectra were acquired at multiple locations across the coated surface, and the intensity of the curcumin band near 1600 cm⁻¹ was used to compare the amount of curcumin among locations; a spatially resolved chemical map was not acquired.

Optical characterization (UV-Vis spectroscopy)

Reflectance characteristics of the coated specimens were evaluated using a UV–Vis spectrophotometer (Evolution One Plus; Thermo Fisher Scientific, Waltham, MA, USA) over the wavelength range of 440–700 nm to confirm the presence and distribution of the curcumin chromophore.

Biological evaluation

Cell culture

Mouse fibroblastic L929 cells (ATCC CCL-1) and human osteosarcoma-derived MG-63 cells (ATCC CRL-1427), both commercially purchased from the National Center for Cell Science, Pune, India, were cultured in Dulbecco’s modified Eagle medium (DMEM) supplemented with 10% fetal bovine serum (FBS), 50 IU mL^− 1 penicillin, and 50 µg mL^− 1 streptomycin, and maintained at 37 °C in a humidified atmosphere containing 5% CO₂.

MTT cytotoxicity assay

Cytotoxicity was assessed in accordance with the principles of ISO 10993-5:2009. L929 mouse fibroblasts and MG-63 human osteoblast-like cells were seeded independently at 1 × 10⁴ cells/mL into 24-well plates containing Cur-Ti or HA-Ti discs (n = 6 discs per group per time point) and incubated for 24 and 48 h. Performing the assay in both cell lines allowed cytocompatibility to be assessed in the ISO 10993-5 fibroblast model (L929) and in an osteoblast-like line (MG-63) that is directly relevant to the osteogenic context of the study. MTT reagent (0.5 mg/mL) was added to each well, the plates were incubated for 4 hours, the formazan crystals were solubilized in DMSO, and absorbance was measured at 540 nm. Results were expressed as percentage viability relative to the untreated control, defined as cells seeded at the same density on tissue-culture polystyrene in wells without a disc and cultured under otherwise identical conditions; this group was set at 100% and served as the negative (blank) reference in the sense of ISO 10993-5 [41]. No uncoated titanium disc group was included in the MTT arm; the uncoated anodized titanium reference was used only in the ALP arm (Sect.  2.3.3). Viability on Cur-Ti and HA-Ti was therefore judged against the 70% non-cytotoxicity threshold of ISO 10993-5 relative to the polystyrene blank, with HA-Ti serving as the material comparator; a consequence of this design is that the contribution of the anodized substrate itself to the MTT signal cannot be separated from that of the coating (Sect.  4.5). Cell-free control wells (discs incubated with MTT reagent in medium without cells) were not included. Because curcumin and other redox-active polyphenols can reduce tetrazolium salts directly and generate formazan in the absence of cells [42], a curcumin-related contribution to the absorbance signal on Cur-Ti cannot be excluded, and the MTT differences between groups are interpreted with this caveat (Sect.  4.5).

Alkaline phosphatase (ALP) activity assay

Early osteogenic differentiation was assessed by measuring ALP activity in MG-63 cells cultured for 7 and 14 days on uncoated, acid-etched, and anodized titanium discs (Control), HA-Ti, and Cur-Ti (n = 6 discs per group per time point). All three groups were maintained in basal DMEM supplemented with 10% fetal bovine serum, L-glutamine, sodium bicarbonate, and antibiotic solution (penicillin 100 U/mL, streptomycin 100 µg/mL, amphotericin B 2.5 µg/mL) at 37 °C in a humidified 5% CO₂ atmosphere, without osteogenic induction supplements (no added ascorbic acid, β-glycerophosphate, or dexamethasone); differences in ALP activity therefore reflect the influence of the respective surface treatments rather than chemical induction by the medium. ALP activity was quantified with p-nitrophenyl phosphate (pNPP) as the chromogenic substrate, with absorbance measured at 405 nm [43], and was calculated as ALP activity (U/L) = (ΔOD/min) × 2754, where ΔOD/min is the change in absorbance per minute at 405 nm [43]. ALP activity was expressed per unit volume (U/L) and was not normalized to total protein or DNA content; total protein, DNA content, and cell number were not measured in parallel, so the reported values represent total well-level enzyme activity rather than specific activity per cell, and an increase in total ALP cannot by itself distinguish greater osteogenic differentiation per cell from a greater number of adherent cells (Sect.  4.5).

Statistical analysis

All biological experiments were performed with six independent replicates (n = 6 separate discs per group per time point), and data are expressed as mean ± standard deviation (SD). For the MTT assay, the two-group comparisons between Cur-Ti and HA-Ti at each time point in each cell line were analyzed with the unpaired two-sided Student’s t-test. For the ALP data, which comprise three groups (Control, HA-Ti, and Cur-Ti) at two time points (day 7 and day 14), a two-way analysis of variance (ANOVA) with group and time as factors was performed, followed by Tukey’s honestly significant difference (HSD) test for pairwise comparisons among the six group × time means. A p-value below 0.05 was considered statistically significant; in the figures, significance is denoted as *p < 0.05, **p < 0.01, and ***p < 0.001. Analyses were performed with GraphPad Prism (version 10; GraphPad Software, San Diego, CA, USA). Normality and homogeneity of variance were assessed with the Shapiro-Wilk and Levene tests, respectively; for the ALP data set the model residuals were normally distributed (Shapiro-Wilk p = 0.22) with homogeneous variances (Levene p = 0.86), supporting the parametric two-way ANOVA. For the MTT comparisons, for which these assumptions were less consistently met at this sample size, significance was additionally confirmed with the non-parametric Mann-Whitney U test (all p < 0.02). Effect sizes are reported as partial η² for the ANOVA factors and as Cohen’s d for the MTT comparisons. Exact p-values are reported; values below the reporting floor of GraphPad Prism (p < 0.0001) were calculated from the test statistic and its degrees of freedom. Percentage differences between groups were calculated from the group means as 100 × (mean of the test group − mean of the reference group)/mean of the reference group, and 95% confidence intervals (CI) for the mean differences were derived from the pooled SD (MTT; t distribution with 10 degrees of freedom) or from the ANOVA residual mean square with Tukey adjustment (ALP). These derived quantities are summarized in Table 1.

Table 1.

MTT cell viability (%) on Cur-Ti and HA-Ti for L929 fibroblasts and MG-63 osteoblast-like cells at 24 and 48 h (mean ± SD; n = 6 discs per group per time point). Per-well values are provided in Supplementary Table S1 and in the data set deposited on Figshare (https://doi.org/10.6084/m9.figshare.33089459)

Cell line	Time point	Cur-Ti (%)	HA-Ti (%)	Difference (percentage points)	Change vs. HA-Ti (%)	p-value	Cohen’s d
L929	24 h	96.4 ± 1.9	85.2 ± 3.9	11.2	+ 13.1	8.6 × 10− 5	3.65
L929	48 h	98.1 ± 4.4	88.7 ± 4.0	9.4	+ 10.6	0.003	2.24
MG-63	24 h	94.8 ± 1.1	83.6 ± 2.3	11.2	+ 13.4	8.1 × 10− 7	6.21
MG-63	48 h	96.5 ± 2.1	86.2 ± 5.1	10.3	+ 11.9	0.001	2.64

Viability is expressed relative to untreated cells on tissue-culture polystyrene (100%). Change vs. HA-Ti = 100 × (Cur-Ti − HA-Ti)/HA-Ti. p-values are from the unpaired two-sided Student’s t-test (df = 10); values below 0.0001 were calculated from the t statistic. Cohen’s d was calculated from the pooled SD

Results

Physicochemical characterization

Surface topographical analysis (FESEM)

FESEM (Fig. 1) showed continuous, conformal coverage of the prepared titanium surface by the curcumin coating, which followed the micro-roughened topography produced by acid etching and anodization without visible cracking, delamination, or large-scale agglomeration at the magnification examined (4000×; horizontal field width 104 μm). Individual nanotubes and particles below approximately 100 nm are not resolvable at this magnification, so the micrograph documents coating continuity and conformity rather than nanoscale morphology; particle size distribution, coating thickness, and quantitative surface coverage were not determined (Sect.  4.5).

Crystalline phase analysis (XRD)

XRD patterns of the Cur-Ti and HA-Ti specimens, together with crystalline curcumin powder as a reference, are shown in Fig. 2. The Cur-Ti pattern showed the characteristic reflections of α-phase titanium at 2θ values of 35.43°, 40.46°, 63.57°, 74.93°, and 78.15°, corresponding to the (100), (101), (110), (103), and (112) crystallographic planes, respectively, consistent with JCPDS reference file 44-1294. No distinct crystalline curcumin reflections were detected, and the low, uniform background across the 2θ range of 10° to 30°, where crystalline curcumin shows its strongest reflections, indicates that curcumin is present in an amorphous state, or at a quantity below the detection limit of XRD, on the disc surface; both interpretations are consistent with deposition from a DMSO solution followed by ambient drying. No anatase or rutile reflections were detected, consistent with the amorphous as-anodized oxide layer. The HA-Ti pattern showed additional reflections at 2θ values of 30.30°, 34.63°, and 39.65°, attributable to a calcium phosphate/HA reflection (approximately 30.3°) and to HA reflections overlapping the Ti (002)/(112) and Ti (101) reflections, respectively, consistent with a partially crystalline calcium phosphate/HA coating. The elevated background in the low-angle region (2θ = 10° to 20°) of the HA-Ti pattern, reaching approximately 17,840 counts at 2θ = 10°, indicates an amorphous calcium phosphate component within the coating, confirming the deposition of a partially crystalline hydroxyapatite layer on the titanium substrate.

Chemical analysis (FTIR)

FTIR analysis (Fig. 3) of the curcumin-loaded Ti disc, compared with the hydroxyapatite (HA)-coated specimens, revealed characteristic absorption bands that confirm the presence of distinct functional groups in each sample. The curcumin-loaded Ti disc spectrum (blue) exhibited a broad absorption band centered at approximately 3400 cm⁻¹, attributable to O–H stretching vibrations, alongside a notable absorption at ~ 2900 cm⁻¹ corresponding to C–H stretching. Prominent absorption bands were observed in the fingerprint region at ~ 1600 cm⁻¹ and ~ 1500 cm⁻¹, consistent with C = C aromatic ring stretching and C = O enol stretching characteristic of the curcumin molecular structure. Additional bands in the 1200–1000 cm⁻¹ region are attributable to C–O–C stretching vibrations. The HA-coated specimens (red) displayed a strong ν₃ P–O stretching envelope at ~ 1030 cm⁻¹ with a shoulder near ~ 1090 cm⁻¹, characteristic of the phosphate group (PO₄³⁻), together with the ν₁ P–O symmetric stretch at ~ 962 cm⁻¹, along with a sharp structural hydroxyl (O–H) stretching band at ~ 3572 cm⁻¹ and the ν₄ O–P–O bending doublet at ~ 563 and ~ 603 cm⁻¹, collectively consistent with the FTIR fingerprint of hydroxyapatite. The presence of these phosphate bands in the HA spectrum and their absence in the curcumin-loaded Ti disc spectrum confirm the distinct chemical compositions of the two coating types and support the successful functionalization of the respective titanium surfaces.

Fig. 3.

FTIR spectra of curcumin-coated titanium and hydroxyapatite-coated titanium

Fourier-transform infrared spectra of the curcumin-loaded Ti disc and hydroxyapatite-coated Ti disc showing characteristic functional-group signatures of the respective coatings. The curcumin-loaded Ti spectrum demonstrates a broad O–H stretching band at approximately 3400 cm⁻¹, C–H stretching near 2900 cm⁻¹, and prominent fingerprint-region bands around 1600 and 1500 cm⁻¹ corresponding to aromatic C = C/C = O and enolic C = O vibrations, with additional C–O–C bands in the 1200–1000 cm⁻¹ region. The HA-Ti spectrum shows characteristic hydroxyapatite bands, including structural O–H stretching near 3572 cm⁻¹, ν₃ phosphate stretching around 1030 cm⁻¹ with a shoulder near 1090 cm⁻¹, ν₁ phosphate stretching near 962 cm⁻¹, and ν₄ O–P–O bending bands at approximately 563 and 603 cm⁻¹. These spectral features confirm the distinct chemical composition of the curcumin and hydroxyapatite coatings on titanium surfaces.

Confocal Raman microscopy

The Raman spectra of the Cur-Ti specimen (Fig. 4) exhibited characteristic bands in the 1625 to 1595 cm⁻¹ region corresponding to aromatic C = C and conjugated C = O stretching vibrations of the curcumin molecule. Additional bands were detected at approximately 1430 cm⁻¹, 1279 to 1204 cm⁻¹, and 970 cm⁻¹, consistent with aromatic ring modes and enolic β-diketone vibrations. Minor shifts from the reference values for crystalline curcumin (for example, a band near 1485 cm⁻¹ versus the literature value of approximately 1430 cm⁻¹) may reflect substrate-molecule interactions on the titanium surface or the amorphous state of the deposit. The region below 750 cm⁻¹ was dominated by the broad bands of the titanium oxide substrate. Spectra acquired at multiple surface locations and referenced to the intensity of the band near 1600 cm⁻¹ indicated a broadly distributed, though microscopically heterogeneous, curcumin coating; because no chemical map was acquired, these spectra do not quantify surface coverage or coating homogeneity.

Fig. 4.

Confocal Raman spectra acquired from the Cur-Ti surface showing the curcumin spectral signature (1250 to 1650 cm⁻¹ region, shaded) and the titanium oxide (TiO₂) background over 0 to 2000 cm⁻¹. Spectra are normalized and vertically offset for clarity; they do not constitute a spatially resolved chemical map

UV-Vis spectroscopy

UV-Vis diffuse reflectance spectroscopy (Fig. 5) was performed on the HA-Ti and Cur-Ti disc specimens across the visible wavelength range of 440 to 700 nm. Both samples exhibited near-zero reflectance at 440 nm, indicating strong absorption in the violet region of the visible spectrum. The HA-Ti specimens exhibited a sharp increase in reflectance between 440 and 465 nm, rising steeply to approximately 35%, after which reflectance remained essentially constant (approximately 35%) across the remainder of the measured spectrum up to 700 nm. This profile is consistent with the optical behavior of hydroxyapatite, which absorbs strongly in the UV and short-wavelength visible range before reaching a high, stable reflectance plateau, indicative of a wide bandgap and an optically inert white coating.

Fig. 5.

UV-Vis diffuse reflectance spectra of the hydroxyapatite-coated titanium disc (HA-Ti) and the curcumin-coated anodized titanium disc (Cur-Ti) over the visible range of 440 to 700 nm

In contrast, the Cur-Ti specimens exhibited a markedly different and more gradual reflectance profile. Reflectance increased progressively from 0% at 440 nm to approximately 9% at 465 nm, 15% at 546 nm, and 36% at 590 nm, before plateauing at approximately 38% by 700 nm. This gradual rise in reflectance across the blue-to-green visible range (440 to 590 nm) is characteristic of the optical absorption behavior of curcumin, which features a strong absorption band centered at approximately 420 to 430 nm attributable to π→π* electronic transitions in the conjugated enol form of the molecule. The progressive increase in reflectance with wavelength marks the absorption edge of curcumin, and the yellow-orange color of the coating is consistent with strong absorption of violet and blue light and reflectance at longer wavelengths above approximately 590 nm. Both coatings converged to a similar reflectance (approximately 35% to 38%) above 590 nm, indicating comparable optical behavior in the red region of the spectrum. These reflectance measurements confirm the presence of the curcumin chromophore on the surface but were not calibrated against deposited curcumin mass and cannot be used to estimate loading.

Biological evaluation

MTT cytotoxicity assay

The MTT assay showed higher cell viability on Cur-Ti than on HA-Ti in both L929 fibroblasts and MG-63 osteoblast-like cells at both time points (Table 1; Fig. 6). In L929 fibroblasts, viability on Cur-Ti was 96.4 ± 1.9% at 24 h and 98.1 ± 4.4% at 48 h, compared with 85.2 ± 3.9% and 88.7 ± 4.0% on HA-Ti, corresponding to relative increases of 13.1% (mean difference 11.2 percentage points, 95% CI 7.3 to 15.1; t(10) = 6.32, p = 8.6 × 10^− 5, Cohen’s d = 3.65) and 10.6% (9.4 points, 95% CI 4.0 to 14.8; t(10) = 3.87, p = 0.003, d = 2.24). In MG-63 cells, viability on Cur-Ti was 94.8 ± 1.1% at 24 h and 96.5 ± 2.1% at 48 h, compared with 83.6 ± 2.3% and 86.2 ± 5.1% on HA-Ti, corresponding to relative increases of 13.4% (11.2 points, 95% CI 8.9 to 13.5; t(10) = 10.76, p = 8.1 × 10^− 7, d = 6.21) and 11.9% (10.3 points, 95% CI 5.3 to 15.3; t(10) = 4.57, p = 0.001, d = 2.64). Relative to the untreated control (100%), viability on Cur-Ti was 1.9% to 5.2% lower and viability on HA-Ti was 11.3% to 16.4% lower (Table 3). Across both cell lines and both time points, both coatings exceeded the 70% viability threshold defined by ISO 10993-5:2009 for a non-cytotoxic classification. The consistent difference in favor of Cur-Ti in fibroblasts and osteoblast-like cells provides evidence across two cell lines of the favorable cytocompatibility of the curcumin coating. Because the MTT assay reports mitochondrial dehydrogenase activity rather than cell number, curcumin is redox-active and a recognized interferent in tetrazolium-based assays [42], and no cell-free disc controls were run, the magnitude of the difference between Cur-Ti and HA-Ti should be interpreted with caution: the non-cytotoxic classification of HA-Ti, which contains no curcumin, is unaffected by this caveat, whereas the Cur-Ti values should be regarded as an upper bound on true viability until confirmed with a label-independent endpoint.

Fig. 6.

MTT cell viability of L929 fibroblasts and MG-63 osteoblast-like cells after 24 h and 48 h on Cur-Ti (green) and HA-Ti (blue) discs (n = 6 per group; mean ± SD). Dashed line, ISO 10993-5 70% viability threshold. Significance thresholds are from the unpaired Student’s t-test; exact p-values, mean differences, and 95% CI are given in Tables 1 and 3

Table 3.

Quantitative comparison of cell viability (MTT) and total alkaline phosphatase (ALP) activity among uncoated anodized Ti (Control), HA-Ti, and Cur-Ti (n = 6 per group per time point)

Endpoint	Comparison	Mean difference (95% CI)	Change (%)	Test statistic	p-value	Effect size
MTT, L929, 24 h	Cur-Ti vs. HA-Ti	11.2 points (7.3 to 15.1)	+ 13.1	t(10) = 6.32	8.6 × 10− 5	d = 3.65
MTT, L929, 48 h	Cur-Ti vs. HA-Ti	9.4 points (4.0 to 14.8)	+ 10.6	t(10) = 3.87	0.003	d = 2.24
MTT, MG-63, 24 h	Cur-Ti vs. HA-Ti	11.2 points (8.9 to 13.5)	+ 13.4	t(10) = 10.76	8.1 × 10− 7	d = 6.21
MTT, MG-63, 48 h	Cur-Ti vs. HA-Ti	10.3 points (5.3 to 15.3)	+ 11.9	t(10) = 4.57	0.001	d = 2.64
ALP, day 7	HA-Ti vs. uncoated Ti	35.95 U/L (31.1 to 40.8)	+ 37.2	q = 31.7	< 0.0001	partial η² (group) = 0.996
ALP, day 7	Cur-Ti vs. uncoated Ti	67.91 U/L (63.0 to 72.8)	+ 70.2	q = 59.9	< 0.0001
ALP, day 7	Cur-Ti vs. HA-Ti	31.96 U/L (27.1 to 36.8)	+ 24.1	q = 28.2	< 0.0001
ALP, day 14	HA-Ti vs. uncoated Ti	71.54 U/L (66.7 to 76.4)	+ 66.9	q = 63.1	< 0.0001	partial η² (time) = 0.985
ALP, day 14	Cur-Ti vs. uncoated Ti	122.88 U/L (118.0 to 127.8)	+ 114.9	q = 108.4	< 0.0001
ALP, day 14	Cur-Ti vs. HA-Ti	51.34 U/L (46.5 to 56.2)	+ 28.8	q = 45.3	< 0.0001
ALP, day 14 vs. day 7	Uncoated Ti	10.23 U/L (5.4 to 15.1)	+ 10.6	q = 9.0	< 0.0001	partial η² (interaction) = 0.953
ALP, day 14 vs. day 7	HA-Ti	45.82 U/L (40.9 to 50.7)	+ 34.5	q = 40.4	< 0.0001
ALP, day 14 vs. day 7	Cur-Ti	65.20 U/L (60.3 to 70.1)	+ 39.6	q = 57.5	< 0.0001

Alkaline phosphatase (ALP) activity

ALP activity (Fig. 7; Table 2), a marker of early osteogenic differentiation, was assessed at day 7 and day 14 in all three groups (n = 6 per group). Two-way ANOVA showed significant main effects of group (F(2,30) = 3560, p < 0.0001, partial η² = 0.996) and time (F(1,30) = 1907, p < 0.0001, partial η² = 0.985) and a significant group × time interaction (F(2,30) = 302, p < 0.0001, partial η² = 0.953). All groups showed an increase in ALP activity between day 7 and day 14 (Tukey HSD, p < 0.0001 for each group). On day 7, the Cur-Ti group exhibited the highest ALP activity (164.62 ± 3.71 U/L), followed by HA-Ti (132.66 ± 2.16 U/L) and the control group (cells on uncoated anodized titanium; 96.71 ± 2.34 U/L). This hierarchy was maintained and accentuated at day 14, with Cur-Ti recording the greatest ALP activity (229.82 ± 2.51 U/L), higher than both HA-Ti (178.48 ± 1.22 U/L) and the control (106.94 ± 3.82 U/L). Relative to the uncoated titanium control, ALP activity was 70.2% higher on Cur-Ti and 37.2% higher on HA-Ti at day 7, and 114.9% higher on Cur-Ti and 66.9% higher on HA-Ti at day 14 (Tukey HSD, all p < 0.0001; Table 3), consistent with the established osteoconductive properties of hydroxyapatite coatings and with a larger effect of the curcumin coating. The Cur-Ti group exhibited 24.1% greater ALP activity than HA-Ti at day 7 (mean difference 31.96 U/L, Tukey-adjusted 95% CI 27.1 to 36.8) and 28.8% greater at day 14 (51.34 U/L, 95% CI 46.5 to 56.2; both p < 0.0001), indicating that Cur-Ti supported higher total ALP activity than the hydroxyapatite coating. Because ALP activity was expressed per unit volume (U/L) and was not normalized to total protein, DNA, or cell number, and Cur-Ti also supported modestly higher cell viability in the MTT assay, this higher total ALP activity cannot be attributed unambiguously to an increase in per-cell osteogenic differentiation. The relative ALP difference between Cur-Ti and HA-Ti (24% to 29%) exceeded the cell-viability difference observed in the MTT assay at 24 to 48 h (11% to 13%), which argues against cell number as the sole explanation, but this comparison is indirect because cell number was not measured at day 7 or day 14.

Fig. 7.

Alkaline phosphatase (ALP) activity of MG-63 osteoblast-like cells on Control (uncoated anodized titanium), HA-Ti, and Cur-Ti surfaces at day 7 (brown) and day 14 (orange) (n = 6 per group; mean ± SD). ***p < 0.001 (two-way ANOVA with Tukey HSD post-hoc test); exact p-values were below 0.0001 for all pairwise comparisons (Table 3). Activity is expressed as U/L and is not normalized to protein, DNA, or cell number

Table 2.

Alkaline phosphatase (ALP) activity of MG-63 cells on uncoated anodized titanium discs (Control), HA-Ti, and Cur-Ti at day 7 and day 14. Values are U/L (ALP activity = ΔOD/min × 2754), expressed as mean ± SD and standard error (SE) from six independent replicates (n = 6); activity is not normalized to protein, DNA, or cell number. Group means were compared by two-way ANOVA (group × time) with Tukey HSD post-hoc test; means not sharing a letter in the Tukey group column differ significantly (p < 0.05), and all six group × time means differed at p < 0.0001 (partial η²: group 0.996, time 0.985, interaction 0.953). Per-well values are provided in Supplementary Table S2 and in the data set deposited on Figshare

Group	Time point	Mean ± SD (U/L)	SE (U/L)	n	Change vs. uncoated Ti (%)	Change vs. HA-Ti (%)	Tukey group
Control	Day 7	96.71 ± 2.34	0.96	6	Reference		a
HA-Ti	Day 7	132.66 ± 2.16	0.88	6	+ 37.2	Reference	c
Cur-Ti	Day 7	164.62 ± 3.71	1.51	6	+ 70.2	+ 24.1	d
Control	Day 14	106.94 ± 3.82	1.56	6	Reference		b
HA-Ti	Day 14	178.48 ± 1.22	0.50	6	+ 66.9	Reference	e
Cur-Ti	Day 14	229.82 ± 2.51	1.02	6	+ 114.9	+ 28.8	f

Comparators are time-matched. Percentage change = 100 × (mean difference)/(reference mean)

Quantitative comparison among uncoated Ti, HA-Ti, and Cur-Ti

To express the biological differences among groups quantitatively, mean differences with 95% confidence intervals, percentage changes relative to the reference group, test statistics, exact p-values, and effect sizes were calculated from the group means and standard deviations in Tables 1 and 2 (n = 6 per group) and are summarized in Table 3. Relative to uncoated titanium, total ALP activity was 37.2% (day 7) and 66.9% (day 14) higher on HA-Ti and 70.2% (day 7) and 114.9% (day 14) higher on Cur-Ti; relative to HA-Ti, it was 24.1% and 28.8% higher on Cur-Ti. In the MTT assay, viability on Cur-Ti exceeded that on HA-Ti by 10.6% to 13.4% across cell lines and time points. Wettability, curcumin loading and release, antibacterial activity, coating adhesion, and corrosion behavior were not measured and are therefore not included.

Mean differences, confidence intervals, and test statistics were calculated from the group means and standard deviations reported in Tables 1 and 2 (n = 6 per group). MTT: unpaired Student’s t-test with pooled SD; 95% CI from the t distribution with 10 degrees of freedom; d, Cohen’s d; p-values below the GraphPad Prism reporting floor (0.0001) were calculated from the t statistic. ALP: Tukey HSD on the six group × time means (k = 6; residual degrees of freedom 30; residual mean square 7.71 (U/L)²); q, studentized range statistic; Tukey-adjusted 95% CI. Change (%) = 100 × (mean difference)/(reference mean). Relative to the untreated control set at 100%, MTT viability on Cur-Ti was 3.6%, 1.9%, 5.2%, and 3.5% lower (L929 24 h, L929 48 h, MG-63 24 h, MG-63 48 h) and on HA-Ti was 14.8%, 11.3%, 16.4%, and 13.8% lower; an uncoated titanium disc group was not included in the MTT arm. The p-values calculated from the summary statistics are consistent with the significance thresholds obtained by GraphPad Prism from the per-well data (p < 0.001 at 24 h and p < 0.01 at 48 h; Fig. 6).

Discussion

The present study provides evidence that a carrier-free curcumin coating on anodized Ti-6Al-4 V supports higher in vitro cytocompatibility and higher total early osteogenic (ALP) activity than a conventional sol-gel hydroxyapatite coating prepared on the same substrate, at the level of the endpoints measured. These findings contribute to the growing body of literature supporting phytochemical-based surface modification as an alternative strategy in implant engineering [14, 31, 44], while the sections that follow distinguish what the data demonstrate from what remains to be tested.

Surface characterization

The physicochemical characterization data collectively confirm the deposition and chemical integrity of curcumin on the anodized titanium substrate. FESEM demonstrated a continuous, conformal coating that followed the micro-roughened topography of the etched and anodized substrate; nanoscale surface features of this kind are known to promote protein adsorption, focal adhesion formation, and subsequent osteoblastic cell attachment [24, 25], although the nanotubular architecture itself was not resolved at the magnification used. In contrast to HA coatings, which are susceptible to microcracking and delamination under mechanical stress [6], the curcumin layer appeared continuous at the micrometer scale; however, adhesion strength was not measured, and this observation should not be read as evidence of mechanical durability.

XRD revealed an important characteristic of the curcumin layer: no distinct crystalline curcumin peaks were detected, and the uniform low-angle background across 2θ = 10° to 30° indicated that curcumin is present in an amorphous state, or below the detection limit of XRD, on the titanium surface. Although amorphous materials lack long-range crystalline order, the amorphous state of curcumin may confer an advantage in this context by enabling more rapid dissolution and local availability at the implant-tissue interface during the early healing period, when osteogenic and anti-inflammatory signaling is most critical [45]; this potential advantage was not tested here, because dissolution and release were not measured. FTIR confirmed the preservation of the essential functional groups of curcumin, including the conjugated β-diketone moiety and phenolic hydroxyl groups, which are critical for its antioxidant and anti-inflammatory properties [10]. Complementary Raman spectroscopy provided molecular-level confirmation and indicated a broad, though microscopically heterogeneous, distribution of curcumin across the titanium surface, offering chemical evidence beyond morphological analysis alone. None of these techniques quantifies the amount of curcumin deposited, and the nanoscale morphology of the deposit was not resolved; for these reasons the coating is described in this manuscript as a curcumin coating rather than a nanoparticle coating (Sect.  4.5).

Biological performance and mechanistic considerations

The higher cytocompatibility observed on Cur-Ti surfaces is consistent with the established role of curcumin as a modulator of cellular redox balance and inflammatory signaling. L929 fibroblast viability exceeded 96% at both time points on Cur-Ti, significantly surpassing HA-Ti and far exceeding the 70% non-cytotoxic threshold of ISO 10993-5:2009. This favorable profile was confirmed in the osteoblast-like MG-63 cell line, where Cur-Ti yielded viability of 94.8 ± 1.1% and 96.5 ± 2.1% at 24 and 48 h, respectively, both significantly exceeding the HA-Ti values (p = 8.1 × 10^− 7 and p = 0.001). This consistency across two cell lines strengthens the evidence for cytocompatibility beyond the standard ISO cell line; nonetheless, because MTT measures metabolic activity rather than cell number and curcumin can interfere with tetrazolium chemistry [42], the absolute magnitude of the differences should be confirmed with a label-independent endpoint. The capacity of curcumin to scavenge reactive oxygen species and attenuate oxidative stress [14] offers a plausible explanation for the favorable response, but reactive oxygen species and inflammatory mediators were not measured in this study, so a redox-mediated mechanism remains a hypothesis rather than a demonstrated effect.

The elevated total ALP activity in the Cur-Ti group is consistent with enhanced early osteogenic activity in MG-63 cells, with the normalization caveat discussed in Sect.  4.5. At day 7, Cur-Ti surfaces yielded ALP activity 24.1% greater than HA-Ti (164.62 ± 3.71 versus 132.66 ± 2.16 U/L), and this advantage was sustained at day 14, with Cur-Ti exceeding HA-Ti by 28.8% (229.82 ± 2.51 versus 178.48 ± 1.22 U/L; p < 0.0001 at both time points). Both coated groups showed progressive ALP upregulation between day 7 and day 14, consistent with advancing osteogenic differentiation. This finding is mechanistically plausible in light of the demonstrated ability of curcumin to upregulate RUNX2, the master transcription factor that governs osteoblast differentiation [15, 16]. Curcumin promotes osteogenesis through several interconnected molecular pathways: it inhibits the NF-κB signaling cascade, thereby relieving the suppression of osteoblast differentiation by pro-inflammatory cytokines such as TNF-α and IL-1β [13]; it activates BMP-2/Smad 1/5/9 signaling [16]; and it protects osteoblasts from oxidative stress-induced dysfunction via the GSK3β-Nrf2 pathway [14]. In the context of bone pathology, curcumin has also been shown to restore Wnt/β-catenin signaling by inhibiting GSK3β-mediated phosphorylation and degradation of β-catenin [18, 19]. Collectively, these mechanisms provide a plausible molecular basis for the osteogenic response observed on curcumin-coated surfaces, although none of them was tested directly in the present coating system.

The osteogenic effects of curcumin are dose-dependent: an optimal concentration range of 5 to 10 µM has been reported for ALP upregulation and osteoblast differentiation, whereas concentrations exceeding 20 to 25 µM are cytotoxic and can induce apoptosis [19, 46]. Critically, the curcumin loading, the release kinetics, and the resulting local concentration were not measured in the present study. Whether the coating achieves and sustains a curcumin concentration within this therapeutic window therefore remains an untested assumption rather than a demonstrated property of the system; accordingly, the release-dependent mechanism proposed here should be regarded as a working hypothesis to be confirmed by direct loading and release measurements (for example, UV-Vis or HPLC quantification over 7 to 14 days under physiological conditions).

Relation to previous curcumin-functionalized titanium systems

The findings of this study complement and extend the work of Sarkar and Bose [31], who demonstrated the controlled release of curcumin and vitamin K2 from plasma-sprayed HA-coated titanium implants over 11 days, with enhanced proliferation of human fetal osteoblasts, as validated in a rat distal femur model. Kushram et al. [32] subsequently showed that HA coatings co-functionalized with curcumin and epigallocatechin gallate achieved approximately a 94% reduction in bacterial growth while maintaining osteogenic potential, and Majumdar and Bose [33] reported a coating adhesion strength of 22 ± 2 MPa for a curcumin and vitamin C dual-release HA coating on Ti-6Al-4 V. In these systems the ceramic layer is the drug reservoir, whereas in the present study curcumin was applied directly to the anodized titanium surface and was compared against an HA coating rather than combined with it. The study by Peng et al. [37], which loaded curcumin onto anodized TiO₂ nanotubes on commercially pure titanium and reported improved bone marrow mesenchymal stem cell adhesion, proliferation, and osteogenic differentiation together with drug release, wettability, and corrosion data, is the closest published comparator. The present work differs in the substrate (Ti-6Al-4 V), in the use of a sol-gel HA benchmark on the identical substrate, and in the ISO 10993-5 dual-cell-line cytocompatibility design, but it does not include the release, wettability, or corrosion measurements reported there. A frequently cited advantage of curcumin coatings over plain HA is intrinsic antimicrobial activity: Alonso-Español et al. [47] demonstrated a significant reduction of all six species in a dynamic multi-species dental biofilm model developed over implant surfaces, including the periodontal pathogens Porphyromonas gingivalis and Aggregatibacter actinomycetemcomitans, and Saha et al. [38] reported reductions of up to 43% and 38% in E. coli and S. aureus populations on drop-cast curcumin on anodized Ti-6Al-4 V. Although antimicrobial activity was not evaluated in the present study, this literature suggests that curcumin functionalization could, in principle, address two of the most critical challenges in implant dentistry, osseointegration and peri-implant infection, within a single coating platform, a possibility that warrants direct testing in future work.

Demonstrated versus proposed functionalities in the context of multifunctional coatings

Contemporary biomedical coatings are increasingly designed to be multifunctional. Yigit et al. [48] summarize the functions that such coatings are expected to deliver simultaneously: biocompatibility, resistance to infection, mechanical durability, corrosion protection, and drug-delivery capability. Recent examples illustrate the quantitative benchmarks now reported for such systems. Akinay et al. [49] showed that modification of Ti₃C₂T_x MXene films with polyhedral oligomeric silsesquioxane reduced the water contact angle from 85.92° to 72.23° and produced antibacterial inhibition of 65.93% against Escherichia coli and 80.63% against Staphylococcus aureus after 4 h. Karatas et al. [50] reported that an alginate-CuO-MXene coating reduced the corrosion current density of an AZ31 magnesium alloy from 27.6 to 6.34 µA cm^− 2, a 77% reduction, while improving antibacterial behavior and cytocompatibility. In the curcumin field, Sun et al. [51] combined curcumin with cyclodextrin and zinc-doped TiO₂ nanorods on titanium and reported hydrophilicity, corrosion resistance, antibacterial, anti-inflammatory, angiogenic, and osteogenic outcomes in vitro and in vivo.

Against this benchmark, the present study makes deliberately narrower claims. Table 4 distinguishes the functions quantitatively demonstrated here from those not tested. Cytocompatibility was demonstrated in two cell lines with percentage differences, confidence intervals, and exact p-values (Table 3), and early osteogenic activity was demonstrated at the level of total ALP activity (70% to 115% above uncoated titanium and 24% to 29% above HA-Ti), with the per-cell normalization caveat noted above. Wettability was not measured, so no comparison with the contact-angle values reported for MXene films [49] or curcumin-loaded nanotubes [37] is possible, and the direction of any change in contact angle produced by the curcumin layer on the anodized surface cannot be inferred from the present data. Antibacterial activity was not evaluated because the study was designed as a first-stage cytocompatibility and osteogenic screen; the antibacterial properties of curcumin reported by others [32, 38, 47] are therefore cited as rationale for future work, not as a property of Cur-Ti. Corrosion protection is neither claimed nor inferred: Ti-6Al-4 V forms a stable passive oxide layer and does not present the corrosion problem that motivates barrier coatings on magnesium alloys [50], and any barrier effect of a thin organic curcumin layer would need to be established by potentiodynamic polarization or electrochemical impedance spectroscopy rather than assumed. Likewise, curcumin loading, release kinetics, coating thickness, adhesion strength, and long-term stability in physiological media were not measured, and the drug-delivery function of the coating remains proposed rather than demonstrated. The comparative statements in this manuscript are accordingly restricted to cytocompatibility and total ALP activity relative to HA-Ti and uncoated titanium.

Table 4.

Functions expected of multifunctional implant coatings [48] and their status in the present study

Function	Status in the present study	Evidence in this manuscript	Measurement required to substantiate
Cytocompatibility	Demonstrated (in vitro)	MTT assay, L929 and MG-63 cells, 24 and 48 h, n = 6 per group (Tables 1 and 3)	Label-independent viability or proliferation assay; cell-free disc controls; uncoated Ti group in the MTT arm
Osteogenic activity	Demonstrated as total (well-level) ALP activity only	ALP at day 7 and day 14, n = 6 per group (Tables 2 and 3)	ALP normalized to DNA or protein; RUNX2 and osteocalcin expression; mineralization
Drug delivery (curcumin loading and release)	Not quantified; proposed	Presence and chemical integrity of curcumin by FESEM, XRD, FTIR, Raman, and UV-Vis	Loading per unit area; cumulative release (%) versus time by UV-Vis or HPLC
Infection resistance (antibacterial activity)	Not evaluated; not claimed	None	Antibacterial assays against S. aureus, E. coli, and oral pathogens
Mechanical durability (adhesion, wear)	Not evaluated; not claimed	Absence of visible cracking or delamination by FESEM (qualitative)	Pull-off or scratch adhesion testing; nanoindentation
Corrosion protection	Not evaluated; not claimed	None	Potentiodynamic polarization and electrochemical impedance spectroscopy in simulated body fluid
Wettability	Not measured	None	Water contact angle
Long-term coating stability and sterilization tolerance	Not evaluated	None	Immersion stability in PBS or medium over 7 to 28 days; effect of sterilization

Limitations and future directions

Several limitations of this study should be acknowledged. First, MG-63 cells, though widely used as an osteoblast model and phenotypically stable across passages [52], are derived from osteosarcoma and differ from primary osteoblasts in miRNA expression profiles and mineralization capacity [53, 54]. Future research should use primary human bone marrow-derived mesenchymal stem cells or hFOB 1.19 cells to improve translational relevance.

Second, the curcumin coating was not characterized quantitatively. Coating thickness, particle size distribution, polydispersity, zeta potential, quantitative surface coverage, and the mass of curcumin deposited per disc were not measured, and the single FESEM micrograph (4000×) documents coating continuity but does not resolve nanoscale morphology. For this reason, the coating is described in this manuscript as a curcumin coating rather than a nanoparticle coating, and the original designation CurNP-Ti has been withdrawn. Cross-sectional FESEM or profilometry for thickness, high-magnification FESEM or atomic force microscopy for morphology, and spectrophotometric quantification of loading (curcumin extracted from coated discs) are recommended to establish quantitative relationships between coating characteristics and biological performance.

Third, the study did not assess curcumin release. Because curcumin is unstable at physiological pH, with about 90% degradation within 30 min in phosphate buffer at pH 7.2 [55], and encapsulation or immobilization has been shown to enhance its stability by 300 to 500 times [56], the curcumin concentration actually experienced by the cells over 24 h to 14 days is unknown. Cumulative release (%) versus time in phosphate-buffered saline and in culture medium at 37 °C, measured by UV-Vis spectrophotometry or HPLC over 7 to 14 days, together with the resulting local concentration relative to the 5 to 10 µM osteogenic window and the 20 to 25 µM cytotoxicity threshold [19, 46], is required to support the release-dependent mechanism proposed in Sect.  4.2.

Fourth, cell viability was assessed solely through the MTT metabolic activity assay, which measures mitochondrial dehydrogenase activity as a proxy for cell viability but does not directly quantify cell number or proliferative capacity. Curcumin is redox-active and a recognized interferent in tetrazolium-based assays [42], and neither cell-free disc controls (discs incubated with MTT reagent in medium without cells) nor an uncoated titanium disc group were included in the MTT arm, so part of the apparent viability advantage of Cur-Ti could reflect a curcumin-assay interaction rather than a true increase in cell survival. The non-cytotoxic classification of HA-Ti is unaffected by this caveat, whereas the Cur-Ti values should be regarded as an upper bound until confirmed. A complementary proliferation assay (BrdU incorporation, Ki-67 immunostaining, or DNA quantification) or an independent viability assay (live/dead fluorescence staining or CCK-8), together with cell-free disc controls and an uncoated titanium group, is recommended in future studies.

Fifth, ALP activity was expressed per unit volume (U/L) and was not normalized to total protein, DNA content, or cell number, which were not measured in parallel. Because the Cur-Ti surfaces also supported higher cell viability in the MTT assay at 24 and 48 h, the elevated total ALP activity at day 7 and day 14 may partly reflect a greater number of adherent cells rather than an increase in ALP expression per cell. Although the relative ALP difference between groups (24% to 29% versus HA-Ti) exceeded the corresponding cell-viability difference (11% to 13%), this comparison is indirect because cell number was not measured at day 7 or day 14, and MTT measurements at 24 to 48 h cannot retrospectively normalize ALP measured at days 7 and 14. A definitive demonstration of enhanced per-cell osteogenic differentiation requires normalization of ALP to total protein (for example, the bicinchoninic acid assay) or DNA (for example, the PicoGreen assay), or parallel DNA or protein quantification, ideally complemented by RUNX2 and osteocalcin expression and alizarin red mineralization endpoints; the present findings should therefore be read as an early osteogenic signal at the well level.

Sixth, antibacterial activity, wettability, coating adhesion strength, corrosion behavior, and long-term coating stability were not evaluated (Table 4). These functions are neither claimed nor implied for Cur-Ti; the comparison with multifunctional coating systems in Sect.  4.4 serves only to position the present findings and to define the measurements required before such claims could be made. Sterilization methods must also be suitable for clinical use; gamma irradiation (25 kGy) is preferred over autoclaving, which causes significant thermal degradation of curcumin.

Seventh, although the two-way ANOVA with Tukey HSD post-hoc test provides appropriate control of type I error for the multi-group ALP comparisons, the modest sample size (n = 6 discs per group), the single-laboratory design, and the use of a single coating batch prepared in one session mean that between-batch reproducibility of the coating and of the biological response was not assessed; confirmation in larger, independently replicated studies with multiple coating batches is needed before firm conclusions regarding the magnitude of the osteogenic advantage can be drawn. Eighth, in vivo testing in an appropriate animal model is essential for assessing long-term coating stability, osseointegration, and potential immune responses under physiological loading conditions.

Conclusion

This study demonstrates that a carrier-free curcumin coating applied by dip-coating to acid-etched and anodized Ti-6Al-4 V provides higher in vitro cytocompatibility in L929 and MG-63 cells and higher total early osteogenic (ALP) activity in MG-63 cells than a conventional sol-gel hydroxyapatite coating on the same substrate, with ALP activity 70% to 115% above uncoated titanium and 24% to 29% above HA-Ti. Physicochemical characterization confirmed the deposition, chemical integrity, and distribution of an amorphous curcumin layer with continuous, conformal coverage at the micrometer scale. The evidence presented is limited to cytocompatibility and well-level ALP activity: curcumin loading and release, per-cell osteogenic differentiation, antibacterial activity, wettability, coating adhesion, corrosion behavior, and long-term stability were not measured and are not claimed. Together with the antioxidant and antimicrobial properties reported for curcumin in other systems, these in vitro findings identify curcumin-based surface functionalization of anodized titanium as a candidate strategy for dental and orthopedic implant coatings whose value will depend on direct quantification of loading and release, normalized osteogenic endpoints, antimicrobial testing, primary osteoblast culture, and in vivo validation before clinical translation.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations:

ALP

alkaline phosphatase

ANOVA

analysis of variance

BMP-2

bone morphogenetic protein-2

CI

confidence interval

Cur-Ti

curcumin-coated anodized titanium

DMEM

Dulbecco's modified Eagle medium

DMSO

dimethyl sulfoxide

FBS

fetal bovine serum

FESEM

field-emission scanning electron microscopy

FTIR

Fourier-transform infrared spectroscopy

HA

hydroxyapatite

HA-Ti

hydroxyapatite-coated anodized titanium

HSD

honestly significant difference

MTT

3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide

NF-κB

nuclear factor kappa-light-chain-enhancer of activated B cells

pNPP

p-nitrophenyl phosphate

RUNX2

runt-related transcription factor 2

SD

standard deviation

SE

standard error

Ti

titanium

TiO2

titanium dioxide

UV-Vis

ultraviolet-visible spectroscopy

XRD

X-ray diffraction

Author contributions

Nandita Suresh: Conceptualization, Investigation, Methodology, Formal Analysis, Writing—Original Draft, Writing—Review & Editing. Nebu G. Thomas: Investigation, Resources. Matti Mauramo: Investigation, Validation. Tuomas Waltimo: Supervision, Resources. Timo Sorsa: Supervision, Writing—Review & Editing. Sukumaran Anil: Conceptualization, Supervision, Writing—Review & Editing, Project Administration.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. Open access funded by Helsinki University Library.

Data availability

The data sets generated and analyzed during the current study are available on Figshare at https://doi.org/10.6084/m9.figshare.33089459 . The per-well values underlying Tables 1 and 2 are also provided as Supplementary Tables S1 and S2 in the Supplementary Information file that accompanies this article.

Declarations

Ethical approval

This study used only commercially available, established cell lines (L929, ATCC CCL-1; MG-63, ATCC CRL-1427) obtained from the National Center for Cell Science, Pune, India. No human participants, human tissues or data, or animals were involved; therefore, ethical approval was not required.

Consent to participate

Not applicable

Competing interests

The authors declare no competing interests.