In contemporary dental clinical practice, conventional photo-activation of resin-based composite (RBC) materials involves continuous light exposure for a specified duration, dependent on light irradiance [1]. A standard protocol utilizes a rigid light guide probe of the light curing unit (LCU) strategically positioned directly above the restoration. The goal is to achieve optimal polymerization through sufficient light transmission by maintaining proximity and a perpendicular orientation to the material surface [2]. Meanwhile, modern RBCs commonly referred to as “bulk-fills” have been specifically engineered for deep cavity restoration in a single layer up to 5 mm in thickness [3]. This design capitalizes on heightened translucency [4] and an improved refractive index match between the organic and inorganic phases [5], facilitating ample light transmittance throughout increasing depth. Nevertheless, the challenge arises from the inevitable light loss (up to 80–90 % irradiance reduction) attributed to scattering and attenuation [6]. Consequently, the concept of “curing from within” has been acknowledged as the ultimate, but still unattained objective for dental applications [7].

At the same time, volumetric shrinkage accompanying polymer chain formation leads to undesired polymerization shrinkage stress (PSS), which is considered detrimental to RBC performance [8]. Therefore, various strategies have been proposed to mitigate PSS, including the incremental RBC insertion technique, modified photo-activation protocols (e.g., soft-start or pulse-delay), and modern self-curing bulk-fill RBCs [9]. By placing the material in smaller increments, each up to 2 mm thickness, and decreasing the ratio of bonded to unbonded surfaces (known as C-factor), the constraint imposed on RBC shrinkage is reduced, allowing a certain amount of flow to partially dissipate the PSS [10]. Another modified insertion technique for PSS reduction involves embedding a dental instrument in the middle of the restoration during photo-activation, resulting in improved marginal adaptation by reducing the volume of the RBC that is polymerized at each step and eliminating stress singular point in the RBC [11,12].

On the other hand, alternative photo-activation techniques, collectively termed “soft-start”, have been introduced as beneficial for PSS reduction due to a slow curing process activated by low light irradiance. This allows sufficient time for material viscous flow and stress relaxation without compromising polymerization efficiency, i.e., the degree of monomer-to-polymer conversion (DC) [13], [14], [15], [16]. Among soft-start protocols, the “pulse-delay” technique has been extensively discussed for its potential ease of clinical adoption [16], [17], [18], [19]. Specifically, this technique involves applying a first pulse of light with low irradiance, followed by a delay when the LCU is turned off, and then applying a second pulse of light with high irradiance [20]. However, conflicting claims exist regarding whether the lower PSS observed following soft-start curing techniques is due to a lower resultant DC [21], [22], [23]. Nevertheless, a recent study by Palagummi et al. [20], supported the clinical success of the pulse-delay protocol, attributing PSS reduction primarily to lower thermal shrinkage (contraction associated with cooling during post-cure).

The objective of the present study was to address the challenges of contraction stress and light loss at increasing depths during dental composite polymerization by introducing a novel two-step photo-activation protocol. The study involved the examination of three bulk-fill RBCs using tooth cavity models. The deformation of the cavity models, considered a secondary manifestation of PSS, was monitored non-invasively in real-time using Digital Holographic Interferometry (DHI). Concurrently, Infrared Thermography (IRT) was employed for the real-time measurement of RBC temperature change, serving as an indicator of the polymerization reaction dynamics. Additionally, the DC was determined based on the RBC spectra obtained through Raman spectroscopy at two distinct time points: immediately after photo-activation and after 24 h of dark storage. As a control, conventional photo-activation was performed on a separate group of samples, following the manufacturer's instructions. Four null hypotheses were proposed, assuming that the employed photo-activation protocol would have no impact on: (1) Cavity model deformation, (2) RBC temperature change, (3) Immediate DC, and (4) 24 h post-cure DC.