The raw materials used in this experiment are shown in Figure 1 . To prepare the non-cement composite, the experiments used RCP, wood sawdust, and oyster shell, which were procured from local sources and oven dried for 24 hours at 120˚C. The dried RCP was sieved through a 0.3 mm sieve, but the wood powder and oyster shell powder were sieved through a 0.6 mm sieve and stored in sealed bags before the experiment.

The experimental plan is shown in Table 1 . It consists of three cases. In case 1, in a plastic container, unground RCP and oyster shell were mixed for 1 minute under dry conditions, and then, the water was added to the mixture and mixed for one more minute manually. The mixture was subjected to pressure at 30 MPa for 10 minutes using an AS ONE H400-15 hot-pressing machine and subsequently demolded, which yielded a hardened body, which was labeled the RCP-OS-UG sample. The molding procedure and apparatus are shown in Figure 2 . It is mentioned that this experiment used a cold press, and no heat was applied.

W: wood powder; OS: osyter shells.

For cases 2 and 3, a mixer mill (MM400, Retsch) was used to grind the raw materials using 50 ml jar and 25 mm ball diameter with a 25 Hz horizontal oscillation frequency.

In case 2, the RCP and oyster shell samples were separately ground for five minutes. The ground materials were poured into a plastic container and mixed for one minute. Then, water was added to the mixture, which was mixed for one more minute. Finally, the mixture was molded under the same conditions as those in case 1, and the obtained sample was marked as RCP-OS-SG.

In case 3, first, RCP and oyster shells were ground separately for five minutes, resulting in fine powders. The powders were subsequently mixed and ground for 2 minutes via the same ball mill machine. This is called combined ground. Next, water was added to the mixture, which was mixed for one minute in a plastic container. Finally, the molding activity was performed under the same conditions as mentioned above.

Similarly, the above procedures were applied to RCP and wood for the three cases, in which only cases 2 and 3 produced hardened bodies. However, in case 1, the mixture could not harden, and thus, no sample was obtained.

Notably, the unground, separate ground and combined ground average particle sizes were measured via SYNC and a NANOTRAC WAVE II Microtrac particle size analyzer. For the particle size measurement of case 3, individual raw materials were ground in two steps, first for five minutes and then for two minutes without mixing, to measure the individual material particle size instead of the mixture size.

All the samples were cured at room temperature for 72 hours before conducting a flexural test.

After obtaining the samples and conducting the flexural strength test the samples were characterized using the following analytical methods to illustrate the microstructure and the possible chemical alteration during grinding.

Mineral composition analysis was conducted via an XRD diffractometer (D8 DISCOVER, Bruker) operating at 40 kV and 40 mA with Cu-Kα radiation. The scanning step size was set to 0.02˚ in the 2θ range of 10˚ to 70˚. The XRD was employed to examine the mineral composition of the samples prepared from unground, separate ground and combined ground samples.

FTIR spectra were acquired using a PerkinElmer Spectrum 3 instrument. The spectra were recorded over a range of 4000 - 400 cm⁻¹ with a resolution of 4 cm⁻¹. The data of the samples containing RCP were normalized to the peak at 874 out-of-plane vibrations of C-O in the pure RCP spectra.

The microstructures of the composite samples were analyzed by SEM‒EDS via a FlexSEM1000II microscope to 1) Characterize the structural features of the composite; 2) Determine the elemental composition and distribution of the samples through elemental mapping; 3) Assess the morphological characteristics of the samples.

The particle size of the particulates is shown in Figure 2 . The unground average particle sizes of RCP, oyster shell, and wood are 215, 340, and 368 µm, respectively. After separate grinding, the size decreased to 92, 54, and 117 µm, and after combined grinding, the size further decreased to 52, 39, and 72 µm. The effects of particle size and grinding method on the flexural strength of oyster shell-based samples and wood-based samples are shown in Figure 3 . As shown in Figure 3 , the composite made of unground RCP and wood powder did not harden, but the composite made of unground RCP and oyster shell powder hardened and had a 0.9 MPa flexural strength. In case 2, when the materials were separately ground for five minutes and their particle sizes decreased, both the oyster-based and the wood-based composites hardened and demonstrated flexural strengths of 1.2 MPa and 0.8 MPa, respectively, confirming the findings of previous studies that indicate an inverse relationship between the size of oyster shell particles and concrete strength [24] [25] . In case 3, when combined ground was applied, the composite flexural strength increased to 4.3 MPa and 1.6 MPa in both oyster shell and wood-based composites, showing 258% and 100% enhancement, respectively. According to Erdugdo et al. [26] , co-gripping leads to strength improvement, which is associated with a reduction in the particle size and homogeneity of the mixture. This result is also consistent with the findings of Onaizi et al. [27] and Dvořák et al. [28] , who reported that co-ground samples presented a greater increase in strength at an early age than did those containing the equivalent ratios of mixtures ground separately.

The X-ray diffraction (XRD) patterns of the unground, separately ground, combined ground samples, and RCP, oyster shell, and wood are shown in Figure 4 and Figure 5 . In RCP, the XRD peaks mostly indicate the presence of CaCO₃ and SiO₂. This also indicates the availability of Ca(OH)₂. In oyster shells, since oysters abundantly contain CaCO₃, the peaks are solely indicative of the existence of CaCO₃. The XRD patterns of the unground, separately ground, and combined ground samples display overlapping features, with no significant deviations between the individual patterns. This finding indicates that the grinding process does not substantially alter the material phases, which aligns with the findings of Onaizi et al. [27] , who reported that co-grinding does not lead to structural transformations in the powder or induce the formation of new phases.

The FTIR spectra in the frequency range of 4000 - 400 cm⁻¹ for the materials and the composite samples from the unground, separate ground, and combined ground samples, and RCP, oyster shell, and wood are shown in Figure 6 and Figure 7 . For both the RCP-oyster shells and RCP-wood samples, the figures indicate that the major peaks for all the samples (unground, separate ground and co-ground) appear at approximately 3438, 2514, 1798, 1420, 1011, 874, and 713 cm⁻¹.

As shown in Table 2 , the peak at 3438 cm⁻¹, which is related to O-H bending and vibrations, arises from water adsorbed on the surface of the samples and the hydration products of C₃S and C₂S [29] [30] . The peaks at 2514 and 1798 cm⁻¹ correspond to CaCO₃ [31] [32] . The peak at 1420 cm⁻¹ indicates the asymmetric C-O stretching of carbonate molecules, which results from calcium hydroxide carbonation [30] [33] . The peak at 1011 cm⁻¹ corresponds to Si-O asymmetric stretching and is due to the formation of hydraulic compounds such as C-S-H [30] [33] . The peak at 874 cm⁻¹ shows the out-of-plane vibration of C-O in CaCO₃ [33] [34] . The peak at 713 cm⁻¹ indicates the in-plane vibration of the C-O of CaCO₃ [33] .

No new peak appeared or disappeared or shifted in any of the three cases, which is consistent with the XRD results and showed that there is no chemical alteration or new phase formation when the raw materials are ground separately or in combination.

Figure 8 and Figure 9 show SEM images of RCP-oyster shells and RCP-wood samples made of the unground, separate ground and combined ground raw materials. As shown in the figures, the size of the particles and voids decreased from the unground samples to the combined ground samples. RCP-OS-CG shows significantly denser microstructures. In contrast, the samples in RCP-OS-SG, which were prepared with separately ground RCP and oyster shell mixtures, were less homogeneous and had more porous microstructures. Compared with the RCP-OS-SG samples, the homogeneous and compacted microstructure of the RCP-OS-CG case considerably contributed to its better flexural strength. Similarly, unlike the RCP-OS-SG sample, the RCP-OS-CG samples exhibited a more uniform dispersion of particles. This may be due to the synergistic effect and fineness achieved by the combined grinding of the RCP and oyster shell or wood combination. The same characterization can be seen in the cases of RCP-W-UG, RCP-W-SG, and RCP-W-CG. The results are in accordance with the results of Onaizi et al. [27] , who reported that the co-ground effectively fills voids and transition zones between particles.

Table 3 shows the EDS data of the RCP-OS-UG, RCP-OS-SG, and RCP-OS-CG samples. The elemental compositions were obtained via SEM‒EDS by mapping analysis of the surface of the samples. The major elements in the samples were O, C, and Ca, which indicate the abundance of CaCO₃ in the samples. The O and Ca contents decreased from the unground samples to the combined ground one. However, the C, Si, Al, and Na contents increased from the unground

samples towards the combined ground samples. The difference in percentage increase was significant between the unground and separate ground samples but was not significant between the separate ground and combined ground samples. Considering the element ratios, the Ca/Si and Ca/Na ratios decreased after grinding. However, the Si/Al ratio increased. Nukah et al. studied the reduction in calcium in ground granular blast furnace slag (GGBS), and increasing the amount of silica improved the mechanical strength [36] .

These results are in accordance with those of previous studies. A reduction in the Ca/Si ratio increased the compressive strength of the cementitious calcium-silicate-hydrate mixture [37] . Co-grinding decreased the Ca/Si ratio and increased the strength of the concrete [27] . Dinh et al. studied fly ash-based geopolymer concrete and reported that a Si/Al ratio in the range of 1.5 - 5 increased the compressive strength [38] . Kunther et al. reported that in calcium silicate hydrate binder (C-S-H), the higher the Ca/Si ratio is, the lower the strength, and the lower the Ca/Si ratio is, the higher the strength [37] .