Hydroxyapatite and Other Calcium Phosphates for the Conservation of Cultural Heritage: A Review

Abstract

The present paper reviews the methods and the performance of in situ formation of calcium phosphates (CaP) for the conservation of materials belonging to cultural heritage. The core idea is to form CaP (ideally hydroxyapatite, HAP, the most stable CaP at pH > 4) by reaction between the substrate and an aqueous solution of a phosphate salt. Initially proposed for the conservation of marble and limestone, the treatment has been explored for a variety of different substrates, including sandstones, sulphated stones, gypsum stuccoes, concrete, wall paintings, archaeological bones and paper. First, the studies aimed at identifying the best treatment conditions (e.g., nature and concentration of the phosphate precursor, solution pH, treatment duration, ionic and organic additions to the phosphate solution, mineralogical composition of the new CaP phases) are summarized. Then, the treatment performance on marble and limestone is reviewed, in terms of protective and consolidating effectiveness, compatibility (aesthetic, microstructural and physical) and durability. Some pilot applications in real case studies are also reported. Recent research aimed at extending the phosphate treatment to other substrates is then illustrated. Finally, the strengths of the phosphate treatment are summarized, in comparison with alternative products, and some aspects needing future research are outlined.

Keywords: ammonium oxalate; ammonium phosphate; calcium phosphates; consolidation; durability; hydroxyapatite; limestone; marble; octacalcium phosphate; protection.

Figure 1

Scheme illustrating in situ formation of hydroxyapatite (HAP), as the reaction product between the substrate and an aqueous solution of a phosphate salt (typically DAP).

Figure 2

Scanning electron microscopy (SEM) images of CaP coatings with different compositions (determined by XRD): (a,b) HAP coating formed over Carrara marble after treatment with a 1 M DAP solution for 24 h; (c,d) HAP + octacalcium phosphate (OCP) coating formed over Carrara marble after treatment with a 1 M DAP + 1 mM CaCl₂ solution for 9 h (c) and 24 h (d) - by comparison with the morphology of the HAP coating illustrated in (a,b), the small flakes in (c) were identified as HAP and the large flakes as OCP [34]; (e,f) OCP coating formed over Carrara marble after treatment with a 0.1 M DAP + 1 mM CaCl₂ in 10 vol % ethanol solution for 24 h. Images (a–d) adapted from reference [34] with permission from Elsevier.

Figure 3

Scheme illustrating the protective action of the phosphate treatment, with an example of a focused ion beam (FIB) SEM cross section of the protective layer formed over marble.

Figure 4

FIB-SEM cross sections of coatings formed by reacting marble for 24 h with solutions containing (a) 1 M DAP + 1 mM CaCl₂ and (b) 0.1 M DAP + 0.1 mM CaCl₂ in 10 vol % ethanol. The much denser coating in (b) was formed thanks to the beneficial effect of ethanol, as described in detail in Section 2.7. Images adapted from reference [38] with permission.

Figure 5

Scheme illustrating the consolidating action of the phosphate treatment, with an example of a SEM image showing HAP growth inside a crack in weathered marble.

Figure 6

Scheme illustrating nanoTiO₂ removal by rain when nanoparticles are directly applied onto marble surface. A significant improvement in durability can be obtained by incorporating nanoTiO₂ in HAP coatings (on the right, a FIB-SEM cross section of a HAP-TiO₂ nanocomposite is illustrated).

Figure 7

Scheme illustrating the consolidating action of the phosphate treatment, with an example of a SEM image showing newly formed HAP bonding calcite grains in limestone.

Figure 8

Scheme illustrating the possible use of HAP as a coupling agent for silicate consolidants applied to limestones. However, the bonding between the silicate consolidant and the HAP layer was found to be mechanical in nature.

Figure 9

Alterations in pore size distribution of Globigerina limestone treated with a 3 M DAP solution, followed by a limewater poultice (“HAP,” left) and with a commercial ethyl silicate (“ES,” right), in comparison with the untreated reference (“UNTR”). Both consolidants were applied by brushing 10 times cylindrical samples (50 mm height, 20 mm diameter) on their lateral surface. The pore size distribution was determined by mercury intrusion porosimetry (MIP) on samples obtained at increasing depth from the treated surface. The “ES” sample was left to cure for 1 month before testing. Image adapted from Reference [49] with permission from Elsevier.

Figure 10

Alterations in water sorptivity of Globigerina limestone treated with a 3 M DAP solution, followed by a limewater poultice (“HAP”) and with a commercial ethyl silicate (“ES”), in comparison with the untreated reference (“UNTR”). Both consolidants were applied by brushing 10 times cubic samples (50 mm side) on a single face, through which water was then let penetrate by capillarity. The “ES” sample was left to cure for 1 month before testing. Image reprinted from Reference [49] with permission from Elsevier.

Figure 11

Variations in weight (left) and E_d (right) of Globigerina limestone subjected to freeze-thaw cycles (FT), untreated (“UNTR”) and treated with a 3 M DAP solution followed by a limewater poultice (“HAP”) and with a commercial ethyl silicate (“ES”). Before testing, the “ES” samples were left to cure for 1 month, then the residual hydrophobicity was eliminated by application of a water poultice [87]. Image reprinted from Reference [50] with permission from Elsevier.

Figure 12

Variations in weight (left) and E_d (right) of Globigerina limestone subjected to salt crystallization cycles (SW), untreated (“UNTR”) and treated with a 3 M DAP solution followed by a limewater poultice (“HAP”) and with a commercial ethyl silicate (“ES”). Before testing, the “ES” samples were left to cure for 1 month, then the residual hydrophobicity was eliminated by application of a water poultice [87]. Open symbols indicate values of samples desalinated after the 5th cycle. Image reprinted from Reference [50] with permission from Elsevier.

Figure 13

Scheme illustrating the de-sulphating action expected from the phosphate treatment.