Materials Science and Engineering C 33 (2013) 527–531
Contents lists available at SciVerse ScienceDirect
Materials Science and Engineering C journal homepage: www.elsevier.com/locate/msec
Influence of water content on hardening and handling of a premixed calcium phosphate cement Johanna Engstrand ⁎, Jonas Aberg, Håkan Engqvist Applied Materials Science, Department of Engineering Sciences, Uppsala University, Sweden
a r t i c l e
i n f o
Article history: Received 24 January 2012 Received in revised form 12 September 2012 Accepted 28 September 2012 Available online 6 October 2012 Keywords: Premixed Calcium phosphate cements Injectability Strength Setting time Water content
a b s t r a c t Handling of calcium phosphate cements is difficult, where problems often arise during mixing, transferring to syringes, and subsequent injection. Via the use of premixed cements the risk of handling complications is reduced. However, for premixed cements to work in a clinical situation the setting time needs to be improved. The objective of this study is to investigate the influence of the addition of water on the properties of premixed cement. Monetite-forming premixed cements with small amounts of added water (less than 6.8 wt.%) were prepared and the influence on injectability, working time, setting time and mechanical strength was evaluated. The results showed that the addition of small amounts of water had significant influence on the properties of the premixed cement. With the addition of just 1.7 wt.% water, the force needed to extrude the cement from a syringe was reduced from 107 (± 15) N to 39 (± 9) N, the compression strength was almost doubled, and the setting time decreased from 29 (±4) min to 19 (± 2) min, while the working time remained 5 to 6 h. This study demonstrates the importance of controlling the water content in premixed cement pastes and how water can be used to improve the properties of premixed cements. © 2012 Elsevier B.V. All rights reserved.
1. Introduction Calcium phosphate based materials are considered ideal for bone replacement, since they resemble the mineral phase of bone; calcium deficient hydroxyapatite. The first calcium phosphate paste was patented in 1975 by Driskell et al. [1], and the first calcium phosphate cements (CPC) were reported in the early 1980's [2,3]. Brushite and apatite are the two main types of CPC found, differing mainly in pH. In an acidic environment (pH b ~4.2) the cement precipitates to brushite [4,5]; although monetite has also been reported [6]. At a higher pH however, apatite precipitation occurs [7]. Both types of CPC have good biocompatibility and are used clinically [8–11]. Furthermore, promising in vivo results have been shown for monetite based materials [12]. When the cements are compared in vivo, the acidic cements have been shown to resorb faster than apatite cements due to their increased solubility in physiological pH [13]. This allows for faster bone in-growth to occur into the cement-filled defect [12]. The mixing of commercially available water-based CPCs at present has to be done in the operating room, since hardening starts immediately after mixing. Each product has specific mixing equipment and instructions that can be quite tedious, making them difficult to work with. All stages of mixing, transfer, and injection are time sensitive, and must be performed within approximately 10 min. If the preparation of the cement goes over the allotted time, the cement hardens and injection of the cement is no longer
possible. To overcome the mentioned problems, premixed calcium phosphate cements (pCPC) have been developed over the last few years. A pCPC is based on calcium phosphate powders mixed with a non-aqueous water-soluble liquid; e.g. glycerol or poly(ethylene glycol) [14–16]. Since the setting of these cements commences when the liquid phase is exchanged with water or body fluid, these cements could be stored and delivered in a syringe. This eliminates the stress related to the preparation of conventional water-mixed CPC's in the operating room. Both apatite [17,18] and acidic [15,16] pCPCs have been reported. In vivo evaluation of pCPCs has demonstrated similar behavior as conventionally water-mixed cements [19,20], making them a viable alternative. Although the water-mixed acidic cements set within a few minutes [15], due to the slow water–glycerol exchange the setting of the acidic pCPC is more time consuming [15,16]. To make the pCPC interesting commercially, the setting time needs to be reduced. It is probable that addition of small amounts of water to the cement mixture could reduce the setting time but it is not known how much is needed in order to achieve clinically relevant times. The objective of this study was to investigate the influence of small amounts of water on injectability, setting time, working time and mechanical properties of an acidic premixed cement mixture. 2. Method 2.1. Cement preparation
⁎ Corresponding author at: Ångströmlab, Lägerhyddsvägen 1, Box 534, 751 21 Uppsala, Sweden. Tel.: +46 18 471 79 46; fax: +46 18 471 35 72. E-mail address: [email protected] (J. Engstrand). 0928-4931/$ – see front matter © 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.msec.2012.09.026
The powders used in the study were β-tricalcium phosphate (β-TCP, Fluka) and monocalcium phosphate hydrate (MCPH, Alfa
528
J. Engstrand et al. / Materials Science and Engineering C 33 (2013) 527–531
Aesar). The MCPH was dried in 110 °C for 3 days in order to remove all water, resulting in monocalcium phosphate anhydrous (MCPA) [21]. MCPA was used to facilitate the control of water content in the cement mixture. β-TCP and MCPA were stored in a desiccator over silica gel. For all cements the molar ratio used for MCPA and β-TCP was 1:1. Glycerol (anhydrous, Sigma) was used as a mixing liquid. Water was deliberately added to the mixture in percent of the total weight; 0, 1.7, 3.4, 5.1, and 6.8 wt.%, and four batches were made per group. The powder to glycerol (P/G) ratio was kept at 4.0 g/ml for all cements. Therefore, the total powder to liquid ratio (P/L) was lower for cements containing more water. A vacuum mixer (Twister, Renfert) was used for all mixing. First water was added to glycerol and hand mixed for some seconds before MCPA was added and all components were mixed for 50 s using the vacuum mixer. β-TCP powder was added and everything was mixed twice for 50 s. Between each mixing the cement attached to the wall and paddle was loosened in order to obtain homogenous cement pastes.
polished to make the sample sides parallel and obtain uniform height. The maximum compressive stress until failure was measured using a universal testing machine (Shimadzu AGS-H), with a cross-head speed of 1 mm/min. A thin plastic film was placed between the sample and the cross-head in order to reduce the effect of potential surface defects. A total of 12 samples were made for each group.
2.6. SEM To study the pore structure of the set cements, the cross-sections were investigated with scanning electron microscopy (SEM, LEO 1550, Zeiss). The samples were polished using 1200 grit SiC paper. A thin gold/palladium coating was sputtered onto the surface before analysis to avoid charging of the surface.
2.7. X-ray diffraction (XRD) 2.2. Working time Working time of the cements was evaluated by preparing two batches with 0, 1.7 and 3.4 wt.% water each. Cements with higher than 3.4 wt.% water content could not be injected after 15 min and were, therefore, not evaluated further. The P/G for 1.7 and 3.4 wt.% was 4.0, while the P/G for 0 wt.% was slightly lower at 3.75, in order to start with comparable extrusion forces at t = 0. Syringes with a barrel diameter of 8.55 mm and an outlet diameter of 1.90 mm were filled with approximately 1 ml of paste at t = 0 and the syringes were stored in a dry environment at room temperature until they were tested. The force needed to extrude the paste from the syringe at a cross-head speed of 60 mm/min was measured continually until the cement was too viscous to be extruded from the syringe, i.e. the piston folded at F = 240 N. The time between measuring points was different for varying water contents in order to get approximately the same amount of measuring points for all cements. 2.3. Injectability
The resulting phase composition of the cements after setting was analyzed using XRD (diffractometer, Siemens). Diffraction angles (2θ) 20–40 were analyzed at 0.45°/min. The set samples were crushed using a mortar prior to analysis.
3. Results 3.1. Working time The working time was clearly altered with the change in water content. As can be seen in Fig. 1, the force needed to extrude the 3.4 wt.% paste increased much faster than the force needed for 1.7 wt.% paste. In addition, the extrusion force of the cement with no water only increased slightly during the first 6 h. Furthermore, the cement containing 3.4 wt.% water was ejectable for 2–2.5 h, the cement with 1.7 wt.% water for 5–6 h, while the cement containing no water could be stored at room temperature up to 2 weeks before it could no longer be ejected from the syringe.
Injectability of the cements was evaluated with the use of a universal testing machine (Shimadzu AGS-H). The force needed to extrude the paste from a 3 ml disposable syringe with a barrel diameter of 8.55 mm and an outlet diameter of 1.90 mm at a cross-head speed of 60 mm/min was measured. The value obtained was the mean force between 10 and 30 mm displacement. From each batch three measurements were performed, the first starting 5 min after finishing mixing and the last starting 9 min after finishing mixing, resulting in a total of 12 measurements per group. 2.4. Setting time (ST) Five cylindrical molds, ∅ 6 mm, height 3 mm, were filled with cement. The filled molds were immersed in 37 °C phosphate buffered saline solution (PBS, pH 7.4, Sigma) 5 min after finishing mixing. Setting time was measured from the immersion in PBS. The cement was considered to have set when the sample could support a 453.5 g Gillmore needle with a tip diameter of 1.06 mm without breaking. The five samples were tested consecutively every 3 min. The mean between the time when the sample supported the weight and the previous time where the sample broke under the load was regarded as the setting time. A total of 8 measurements were made per group. 2.5. Compressive strength Cylindrical molds, Ø 6 mm and height 12 mm were filled with cement and immersed in 50 ml PBS at 37 °C in a sealed beaker. After 24 h the samples were removed from the molds and carefully
Fig. 1. Working time test for 0 wt.%, 1.7 wt.% and 3.4 wt.% water. (n = 2).
J. Engstrand et al. / Materials Science and Engineering C 33 (2013) 527–531
529
Fig. 2. Extrusion forces for cements with different water contents. Fig. 4. Compression strength for cements with different water contents.
3.2. Ejectability The force needed to extrude the cement from the syringe clearly decreased with water content (Fig. 2), from 107 (±15) N with no water added, to 39 (± 9) N with only 1.7 wt.% water added and to 16 (± 3) N with 3.4 wt.% water added. Immediately after mixing both 5.1 wt.% and 6.8 wt.% pastes were very fluid. However, the extrusion force for 6.8 wt.% water was not measured since it sets in the syringe prior to the injection test. Furthermore, due to the high amounts of water present in the paste containing 5.1 wt.% water, the setting had already begun in the syringe. This led to an increased extrusion force with time, and the measured forces were higher for every consecutive sample within the same batch. Hence, only the results for the three stable compositions are reported. 3.3. Setting time The setting time for the cement containing 0 wt.% water was 29 (±4) min, and a decrease in setting time can be seen with an increasing amount of water present in the paste (Fig. 3). With the addition of 1.7 wt.% water the setting time was reduced to 19 (±2). The setting time for 6.8 wt.% water was not measured since it sets already in the syringe before injection into the molds. 3.4. Compression strength The compressive strength showed a maximum at 1.7 wt.% water content, 11.2 (± 1.5) MPa (Fig. 4). The addition of small amounts of
Fig. 3. Setting times for cements with different water contents.
water increased the strength compared to the dry samples; however, when further water was added the strength was gradually reduced. 3.5. SEM The cement containing 0 wt.% water consisted of much larger grains than the samples containing no water (Fig. 5). It should also be noted that this cement showed quite a wide distribution of the grain sizes (0.5–5 μm). Samples containing 5.1 and 6.8 wt.% water seemed to have quite similar structures, but larger pores were visible in the 6.8 wt.% sample. The sample containing 3.4 wt.% water had about the same grain size as the 5.1 and 6.8 wt.% samples (0.1– 0.5 μm), but the grains seemed to be slightly more connected to each other. The sample containing only 1.7 wt.% water showed medium sized grains (0.5–1 μm) that also seemed to be slightly fused together. 3.6. XRD In Fig. 6 the spectra for the composition extremes, 0 and 6.8 wt.% water content, are shown. However, all cement formulations, regardless of water content, formed monetite after setting. 4. Discussion As established in the results section, the addition of small amounts of water had a great influence on the properties of the premixed cement. As expected, the working time and setting time of the cements is strongly dependent on the water content. This was clearly exemplified as the paste, which was prepared without the addition of water, was injectable for 2 weeks, while the working time was decreased to a couple of hours of when only 1.7 wt.% water was added. A further reduction to merely 10 min was seen when more than 5 wt.% water was added (Fig. 1), working times that are comparable with the working time of conventional water mixed cements [22,23]. The same trend was seen for the final setting time, which was reduced from 29 min to 19 min with the addition of only 1.7 wt.% of water (Fig. 3). To avoid stressful handling in the operating room setting times around 20 min are preferable, which is only achieved by the addition of either glycolic or citric acid to regular CPCs [24]. Since water triggers the setting of the cements, it is natural that both the working time and the setting time decrease when water is added. Furthermore, the extrusion force decreased with the addition of water (Fig. 2). The added water contributes to a paste with lower viscosity in three ways; a lower P/L, dissolution of parts of the
530
J. Engstrand et al. / Materials Science and Engineering C 33 (2013) 527–531
Fig. 5. Representative SEM micrographs of the microstructure of set cements containing (A) 0 wt.%, (B) 1.7 wt.%, (C) 3.4 wt.%, (D) 5.1 wt.%, and (E) 6.8 wt.% water.
powder phase already in the syringe, and lowering the viscosity of the liquid phase. However, it is interesting to note that the addition of merely 1.7 wt.% reduced the extrusion force by 64%. The compressive strengths measured herein are comparable with other premixed cements [25]; however, the conventional water mixed CPCs have a wider range of compressive strengths that could reach up to as high as tree times the maximum of the premixed ones [22]. The change in compressive strength for the samples containing different amounts of water (Fig. 4) could possibly be explained by the difference in grain size that was observed in the SEM images in Fig. 5, combined with the probable variation in porosity for the cements. It is well known that smaller grains results in a stronger material, and the increased strength for cements containing 1.7 wt.% water compared to no added water can be explained by this. Although the SEM micrographs show that the particles are decreasing in size with increased water content up to 5.1 wt.%, a plateau in compressive strength can be seen between 1.7 and 3.4 wt.%. Furthermore, for samples containing ≥5.1 wt.% water the strength is decreasing. This trend is likely due to the higher porosity of these samples, resulting from the higher liquid (water+glycerol) content, which is well known to increase the porosity of a cement [22]. The difference in crystal size is caused by the presence of water during mixing. In the cements where water was added, many nucleation sites are formed simultaneously throughout the cement volume when the raw materials are dissolved in the water. The large
amount of nucleation sites results in many small grains. In the cements with 0 wt.% water, some nuclei may form but they start to grow first when water diffuses into the sample. Eq. (1) shows that a new nucleus can only be formed if a critical radius can be exceeded. ΔG ¼
4 3 2 πr Gv þ 4πr σ 3
ð1Þ
(where ΔG = Gibbs free energy, Gv = free energy per volume unit, σ = free energy per surface area). If not enough ions are available for the formation of a new nucleus the growth of already existing nuclei is instead favored. Thus, when only a low amount of water is added the formation of new nuclei of the critical size is less likely to occur than the growth of already existing nuclei. The crystals in the samples containing 0 wt.% water will therefore be large rather than many. XRD showed that the phase formed after setting of all investigated cements was monetite and not brushite, as is the case for water-mixed acidic cements. This is in accordance with results from previous studies [15] on premixed cements. Although monetite is more stable than brushite under biological conditions, the nucleation of monetite is difficult and brushite is formed in water-mixed cements when they are mixed at room temperatures; however, at temperatures around 37 °C normally a mixture of brushite and monetite is formed [26,27].
J. Engstrand et al. / Materials Science and Engineering C 33 (2013) 527–531
531
shown to work for apatite forming premixed cements [29]. Preliminary studies have shown that monetite forming pastes can also be kept for a long time in the freezer and after thawing the setting time is comparable with the setting time before freezing. However, this requires further investigation to find an optimized method. 5. Conclusion This study showed that the addition of small amounts of water decreased the force needed for extruding the cement from the syringe, decreased the setting time, and increased the compression strength, while maintaining a working time of several hours. This can be exemplified by the force needed to extrude the cement from a syringe that was reduced from 107 N to 39 N when adding 1.7 wt.% water. The compressive strength for the same example was increased from around 6 MPa to just above 11 MPa, and the setting time was decreased with 10 min to 19 min; however, the working time was reduced but still remained at about 5 h. Acknowledgment The Swedish Research Council and the European project Biodesign are gratefully acknowledged for financial support. References Fig. 6. XRD of cements containing 0 and 6.8 wt.% water after setting. Monetite reference pattern from PDF #00-009-0080.
Furthermore, due to the low amount of water present and temperature of 37 °C during setting of premixed cements, it is not surprising that the stable monetite is formed exclusively. As previously mentioned, results indicated that the addition of small amounts of water reduced the setting time and increased the strength of these cements, but water addition also gave another advantage by lowering the force needed to extrude the paste from the syringe. The P/L affects both setting time and strength of the material. Since the extrusion force is decreased when some water is added, P/G could be altered some, while still keeping the paste ejectable, suggesting that the setting time and compression strength thus could be improved even more. The cement prepared in this study had similar setting time and strength as that of previously studied cements where MCPM was used [15]. This shows that MCPA can be used instead of MCPM in acidic premixed cements without significantly altering the cement properties. This is interesting since this could possibly increase the shelf life of the premixed cement. Previous publications have shown that a mixture of dry MCPM and β-TCP powders react and form monetite, when stored at moderate temperatures [28]. This reaction will produce water when the water in the MCPM is released, which then catalyzes the reaction. By using MCPA, this catalytic effect would be eliminated, which would likely increase the shelf life for the cement. By the addition of small amounts of water (1.5–3.5 wt.%) an injectable CPC is obtained that has suitable characteristics regarding handling, setting time, and strength for use as a bone void filler. The working time of these cements can be measured in hours, compared to minutes for existing water-based CPC. This will give the surgeon plenty of time for placing the paste. However, the problem with obtaining sufficient shelf life becomes more crucial when water is added. Ways of improving shelf life could, for example, be to use a dual compartment syringe setup with two pastes; one containing MCPA and glycerol, and one containing β-TCP, water and glycerol, which are mixed during extrusion from the syringe through a mixing-tip. Another way of improving shelf-life could be to store the pastes in a freezer which has been
[1] T.D. Driskell, A.L. Heller, J.H. Koenigs, in: U.S. patent (Ed.), Miter, Inc. (Worthington, OH) 1975. [2] W. Brown, L.C. Chow, J. Dent. Res. 62 (1983) 672. [3] R. Legeros, A. Chohayeb, A. Shulman, J. Dent. Res. 61 (1982) 343. [4] M. Bohner, P. Van Landuyt, H.P. Merkle, J. Lemaitre, J. Mater. Sci. Mater. Med. 8 (1997) 675–681. [5] A.A. Mirtchi, J. Lemaitre, N. Terao, Biomaterials 10 (1989) 475–480. [6] T.R. Desai, S.B. Bhaduri, A.C. Tas, Ceram. Eng. Sci. Proc. 27 (2007) 61–69. [7] C.S. Liu, W. Shen, Y.F. Gu, L.M. Hu, J. Biomed. Mater. Res. 35 (1997) 75–80. [8] A. Joeris, S. Ondrus, L. Planka, P. Gal, T. Slongo, Eur. J. Pediatr. Surg. 20 (2010) 24–28. [9] K.D. Wolff, S. Swaid, D. Nolte, R.A. Bockmann, F. Holzle, C. Muller-Mai, J. Cranio-Maxillofac. Surg. 32 (2004) 71–79. [10] M. Bohner, Eur.Cells & Mater. 20 (2010) 1–12. [11] F. Tamimi, Z. Sheikh, J. Barralet, Acta Biomater. 8 (2012) 474–487. [12] F. Tamimi, J. Torres, D. Bassett, J. Barralet, E.L. Cabarcos, Biomaterials 31 (2010) 2762–2769. [13] D. Apelt, F. Theiss, A.O. El-Warrak, K. Zlinszky, R. Bettschart-Wolfisberger, M. Bohner, S. Matter, J.A. Auer, B. von Rechenberg, Biomaterials 25 (2004) 1439–1451. [14] A. Sugawara, L.C. Chow, S. Takagi, H. Chohayeb, J Endod. 16 (1990) 162–165. [15] J. Aberg, H. Brisby, H.B. Henriksson, A. Lindahl, P. Thomsen, H. Engqvist, J. Biomed. Mater. Res. B Appl. Biomater. 93B (2010) 436–441. [16] B. Han, P.-W. Ma, L.-L. Zhang, Y.-J. Yin, K.-D. Yao, F.-J. Zhang, Y.-D. Zhang, X.-L. Li, W. Nie, Acta Biomater. 5 (2009) 3165–3177. [17] L.E. Carey, H.H.K. Xu, C.G. Simon, S. Takagi, L.C. Chow, Biomaterials 26 (2005) 5002–5014. [18] S. Takagi, L.C. Chow, S. Hirayama, A. Sugawara, Journal of Biomedical Materials Research Part B—Applied Biomaterials 67B (2003) 689–696. [19] A. Sugawara, K. Fujikawa, S. Hirayama, S. Takagi, L.C. Chow, J. Res. Natl. Inst. Stand. 115 (2010) 277–290. [20] J. Aberg, H.B. Henriksson, H. Engqvist, A. Palmquist, C. Brantsing, A. Lindahl, P. Thomsen, H. Brisby, J. Biomed. Mater. Res. A 100A (2012) 1269–1278. [21] H.W.E. Larson, Ind. Eng. Chem. Anal. Ed. 7 (1935) 401–406. [22] M.P. Hofmann, A.R. Mohammed, Y. Perrie, U. Gbureck, J.E. Barralet, Acta Biomater. 5 (2009) 43–49. [23] Z. Huan, J. Chang, Acta Biomater. 5 (2009) 1253–1264. [24] F.T. Mariño, J. Torres, M. Hamdan, C.R. Rodríguez, E.L. Cabarcos, J. Biomed. Mater. Res. B Appl. Biomater. 83B (2007) 571–579. [25] B. Han, P.W. Ma, L.L. Zhang, Y.J. Yin, K.D. Yao, F.J. Zhang, Y.D. Zhang, X.L. Li, W. Nie, Acta Biomater. 5 (2009) 3165–3177. [26] J.C. Elliott, Rev. Mineral. Geochem. 48 (2002) 427–453. [27] L.M. Grover, U. Gbureck, A.M. Young, A.J. Wright, J.E. Barralet, J. Mater. Chem. 15 (2005) 4955–4962. [28] U. Gbureck, S. Dembski, R. Thull, J.E. Barralet, Biomaterials 26 (2005) 3691–3697. [29] I. Rajzer, O. Castano, E. Engel, J.A. Planell, J. Mater. Sci. Mater. Manag. 21 (2010) 2049–2056.
Contents lists available at SciVerse ScienceDirect
Materials Science and Engineering C journal homepage: www.elsevier.com/locate/msec
Influence of water content on hardening and handling of a premixed calcium phosphate cement Johanna Engstrand ⁎, Jonas Aberg, Håkan Engqvist Applied Materials Science, Department of Engineering Sciences, Uppsala University, Sweden
a r t i c l e
i n f o
Article history: Received 24 January 2012 Received in revised form 12 September 2012 Accepted 28 September 2012 Available online 6 October 2012 Keywords: Premixed Calcium phosphate cements Injectability Strength Setting time Water content
a b s t r a c t Handling of calcium phosphate cements is difficult, where problems often arise during mixing, transferring to syringes, and subsequent injection. Via the use of premixed cements the risk of handling complications is reduced. However, for premixed cements to work in a clinical situation the setting time needs to be improved. The objective of this study is to investigate the influence of the addition of water on the properties of premixed cement. Monetite-forming premixed cements with small amounts of added water (less than 6.8 wt.%) were prepared and the influence on injectability, working time, setting time and mechanical strength was evaluated. The results showed that the addition of small amounts of water had significant influence on the properties of the premixed cement. With the addition of just 1.7 wt.% water, the force needed to extrude the cement from a syringe was reduced from 107 (± 15) N to 39 (± 9) N, the compression strength was almost doubled, and the setting time decreased from 29 (±4) min to 19 (± 2) min, while the working time remained 5 to 6 h. This study demonstrates the importance of controlling the water content in premixed cement pastes and how water can be used to improve the properties of premixed cements. © 2012 Elsevier B.V. All rights reserved.
1. Introduction Calcium phosphate based materials are considered ideal for bone replacement, since they resemble the mineral phase of bone; calcium deficient hydroxyapatite. The first calcium phosphate paste was patented in 1975 by Driskell et al. [1], and the first calcium phosphate cements (CPC) were reported in the early 1980's [2,3]. Brushite and apatite are the two main types of CPC found, differing mainly in pH. In an acidic environment (pH b ~4.2) the cement precipitates to brushite [4,5]; although monetite has also been reported [6]. At a higher pH however, apatite precipitation occurs [7]. Both types of CPC have good biocompatibility and are used clinically [8–11]. Furthermore, promising in vivo results have been shown for monetite based materials [12]. When the cements are compared in vivo, the acidic cements have been shown to resorb faster than apatite cements due to their increased solubility in physiological pH [13]. This allows for faster bone in-growth to occur into the cement-filled defect [12]. The mixing of commercially available water-based CPCs at present has to be done in the operating room, since hardening starts immediately after mixing. Each product has specific mixing equipment and instructions that can be quite tedious, making them difficult to work with. All stages of mixing, transfer, and injection are time sensitive, and must be performed within approximately 10 min. If the preparation of the cement goes over the allotted time, the cement hardens and injection of the cement is no longer
possible. To overcome the mentioned problems, premixed calcium phosphate cements (pCPC) have been developed over the last few years. A pCPC is based on calcium phosphate powders mixed with a non-aqueous water-soluble liquid; e.g. glycerol or poly(ethylene glycol) [14–16]. Since the setting of these cements commences when the liquid phase is exchanged with water or body fluid, these cements could be stored and delivered in a syringe. This eliminates the stress related to the preparation of conventional water-mixed CPC's in the operating room. Both apatite [17,18] and acidic [15,16] pCPCs have been reported. In vivo evaluation of pCPCs has demonstrated similar behavior as conventionally water-mixed cements [19,20], making them a viable alternative. Although the water-mixed acidic cements set within a few minutes [15], due to the slow water–glycerol exchange the setting of the acidic pCPC is more time consuming [15,16]. To make the pCPC interesting commercially, the setting time needs to be reduced. It is probable that addition of small amounts of water to the cement mixture could reduce the setting time but it is not known how much is needed in order to achieve clinically relevant times. The objective of this study was to investigate the influence of small amounts of water on injectability, setting time, working time and mechanical properties of an acidic premixed cement mixture. 2. Method 2.1. Cement preparation
⁎ Corresponding author at: Ångströmlab, Lägerhyddsvägen 1, Box 534, 751 21 Uppsala, Sweden. Tel.: +46 18 471 79 46; fax: +46 18 471 35 72. E-mail address: [email protected] (J. Engstrand). 0928-4931/$ – see front matter © 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.msec.2012.09.026
The powders used in the study were β-tricalcium phosphate (β-TCP, Fluka) and monocalcium phosphate hydrate (MCPH, Alfa
528
J. Engstrand et al. / Materials Science and Engineering C 33 (2013) 527–531
Aesar). The MCPH was dried in 110 °C for 3 days in order to remove all water, resulting in monocalcium phosphate anhydrous (MCPA) [21]. MCPA was used to facilitate the control of water content in the cement mixture. β-TCP and MCPA were stored in a desiccator over silica gel. For all cements the molar ratio used for MCPA and β-TCP was 1:1. Glycerol (anhydrous, Sigma) was used as a mixing liquid. Water was deliberately added to the mixture in percent of the total weight; 0, 1.7, 3.4, 5.1, and 6.8 wt.%, and four batches were made per group. The powder to glycerol (P/G) ratio was kept at 4.0 g/ml for all cements. Therefore, the total powder to liquid ratio (P/L) was lower for cements containing more water. A vacuum mixer (Twister, Renfert) was used for all mixing. First water was added to glycerol and hand mixed for some seconds before MCPA was added and all components were mixed for 50 s using the vacuum mixer. β-TCP powder was added and everything was mixed twice for 50 s. Between each mixing the cement attached to the wall and paddle was loosened in order to obtain homogenous cement pastes.
polished to make the sample sides parallel and obtain uniform height. The maximum compressive stress until failure was measured using a universal testing machine (Shimadzu AGS-H), with a cross-head speed of 1 mm/min. A thin plastic film was placed between the sample and the cross-head in order to reduce the effect of potential surface defects. A total of 12 samples were made for each group.
2.6. SEM To study the pore structure of the set cements, the cross-sections were investigated with scanning electron microscopy (SEM, LEO 1550, Zeiss). The samples were polished using 1200 grit SiC paper. A thin gold/palladium coating was sputtered onto the surface before analysis to avoid charging of the surface.
2.7. X-ray diffraction (XRD) 2.2. Working time Working time of the cements was evaluated by preparing two batches with 0, 1.7 and 3.4 wt.% water each. Cements with higher than 3.4 wt.% water content could not be injected after 15 min and were, therefore, not evaluated further. The P/G for 1.7 and 3.4 wt.% was 4.0, while the P/G for 0 wt.% was slightly lower at 3.75, in order to start with comparable extrusion forces at t = 0. Syringes with a barrel diameter of 8.55 mm and an outlet diameter of 1.90 mm were filled with approximately 1 ml of paste at t = 0 and the syringes were stored in a dry environment at room temperature until they were tested. The force needed to extrude the paste from the syringe at a cross-head speed of 60 mm/min was measured continually until the cement was too viscous to be extruded from the syringe, i.e. the piston folded at F = 240 N. The time between measuring points was different for varying water contents in order to get approximately the same amount of measuring points for all cements. 2.3. Injectability
The resulting phase composition of the cements after setting was analyzed using XRD (diffractometer, Siemens). Diffraction angles (2θ) 20–40 were analyzed at 0.45°/min. The set samples were crushed using a mortar prior to analysis.
3. Results 3.1. Working time The working time was clearly altered with the change in water content. As can be seen in Fig. 1, the force needed to extrude the 3.4 wt.% paste increased much faster than the force needed for 1.7 wt.% paste. In addition, the extrusion force of the cement with no water only increased slightly during the first 6 h. Furthermore, the cement containing 3.4 wt.% water was ejectable for 2–2.5 h, the cement with 1.7 wt.% water for 5–6 h, while the cement containing no water could be stored at room temperature up to 2 weeks before it could no longer be ejected from the syringe.
Injectability of the cements was evaluated with the use of a universal testing machine (Shimadzu AGS-H). The force needed to extrude the paste from a 3 ml disposable syringe with a barrel diameter of 8.55 mm and an outlet diameter of 1.90 mm at a cross-head speed of 60 mm/min was measured. The value obtained was the mean force between 10 and 30 mm displacement. From each batch three measurements were performed, the first starting 5 min after finishing mixing and the last starting 9 min after finishing mixing, resulting in a total of 12 measurements per group. 2.4. Setting time (ST) Five cylindrical molds, ∅ 6 mm, height 3 mm, were filled with cement. The filled molds were immersed in 37 °C phosphate buffered saline solution (PBS, pH 7.4, Sigma) 5 min after finishing mixing. Setting time was measured from the immersion in PBS. The cement was considered to have set when the sample could support a 453.5 g Gillmore needle with a tip diameter of 1.06 mm without breaking. The five samples were tested consecutively every 3 min. The mean between the time when the sample supported the weight and the previous time where the sample broke under the load was regarded as the setting time. A total of 8 measurements were made per group. 2.5. Compressive strength Cylindrical molds, Ø 6 mm and height 12 mm were filled with cement and immersed in 50 ml PBS at 37 °C in a sealed beaker. After 24 h the samples were removed from the molds and carefully
Fig. 1. Working time test for 0 wt.%, 1.7 wt.% and 3.4 wt.% water. (n = 2).
J. Engstrand et al. / Materials Science and Engineering C 33 (2013) 527–531
529
Fig. 2. Extrusion forces for cements with different water contents. Fig. 4. Compression strength for cements with different water contents.
3.2. Ejectability The force needed to extrude the cement from the syringe clearly decreased with water content (Fig. 2), from 107 (±15) N with no water added, to 39 (± 9) N with only 1.7 wt.% water added and to 16 (± 3) N with 3.4 wt.% water added. Immediately after mixing both 5.1 wt.% and 6.8 wt.% pastes were very fluid. However, the extrusion force for 6.8 wt.% water was not measured since it sets in the syringe prior to the injection test. Furthermore, due to the high amounts of water present in the paste containing 5.1 wt.% water, the setting had already begun in the syringe. This led to an increased extrusion force with time, and the measured forces were higher for every consecutive sample within the same batch. Hence, only the results for the three stable compositions are reported. 3.3. Setting time The setting time for the cement containing 0 wt.% water was 29 (±4) min, and a decrease in setting time can be seen with an increasing amount of water present in the paste (Fig. 3). With the addition of 1.7 wt.% water the setting time was reduced to 19 (±2). The setting time for 6.8 wt.% water was not measured since it sets already in the syringe before injection into the molds. 3.4. Compression strength The compressive strength showed a maximum at 1.7 wt.% water content, 11.2 (± 1.5) MPa (Fig. 4). The addition of small amounts of
Fig. 3. Setting times for cements with different water contents.
water increased the strength compared to the dry samples; however, when further water was added the strength was gradually reduced. 3.5. SEM The cement containing 0 wt.% water consisted of much larger grains than the samples containing no water (Fig. 5). It should also be noted that this cement showed quite a wide distribution of the grain sizes (0.5–5 μm). Samples containing 5.1 and 6.8 wt.% water seemed to have quite similar structures, but larger pores were visible in the 6.8 wt.% sample. The sample containing 3.4 wt.% water had about the same grain size as the 5.1 and 6.8 wt.% samples (0.1– 0.5 μm), but the grains seemed to be slightly more connected to each other. The sample containing only 1.7 wt.% water showed medium sized grains (0.5–1 μm) that also seemed to be slightly fused together. 3.6. XRD In Fig. 6 the spectra for the composition extremes, 0 and 6.8 wt.% water content, are shown. However, all cement formulations, regardless of water content, formed monetite after setting. 4. Discussion As established in the results section, the addition of small amounts of water had a great influence on the properties of the premixed cement. As expected, the working time and setting time of the cements is strongly dependent on the water content. This was clearly exemplified as the paste, which was prepared without the addition of water, was injectable for 2 weeks, while the working time was decreased to a couple of hours of when only 1.7 wt.% water was added. A further reduction to merely 10 min was seen when more than 5 wt.% water was added (Fig. 1), working times that are comparable with the working time of conventional water mixed cements [22,23]. The same trend was seen for the final setting time, which was reduced from 29 min to 19 min with the addition of only 1.7 wt.% of water (Fig. 3). To avoid stressful handling in the operating room setting times around 20 min are preferable, which is only achieved by the addition of either glycolic or citric acid to regular CPCs [24]. Since water triggers the setting of the cements, it is natural that both the working time and the setting time decrease when water is added. Furthermore, the extrusion force decreased with the addition of water (Fig. 2). The added water contributes to a paste with lower viscosity in three ways; a lower P/L, dissolution of parts of the
530
J. Engstrand et al. / Materials Science and Engineering C 33 (2013) 527–531
Fig. 5. Representative SEM micrographs of the microstructure of set cements containing (A) 0 wt.%, (B) 1.7 wt.%, (C) 3.4 wt.%, (D) 5.1 wt.%, and (E) 6.8 wt.% water.
powder phase already in the syringe, and lowering the viscosity of the liquid phase. However, it is interesting to note that the addition of merely 1.7 wt.% reduced the extrusion force by 64%. The compressive strengths measured herein are comparable with other premixed cements [25]; however, the conventional water mixed CPCs have a wider range of compressive strengths that could reach up to as high as tree times the maximum of the premixed ones [22]. The change in compressive strength for the samples containing different amounts of water (Fig. 4) could possibly be explained by the difference in grain size that was observed in the SEM images in Fig. 5, combined with the probable variation in porosity for the cements. It is well known that smaller grains results in a stronger material, and the increased strength for cements containing 1.7 wt.% water compared to no added water can be explained by this. Although the SEM micrographs show that the particles are decreasing in size with increased water content up to 5.1 wt.%, a plateau in compressive strength can be seen between 1.7 and 3.4 wt.%. Furthermore, for samples containing ≥5.1 wt.% water the strength is decreasing. This trend is likely due to the higher porosity of these samples, resulting from the higher liquid (water+glycerol) content, which is well known to increase the porosity of a cement [22]. The difference in crystal size is caused by the presence of water during mixing. In the cements where water was added, many nucleation sites are formed simultaneously throughout the cement volume when the raw materials are dissolved in the water. The large
amount of nucleation sites results in many small grains. In the cements with 0 wt.% water, some nuclei may form but they start to grow first when water diffuses into the sample. Eq. (1) shows that a new nucleus can only be formed if a critical radius can be exceeded. ΔG ¼
4 3 2 πr Gv þ 4πr σ 3
ð1Þ
(where ΔG = Gibbs free energy, Gv = free energy per volume unit, σ = free energy per surface area). If not enough ions are available for the formation of a new nucleus the growth of already existing nuclei is instead favored. Thus, when only a low amount of water is added the formation of new nuclei of the critical size is less likely to occur than the growth of already existing nuclei. The crystals in the samples containing 0 wt.% water will therefore be large rather than many. XRD showed that the phase formed after setting of all investigated cements was monetite and not brushite, as is the case for water-mixed acidic cements. This is in accordance with results from previous studies [15] on premixed cements. Although monetite is more stable than brushite under biological conditions, the nucleation of monetite is difficult and brushite is formed in water-mixed cements when they are mixed at room temperatures; however, at temperatures around 37 °C normally a mixture of brushite and monetite is formed [26,27].
J. Engstrand et al. / Materials Science and Engineering C 33 (2013) 527–531
531
shown to work for apatite forming premixed cements [29]. Preliminary studies have shown that monetite forming pastes can also be kept for a long time in the freezer and after thawing the setting time is comparable with the setting time before freezing. However, this requires further investigation to find an optimized method. 5. Conclusion This study showed that the addition of small amounts of water decreased the force needed for extruding the cement from the syringe, decreased the setting time, and increased the compression strength, while maintaining a working time of several hours. This can be exemplified by the force needed to extrude the cement from a syringe that was reduced from 107 N to 39 N when adding 1.7 wt.% water. The compressive strength for the same example was increased from around 6 MPa to just above 11 MPa, and the setting time was decreased with 10 min to 19 min; however, the working time was reduced but still remained at about 5 h. Acknowledgment The Swedish Research Council and the European project Biodesign are gratefully acknowledged for financial support. References Fig. 6. XRD of cements containing 0 and 6.8 wt.% water after setting. Monetite reference pattern from PDF #00-009-0080.
Furthermore, due to the low amount of water present and temperature of 37 °C during setting of premixed cements, it is not surprising that the stable monetite is formed exclusively. As previously mentioned, results indicated that the addition of small amounts of water reduced the setting time and increased the strength of these cements, but water addition also gave another advantage by lowering the force needed to extrude the paste from the syringe. The P/L affects both setting time and strength of the material. Since the extrusion force is decreased when some water is added, P/G could be altered some, while still keeping the paste ejectable, suggesting that the setting time and compression strength thus could be improved even more. The cement prepared in this study had similar setting time and strength as that of previously studied cements where MCPM was used [15]. This shows that MCPA can be used instead of MCPM in acidic premixed cements without significantly altering the cement properties. This is interesting since this could possibly increase the shelf life of the premixed cement. Previous publications have shown that a mixture of dry MCPM and β-TCP powders react and form monetite, when stored at moderate temperatures [28]. This reaction will produce water when the water in the MCPM is released, which then catalyzes the reaction. By using MCPA, this catalytic effect would be eliminated, which would likely increase the shelf life for the cement. By the addition of small amounts of water (1.5–3.5 wt.%) an injectable CPC is obtained that has suitable characteristics regarding handling, setting time, and strength for use as a bone void filler. The working time of these cements can be measured in hours, compared to minutes for existing water-based CPC. This will give the surgeon plenty of time for placing the paste. However, the problem with obtaining sufficient shelf life becomes more crucial when water is added. Ways of improving shelf life could, for example, be to use a dual compartment syringe setup with two pastes; one containing MCPA and glycerol, and one containing β-TCP, water and glycerol, which are mixed during extrusion from the syringe through a mixing-tip. Another way of improving shelf-life could be to store the pastes in a freezer which has been
[1] T.D. Driskell, A.L. Heller, J.H. Koenigs, in: U.S. patent (Ed.), Miter, Inc. (Worthington, OH) 1975. [2] W. Brown, L.C. Chow, J. Dent. Res. 62 (1983) 672. [3] R. Legeros, A. Chohayeb, A. Shulman, J. Dent. Res. 61 (1982) 343. [4] M. Bohner, P. Van Landuyt, H.P. Merkle, J. Lemaitre, J. Mater. Sci. Mater. Med. 8 (1997) 675–681. [5] A.A. Mirtchi, J. Lemaitre, N. Terao, Biomaterials 10 (1989) 475–480. [6] T.R. Desai, S.B. Bhaduri, A.C. Tas, Ceram. Eng. Sci. Proc. 27 (2007) 61–69. [7] C.S. Liu, W. Shen, Y.F. Gu, L.M. Hu, J. Biomed. Mater. Res. 35 (1997) 75–80. [8] A. Joeris, S. Ondrus, L. Planka, P. Gal, T. Slongo, Eur. J. Pediatr. Surg. 20 (2010) 24–28. [9] K.D. Wolff, S. Swaid, D. Nolte, R.A. Bockmann, F. Holzle, C. Muller-Mai, J. Cranio-Maxillofac. Surg. 32 (2004) 71–79. [10] M. Bohner, Eur.Cells & Mater. 20 (2010) 1–12. [11] F. Tamimi, Z. Sheikh, J. Barralet, Acta Biomater. 8 (2012) 474–487. [12] F. Tamimi, J. Torres, D. Bassett, J. Barralet, E.L. Cabarcos, Biomaterials 31 (2010) 2762–2769. [13] D. Apelt, F. Theiss, A.O. El-Warrak, K. Zlinszky, R. Bettschart-Wolfisberger, M. Bohner, S. Matter, J.A. Auer, B. von Rechenberg, Biomaterials 25 (2004) 1439–1451. [14] A. Sugawara, L.C. Chow, S. Takagi, H. Chohayeb, J Endod. 16 (1990) 162–165. [15] J. Aberg, H. Brisby, H.B. Henriksson, A. Lindahl, P. Thomsen, H. Engqvist, J. Biomed. Mater. Res. B Appl. Biomater. 93B (2010) 436–441. [16] B. Han, P.-W. Ma, L.-L. Zhang, Y.-J. Yin, K.-D. Yao, F.-J. Zhang, Y.-D. Zhang, X.-L. Li, W. Nie, Acta Biomater. 5 (2009) 3165–3177. [17] L.E. Carey, H.H.K. Xu, C.G. Simon, S. Takagi, L.C. Chow, Biomaterials 26 (2005) 5002–5014. [18] S. Takagi, L.C. Chow, S. Hirayama, A. Sugawara, Journal of Biomedical Materials Research Part B—Applied Biomaterials 67B (2003) 689–696. [19] A. Sugawara, K. Fujikawa, S. Hirayama, S. Takagi, L.C. Chow, J. Res. Natl. Inst. Stand. 115 (2010) 277–290. [20] J. Aberg, H.B. Henriksson, H. Engqvist, A. Palmquist, C. Brantsing, A. Lindahl, P. Thomsen, H. Brisby, J. Biomed. Mater. Res. A 100A (2012) 1269–1278. [21] H.W.E. Larson, Ind. Eng. Chem. Anal. Ed. 7 (1935) 401–406. [22] M.P. Hofmann, A.R. Mohammed, Y. Perrie, U. Gbureck, J.E. Barralet, Acta Biomater. 5 (2009) 43–49. [23] Z. Huan, J. Chang, Acta Biomater. 5 (2009) 1253–1264. [24] F.T. Mariño, J. Torres, M. Hamdan, C.R. Rodríguez, E.L. Cabarcos, J. Biomed. Mater. Res. B Appl. Biomater. 83B (2007) 571–579. [25] B. Han, P.W. Ma, L.L. Zhang, Y.J. Yin, K.D. Yao, F.J. Zhang, Y.D. Zhang, X.L. Li, W. Nie, Acta Biomater. 5 (2009) 3165–3177. [26] J.C. Elliott, Rev. Mineral. Geochem. 48 (2002) 427–453. [27] L.M. Grover, U. Gbureck, A.M. Young, A.J. Wright, J.E. Barralet, J. Mater. Chem. 15 (2005) 4955–4962. [28] U. Gbureck, S. Dembski, R. Thull, J.E. Barralet, Biomaterials 26 (2005) 3691–3697. [29] I. Rajzer, O. Castano, E. Engel, J.A. Planell, J. Mater. Sci. Mater. Manag. 21 (2010) 2049–2056.
