Phosphorus chemistry: from small molecules, to polymers, to pharmaceutical and industrial applications

DSpace https://erepo.uef.fi

Rinnakkaistallenteet Terveystieteiden tiedekunta

2018

Phosphorus chemistry: from small molecules, to polymers, to

pharmaceutical and industrial applications

Papathanasiou, Konstantinos E

Walter de Gruyter GmbH

Tieteelliset aikakauslehtiartikkelit

© IUPAC & De Gruyter

CC BY-NC-ND https://creativecommons.org/licenses/by-nc-nd/4.0/

http://dx.doi.org/10.1515/pac-2018-1012

https://erepo.uef.fi/handle/123456789/7215

Downloaded from University of Eastern Finland's eRepository

Conference paper

Konstantinos E. Papathanasiou, Maria Vassaki, Argyro Spinthaki, Fanouria-Eirini G. Alatzoglou, Eleftherios Tripodianos, Petri Turhanen and Konstantinos D. Demadis*

Phosphorus chemistry: from small molecules, to polymers, to pharmaceutical and industrial applications

https://doi.org/10.1515/pac-2018-1012

Abstract: (Poly)phosphonic acids constitute an exciting family of phosphorus compounds. One of the attrac- tive attributes of these molecules is the rich chemistry of the phosphonate moiety, and, in particular, its high affinity for metal ions and mineral surfaces. Whether the phosphonate group belongs to a “small” molecule or to a polymeric matrix, phosphonate-containing compounds have found a phalanx of real-life applica- tions. Herein, we address a special category of phosphorus compounds called bisphosphonates (BPs, a.k.a.

“-dronates”) and also phosphonate containing polymers. The success of BPs in mitigating osteoporosis not- withstanding, these “-dronate” drugs present a number of challenges. Nevertheless, the main drawback of BPs is their limited oral bioavailability. It is, therefore, imperative to design and fabricate “smart” systems that allow controlled delivery of the active BP agent. Here, easy-to-prepare drug delivery systems are pre- sented based on silica gels. These have been synthesized, characterized, and studied as hosts in the control release of several BP drugs. They exhibit variable release rates and final % release, depending on the nature of bisphosphonate (side-chain length, hydro-philicity/-phobicity, water-solubility), cations present, pH and temperature. These gels are robust, injectable, re-loadable and re-usable. Furthermore, alternative drug delivery systems are presented that are based on metal-organic frameworks (MOFs). In these biologically acceptable inorganic metal ions have been incorporated, together with BPs as the organic portion. These materials have been synthesized, characterized, and studied for the self-sacrificial release (by pH-driven dis- solution) of the BP active ingredient. Several such materials were prepared with a variety of bisphosphonate drugs. They exhibit variable release rates and final % release, depending on the actual structure of the metal- bisphosphonate material. Lastly, we will present the use of phosphonate-grafted polymers as scale inhibitors for water treatment applications.

Keywords: controlled release; hydrogels; osteoporosis; phosphonates; POC-2018; polymers; scale inhibitors;

silica scale; water treatment.

Article note: A collection of papers presented at the 17^th Polymers and Organic Chemistry (POC-2018) conference held 4–7 June 2018 in Le Corum, Montpelier, France.

This paper is based on an invited presentation at the 17^th IUPAC conference on Polymers and Organic Chemistry (POC-17) that was held in Palavas Les Flots, France from 3 to 7 of June 2018.

*Corresponding author: Konstantinos D. Demadis, Crystal Engineering, Growth and Design Laboratory, Department of Chemistry, University of Crete, Voutes Campus, Crete, GR-71003, Greece, e-mail: demadis@uoc.gr

Konstantinos E. Papathanasiou: Department of Chemistry and Biochemistry, Florida International University, MMC 11200 SW 8th Street, Miami, Florida 33199, USA

Maria Vassaki, Argyro Spinthaki, Fanouria-Eirini G. Alatzoglou and Eleftherios Tripodianos: Crystal Engineering, Growth and Design Laboratory, Department of Chemistry, University of Crete, Voutes Campus, Crete, GR-71003, Greece

Petri Turhanen: University of Eastern Finland, School of Pharmacy, Biocenter Kuopio, P.O. Box 1627, Kuopio FIN-70211, Finland

Introduction

Phosphorus (P) chemistry is widespread among literally all chemistry disciplines [1]. A few examples include the fact that P is an important element for plants [2], and a key ingredient in all forms of fertilizers [3]. In nature, P exists in the “V” oxidation state, in the form of phosphate anion. Rocks containing phosphate are fluoro- apatite [3Ca₃(PO₄)₂ · CaF₂] [4], chloroapatite [3Ca₃(PO₄)₂ · CaCl₂] [5], and hydroxyapatite [3Ca₃(PO₄)₂ · Ca(OH)₂] [6]. Phosphate is the major inorganic constituent of bones and teeth in mammals and fish [7]. From an indus- trial and economical view point, P-containing compounds are important commodities [8]. Thus, the chemis- try of P has academic, commercial and industrial significance.

The present paper focuses on two types of P compounds. The first embraces chemicals that are called phosphonic acids (Fig. 1) [9]. The second includes P-containing polymers of various types (Fig. 2) [10].

P HO OH

O R

P OH HO

O NH⁺

P

HO O^–

O R

P OH HO

O

P OH

OH O

R OH

Fig. 1: Schematic structures of three types of phosphonic acids: monophosphonic acids (left), alkylamino- bis(methylenephosphonic acids) (center), and bisphosphonic acids (right).

Fig. 2: Schematic structures of three types of P-containing polymers: Phosphonium bolaamphiphiles (PEGP, upper left), phosphonomethylated polyethyleneimine (PPEI, upper right), and phosphonomethylated chitosan (PCH, lower). The inhibitory effects additives PEGP [11] and PCH [12, 13] are not presented herein, as they have been reported before.

Bisphosphonate compounds (Fig. 1, right) are well-known drugs for several pathological conditions of calcium metabolism since 1977, with osteoporosis being the most prominent [14]. Their commercial names have the ending “-dronate”. For example, etidronate is a member of the first generation osteoporosis drugs.

In spite of their widespread use, they demonstrate a number of problematic issues including fast excretion [15], and several side effects, such as osteonecrosis of the jaw, hypocalcemia, esophageal cancer, ocular inflammation, atrial fibrillation, etc. [16]. Their main disadvantage, however, is their low oral bioavailability (<6 %). This leads to increase drug dosage to achieve the necessary therapeutic level. Hence, design and fabrication of systems that allow controlled delivery of the active bisphosphonate drug are necessary. Some preliminary results have been published recently [17].

In this context, we developed three types of controlled delivery systems: (a) hydrogels based on silica, (b) surface-grafted silica hydrogels with (3-aminopropyl)triethoxysilane (APTES), and (c) metal bisphos- phonate compounds/coordination polymers. In systems (a) and (b) the bisphosphonate drug is entrapped into the silica (or modified silica) matrix, whereas in system (c) the bisphosphonate ligand/linker is the active drug.

The P-containing polymers that are presented here are aimed towards a different application. They are efficient inhibitors of silicic acid polycondensation, a process that produces colloidal silica. The latter is a problematic precipitate/deposit in industrial water systems. The silicic acid polycondensation process and the physicochemical factors that influence it have been part of on-going investigations in our laboratory. The goal is to prevent silica formation, by stabilizing silicic acid, through the discovery and application of chemi- cal scale inhibitors and additives [18–23].

Experimental section

Materials

The bisphosphonate compounds studied in this work are shown in Table 1. Sodium silicate pentahydrate, Na₂SiO₃ · 5H₂O, silicic acid (<20 micron, refined, 99.9 %) and potassium hydroxide was purchased from Sigma- Aldrich. Etidronic acid (either as solid tetrasodium salt or as acid in aqueous solution) was used as received from Solutia Inc. Deuterium oxide (99.9 atom % D) that contained 0.05 wt. % 3-(trimethylsilyl)propionic- 2,2,3,3-d₄ acid, sodium salt purchased also from Sigma-Aldrich. Deionized water from an ion-exchange resin was used for all experiments and stock solution preparations.

Table 1: Bisphosphonate notation and structures.

Bisphosphonate

common name Bisphosphonate

chemical name Bisphosphonate

abbreviation Bisphosphonate

structure Available

as “acid” Available as

“salt”

Etidronic acid 1-hydroxyethane- 1,1-bisphosphonic acid

ETID ^H3C _C ^OH

P OH HO

O P

OH

HO O

YES Tetrasodium

salt

Not available 1-hydroxybutane- 1,1-bisphosphonic acid disodium salt

C3BP

C OH

P OH HO

O P

OH

HO O

NO Disodium

salt

Instrumentation

An AVANCE 300 (Bruker, Karlsruhe, Germany) spectrometer was used for the BP release experiments. SEM data and images collected with a JEOL JSM-6390LV electron microscope.

General comments on the synthesis of bisphosphonates

Alendronate and neridronate [24], C3BP (disodium salt) and C5BP (disodium salt) [25] and C3NBP [24, 25]

were synthesized and characterized as reported in the literature.

Bisphosphonate

common name Bisphosphonate

chemical name Bisphosphonate

abbreviation Bisphosphonate

structure Available

as “acid” Available as

“salt”

Not available 1-hydroxyhexane- 1,1-bisphosphonic acid disodium salt

C5BP

C OH

P OH HO

O P

OH

HO O

NO Disodium

salt

Pamidronic acid 3-Amino-1- hydroxypropane- 1,1-diphosphonic acid

PAM

C OH

P OH HO

O P

OH

HO O H₂N

YES NO

Alendronic acid 3-Amino-1- hydroxybutane- 1,1-diphosphonic acid

ALE

C OH

P OH HO

O P

OH

HO O NH₂

YES Disodium

salt

Not available 4-Amino-1- hydroxypentane- 1,1-diphosphonic acid

C4NBP

C OH

P OH HO

O P

OH

HO O H2N

YES NO

Neridronic acid 6-Amino-1- hydroxyhexane- 1,1-bisphosphonic acid

NER

C OH

P OH HO

O P

OH

HO O NH₂

YES NO

Table 1 (continued)

Syntheses of selected bisphosphonates

General

1H and ³¹P NMR spectra were recorded on a 600 MHz spectrometer operating at 600.2 and 243.0 MHz, respec- tively; ¹³C NMR spectra were recorded on a 500 MHz spectrometer operating at 125.8 MHz. The solvent residual peak was used as a standard for ¹H measurements in D₂O (4.79 ppm) and in ¹³C measurements CD₃OD were added to be as a reference (49.00 ppm) [26]. Eighty five percent H₃PO₄ was used as an external standard in the

31P measurements. The ⁿJ_HH couplings were calculated from proton spectra and all J values are given in Hz. The

nJ_CP couplings were calculated from carbon spectra with the coupling constants given in parenthesis as Hz.

Mass spectra were recorded with a Finnigan LCQ quadrupole ion trap mass spectrometer (Finnigan MAT, San Jose, CA, USA) equipped with an electrospray ionization source. The purity of the products was determined from ¹H and ³¹P NMR spectra and was ≥95 % unless stated otherwise.

Synthesis of 1-hydroxybutane-1,1-bisphosphonic acid disodium salt (C3BP)

Butyric acid (12.5 g, 0.14 mol), phosphorus acid (11.6 g, 0.14 mol) and methanesulfonic acid (40 mL) were heated to 65 °C with stirring in a flask with a reflux condenser and a CaCl₂-tube. PCl₃ (25.0 mL, 0.29 mol) was added drop-wise during 0.5 h and the mixture maintained at 65–70 °C overnight. Water (100 mL) was added to the cooled mixture and the solution was refluxed overnight. The reaction mixture was cooled to room temperature and the product was solidified after pH adjustment to approx. 6–7 with 50 % NaOH. After the solids were filtered, washed with EtOH (300 mL), the crude product was purified by stirring first with distilled water (120 mL) and after 0.5 h stirring EtOH (350 mL) was added and the mixture stirred additional 1 h. Pure product was filtered, washed with EtOH (400 mL) and finally with acetone (300 mL) and dried under vacuum.

1-hydroxybutane-1,1-bisphosphonic acid disodium salt (33.5 g, 85 %) was obtained as a white powder. ¹H NMR (D₂O): δ 1.88–1.79 (m, 2H, CH₂), 1.56–1.47 (m, 2H, CH₂), 0.89 (t, 3H, ³J = 7.4). ¹³C NMR (D₂O, CD₃OD as reference) δ 75.4 (t, ¹J_CP = 130.8, P-C-P), 37.2, 18.2 (t, ²J_CP = 6.3), 15.2. ³¹P NMR (D₂O) δ 18.79. MS (ESI⁻) calcd. for C₄H₁₁O₇P₂⁻ [M-H]⁻ 232.9986, found: 232.9985.

Synthesis of 1-hydroxyhexane-1,1-bisphosphonic acid disodium salt (C5BP)

C5BP was prepared in a similar fashion to compound C3BP, but starting from hexanoic acid (5.0 g, 0.043 mol).

1-hydroxyhexane-1,1-bisphosphonic acid disodium salt (11.0 g, 83 %) was obtained as a white powder.

1H NMR (D₂O): δ 1.92–1.83 (m, 2H, CH₂), 1.57–1.50 (m, 2H, CH₂), 1.36–1.24 (m, 4H, 2 × CH₂), 0.87 (t, 3H, ³J = 7.2).

13C NMR (D₂O, CD₃OD as reference) δ 75.3 (t, ¹J_CP = 132.0, P-C-P), 34.9, 33.1, 24.4 (t, ²J_CP = 5.7), 22.9,14.4. ³¹P NMR (D₂O) δ 18.72. MS (ESI⁻) calcd. for C₆H₁₅O₇P₂⁻ [M-H]⁻ 261.0299, found: 261.0297.

Preparation of gels

Herein, the synthesis of an ETID-loaded gel is described, as an example. All other BP-loaded gels were pre- pared in the same manner. The synthesis of each BP-loaded gel was repeated four (4) times using identical shape and diameter (5 cm inner diameter) borosilicate glass beakers. In a beaker 10 mL of DI water was added. In this a quantity (0.66 g, 3.14 mmol) of sodium metasilicate pentathydrate was dissolved, together with 0.50 g, (1.70 mmol) of tetrasodium ETID, while keeping the solution under stirring. The pH value of this solution was ~12.5. The pH was adjusted to 7.00 with the use 0.75 mL of concentrated HCl (37 %). This particu- lar pH value was selected because the polymerization of silicic acid has the highest rate there. Gel formation commences within 10 min, however the freshly formed and “loose” gel was allowed to mature for 12 h, after

which a shapely and almost translucent gel formed. Gel preparation can be reproducibly repeated and can be modified by altering the amount of Na⁺ ions, replacing the alkali ion, or changing the entrapped BP. BP-con- taining gels for all remaining BPs were prepared in the same manner, using quantities shown in the Table 2.

Controlled release of BPs from gels

On top of the solidified gel (see above), a volume of deionized (DI) water (50 mL), pre-acidified to pH ~ 3 was carefully poured. This marked the initiation of the controlled release process (t = 0), which continued for 48 h. For the initial 6-h period an aliquot of 0.350 mL was withdrawn from the supernatant every hour. After the 6^th h and for the next 12 h, sampling was performed every 3 h. Finally, after the 18^th h and until the end of the release experiment (at the 48^th h) sampling was performed every 8 h. The withdrawn samples were mixed with 0.150 mL of deuterium oxide (99.9 atom % D) that contained 0.05 wt. % (4.3375 μmol) 3-(trimethylsilyl) propionic-2,2,3,3-d₄ acid, sodium salt, TSP) as standard. ¹H NMR spectra were recorded on a Bruker AVANCE 300 MHz NMR (Bruker, Karlsruhe, Germany) spectrometer at 293.2 K operating at a proton NMR frequency of 300.13 MHz. Standard solvent (D₂O) was used as internal lock. Each ¹H spectrum consisted of 32 scans requiring 3 min. and 39 min. acquisition time with the following parameters: spectral width = 20.5671 μs, pulse width (P1) = 15.000 μs, and relaxation delay (D1) = 4.000 s. Polynomial 4^th-order baseline correction was performed before manual integration of all NMR spectra. Proton chemical shifts in D₂O are reported rela- tive to TMS and peaks were integrated according to TSP. The characteristic peaks for each compound were integrated using the integration tool available from the Bruker software (TopSpin 3.2). For each compound we selected the integration value of the sharpest peak. All the integration values were cross-checked in order to ensure the best result for each compound.

Mathematical data treatment

All aliquot samples withdrawn from the supernatant were measured using ¹H, ¹³C, and ³¹P NMR spectroscopy.

The quantification of the released BP was based on integration of peaks in the ¹H spectrum and was made possible by using deuterium oxide 99.9 atom % D as solvent, that contains 0.05 wt. % 3-(trimethylsilyl)propi- onic-2,2,3,3-d₄ acid sodium salt, TSP, as standard. Each release experiment was repeated 4 times in order to achieve maximum reproducibility and satisfactory statistics. Each release experiment consisting of 15 sam- plings/measurements, and repeated 4 times, was treated with the IGOR Pro 6.05 software.

The last step in the mathematical treatment of the results was the creation of a universal curve in the form f(t) = a * e^−b*t + c that depicts the average value of the diffusion of the phosphonate from the silica hydrogel, including standard deviation (Fig. 3).

Factor “a” of the equation is the “frequency factor” and it is related to the entropy difference between the gel and the liquid phase. Factor “b” is the exponential parameter that describes the energy statistical

Table 2: Quantities of bisphosphonates used for hydrogel preparation.

Bisphosphonate Mass (g) mmol

ETID 0.50 1.70

C3BP 0.38 1.36

C5BP 0.48 1.56

PAM 0.29 1.22

ALE 0.37 1.25

C4NBP 0.35 1.31

NER 0.35 1.25

distribution of the molecules through desorption. Constant “c” describes all the remaining interactions between the different phases (diffusion inside the gel, diffusion in the liquid, adsorption etc). The variable

“t” is the time (Table 3).

00 20 40 60

% Release of ETID

80 100

0 20 40 60

% Average release of ETID

80 100

0 20 40 60

% Release of ETID

80 100

0 20 α = –71.447 ± 0.765 40

β = –0.19639 ± 0.00521 c = 73.557 ± 0.526

α = –70.948 ± 1.09 β = –0.19861 ± 0.00692 c = 73.255 ± 0.581 60

% Average release of ETID

80 100

10 20 30

Time (h)

40 50

1^st Gel 2^nd Gel 3^rd Gel 4^th Gel

0 10 20 30

Time (h)

40 50

0 10 20 30

All points from the 4 similar gels

Average with standard deviation

Average with standard deviation

Time (h)

40 50 0 10 20 30

Time (h)

40 50

Fig. 3: Mathematical treatment for ETID samples. Fitting the raw agrees with fitted average quadruplets.

Table 3: Values of a, b, and c (in bold) and their standard deviations in the equation f(t) = a ^* e^−b*t + c, that describe diffusion phenomena for the release of BPs.

Bisphosphonates a b c

ETID(4Na)  − 70.928

( ± 1.02)  − 0.19862

( ± 0.00692) 73.235

( ± 0.581)

C3BP(2Na)  − 52.119

( ± 0.789)  − 0.22025

( ± 0.00735) 53.258

( ± 0.398)

C5BP(2Na)  − 45.92

( ± 0.768)  − 0.18582

( ± 0.00707) 48.284

( ± 0.428)

PAM(acid) 80.10

( ± 1.47) 0.22895

( ± 0.00916) 80.984

( ± 0.724)

ALE(2Na) 70.386

( ± 2.09)  − 0.16825

( ± 0.012) 73.371

( ± 1.24) C4NBP

(acid) 66.746

( ± 1.96)  − 0.22095

( ± 0.00144) 68.54

( ± 0.989)

NER(acid) 70.026

(1.28) 0.20366

(0.0084) 72.527

(0.674)

Synthesis of Ca-ETID

Ca-ETID [Ca(HO₃P)₂C(OH)(CH₃)(H₂O) · 2.5H₂O] was synthesized according to a literature procedure [27].

Tablet preparation for controlled release

Tablets were prepared that contained the ETID drug as active agent (0.250 g, 850.5 μmol of tetrasodium ETID) and three common excipients, cellulose (0.250 g), lactose (0.250 g) and silica (0.250 g). The solid ingredients were mechanically ground in a mortar and pestle, until a free-flowing powdery solid sample was obtained. The powder was then transferred to a hydraulic press that produced the tablet (10 bar pressure).

Those tablets were referred to as the “control” curve in the release diagrams. Also, tablets were prepared that contained Ca-ETID as the active agent in equimolar quantity as the “control” tablet (0.260 g, 850.5 μmol of Ca-ETID) and the same excipients cellulose (0.247 g), lactose (0.247 g) and silica (0.247 g). The total weight of each tablet was 1.000 g.

Quantification of ETID released from tablets

The quantification of ETID release vs. time was carried out with NMR spectroscopy. Specifically, in a 100 mL glass beaker, 50 mL of deionized water were added and the pH was adjusted to 1.3 with hydrochloric acid.

The tablet (as prepared above) was placed in a plastic net and was immersed into the solution just above the stirring bar. Mild stirring was applied to ensure solution homogeneity. The solution was sampled (sample volume 350 μL) every hour for the first 6 h, and every 3 h until the 12^th h, and every 12 h until the 48^th h of the release experiment. After the 48^th h, samples were withdrawn every 24 h or every 48 h or longer, if necessary.

Each sample was placed in a NMR tube, and then the D₂O standard solution (150 μL) was added. The concen- tration of the D₂O TSP standard solution was 4.337 μmol. Quantification of ETID concentration in each sample was achieved by peak (-CH₃) integration in the ¹H NMR spectrum and its comparison to the peak of the TSP standard solution peak [-Si(CH₃)₃].

Synthesis of PPEI polymers

All PPEI polymers were synthesized according to literature procedures [28, 29].

Protocols for the stabilization of silicic acid

These procedures have been published before [12, 30].

Determination of molybdate-reactive silica (silicomolybdate method)

Molybdate-reactive silica (mainly mono- and some disilicic acid) was quantified using the well-established silicomolybdate spectrophotometric method [31]. As in our previous studies, we used the “yellow molyb- date” method (using Spectrophotometer HACH DR/890) as follows: 2 mL from the working solution is filtered through a 0.45 μm syringe filter and then diluted to 25 mL in a special cylindrical cell of 1 cm path length, made of quartz. Next, 1 mL of ammonium molybdate stock solution and 0.5 mL of HCl (1:1 dilution of the concen- trated solution) are added to the sample cell, the solution is shaken well and left standing for 10 min. After- ward 1 mL of oxalic acid solution is added and the cell contents are mixed well. The solution is set aside for 2 min. The photometer is now set to “zero absorbance” using a sample of deionized water (“blank”). Finally,

the sample absorbance is measured (at 452 nm) and is expressed as “ppm SiO₂”. The detectable concentra- tion range for this specific protocol is 6 − 75 ppm. To calculate the concentration in the original solution, an appropriate dilution factor (×27.5/2) is applied. The basic working principal of the silicomolybdate test is that ammonium molybdate reacts only with mono- and disilicic acid and any phosphate present and forms yellow- colored complexes. This reaction requires acidic environment in order to take place, and this is why the hydro- chloric acid is added to the samples. It should be noted that colloidal silica does not participate in the reaction and thus does not affect the intensity of yellow color, which is proportional to the concentration of the reactive silica present in the sample experiment. Oxalic acid is added to destroy any molybdophosphoric acid formed, leaving the silicomolybdate complex intact, and thus eliminating any color interference from phosphates.

Results and discussion

Characterization of BP-loaded hydrogel and dried gel systems

All studied BPs possess the same substituent environment around the central carbon atom (two phosphonate and one hydroxyl moieties), while the R side chain is variable. Hence, the selection criteria were based on fea- tures of R such as polarity, hydrophobicity-philicity, and end-groups. Two groups of BPs have been selected, one that possess a systematically elongated hydrophobic alkyl chain (ETID, C3BP, and C5BP), and the second with the systematically elongated aminoalkyl chain (PAM, ALE, C4NBP and NER). It has been shown in previous studies that the nature of the BP side chain affects drug effectiveness in case of the commercial BP drugs [32].

Three types of hydrogels, “control” (i.e. without BP), “loaded” with BP and “empty” after BP release have been studied by SEM-EDS (after supercritical drying, Fig. 4) and, as expected, they display amorphous fea- tures (by powder XRD). Samples were treated several times and in series with Ethanol/H₂O solutions, 30/70, 50/50, 70/30, 90/10, and finally 100/0, in order to achieve complete dehydration. Then a typical protocol of CPD was followed before SEM studies.

ATR-IR spectra were collected for all “pure” BPs and dried “loaded” gels. Some selective spectra of drug- loaded (ETID, C3BP, and C5BP) dried gels are shown in Fig. 5. The ATR-IR spectra reveal the presence of BPs based on the characteristic peaks of phosphonate group. Although full assignments are difficult, the region of 900–1100 cm⁻¹ can be used as “fingerprint” [33]. Unfortunately, the spectra are complicated by the fact that the Si-O-Si vibrations from the gel appear in the same region [34].

BP-loaded hydrogels were exposed to excess of water and their water uptake/absorption ability has been evaluated. The results (Figs. 6 and 7) prove that they do not absorb additional water, and the synthetic proce- dure followed produces hydrogels that are saturated in water. Their weight differences vary slightly between

~0.5 % and ~3.2 %.

On the other hand, the swelling behavior of dried gels was also studied. Variable behavior was noted. For example, the “empty” (without BP drug) dried gel re-absorbed only ~1.3 % water. The PAM-loaded gel gained

~12.2 % of its weight. In contrast, the ETID-loaded gel lost ~34.6 % of its weight. We assign this weight loss to the hydrophobic nature of the –CH₃ group on the side-chain of ETID.

All hydrogels collapse upon drying. “Empty”, ETID-loaded, and PAM-loaded hydrogels were used for N₂ sorption measurements in order to obtain nitrogen sorption isotherms at 77 K and BET plots. Results of the adsorption studies presented are shown in Fig. 8. The downward trend is consistent with the BP drug par- tially filling the voids within the gel, a fact that is corroborated by the observation that when the BP molecular size increases (from ETID to PAM) the SSA dramatically drops.

Controlled release of BP drugs from hydrogel and dried gel systems

BP-loaded gels were subjected to 48-h controlled release experiments (each repeated at least 4 times, with excellent reproducibility, ±3 % error). ¹H NMR signals were used to quantify the amount of BP released to

Fig. 4: Appearance of the hydrogel (top), representative SEM images (bar = 5 μm) and EDS spectra of ETID-loaded gel before (left) and after (right) BP release (48 h) experiments. The amount of P (in the white circle) in the EDS spectra is dramatically reduced after release, in agreement with the controlled release experiments.

4500 68 70 72 74 76 78 80 82

% Transmittance

84

Dried ETID-loaded gel Dried C3BP-loaded gel Dried C5BP-loaded gel

Dried silica matrix with C5BP Dried silica matrix with C3BP

86 88 90 92 94 96 98 100

4000 3500 3000 2500 2000

Wavenumbers (cm^–1) 1500 1000 500

Fig. 5: ATR-IR spectra of ETID-, C3BP-, and C5BP-loaded dried gels.

the supernatant aqueous solution. Release curves for the family of BPs with hydrophobic side chains (ETID, C3BP, C5BP) are presented in Fig. 9. The BP with the shortest alkyl side-chain (ETID) exhibits the fastest release, followed by C3BP (with a n-propyl side-chain). The BP with longest n-pentyl side-chain (C5BP) demonstrates the slowest release rate. All release experiments reach a plateau after ~20 h, most likely due to saturation of the aqueous phase. The final % BP released in solution after the plateau is reached follows the same ranking, ETID (72 %) > C3BP (54 %) > C5BP (48 %).

Controlled release results for the family of hydrophilic, amine-containing side-chains BPs are quite intriguing (Fig. 10). It is clear that the presence of the amine group on the side chain enhances release rate as well as the final value of % release. The shortest member of the family (PAM) exhibits the fastest release rate and the highest plateau (82 %), a fact that agrees with the previous studies for the non-polar side chain BPs (Fig. 9). Also, side chain elongation decelerates release and lowers the final values. Nevertheless, side- chain length increase beyond three atoms (two C’s, one N) does not induce systematic release reduction, as the results are nearly indistinguishable, thus pointing to a strong effect from the amine end-functionality overwhelming that of the chain lengthening.

The above results (Figs. 9 and 10) that exhibit some important trends for the amino-BPs and non-polar side-chain BPs release rates and final % release can be correlated with their aqueous solubility trends as described by Alanne et al. [24, 35]. Based on the solubility data reported in there, solubility curves can be constructed of the alkyl-BPs (non-polar side-chain) and the aminoalkyl-BPs (polar side-chain) (Fig. 11).

0 ‘‘Empty’’ gel (no

drug) ETID-loaded gel PAM-loaded gel –2.36%

difference Gel weight (g, before swelling) Gel weight (g, after swelling) –3.26%

difference –0.58%

difference

5 10 15

Fig. 6: Water absorption from freshly formed hydrogels. Essentially, there is no swelling.

0 ‘‘Empty’’ gel (no

drug) ETID-loaded gel PAM-loaded gel +12.20%

difference Dry gel weight (g, before swelling) Gel weight (g, after swelling)

–34.65%

difference +1.35%

difference

0.5 10

Gel weight (g)

15

Fig. 7: Water absorption from dried gels.

It is worth-noting that the solubility of alkyl-BPs is ~ two orders of magnitude higher than that of the aminoalkyl-BPs. This may seem counter-intuitive at first glance, because the aminoalkyl-BPs possess an amino end that enhances hydration, and hence, should increase solubility. The explanation for the dramati- cally reduced water solubility of aminoalkyl-BPs lies with their crystal structures. In the crystal structures of PAM [36], ALE [37], and NER [38] acids, the basic amino group is always protonated and participates in a rich network of hydrogen bonds, usually with the phosphonate moieties of neighboring molecules. This creates

30.06 4

4 60

70 80 90 100 110 120 130 140

6 8 10 12 14 16 18 20

5 6 7 8 9 10

0.08 0.10 0.12 0.14 0.16 0.18 BET summary Slope = 41.349 Intercept = 4.222e-01 Correlation coefficient, r = 0.999929 C constant = 98.933

Surface area = 83.4 m²/g

BET summary Slope = 83.547 Intercept = 9.702e-01 Correlation coefficient, r = 0.999965 C constant = 87.114

Surface area = 41.2 m²/g

BET summary Slope = 461.749 Intercept = 9.870e+00 Correlation coefficient, r = 0.999808 C constant = 47.783

Surface area = 7.4 m²/g

Relative pressure, p/p_o

‘‘Empty’’ gel

(SSA = 83.4 m²/g) ETID-loaded gel

(SSA = 41.2 m²/g) PAM-loaded gel

(SSA = 7.4 m²/g)

1/[W((Po/P) – 1)] 1/[W((Po/P) – 1)] 1/[W((Po/P) – 1)]

Relative pressure, p/p_o

Relative pressure, p/p_o

Relative pressure, p/p_o 0.20 0.25 0.24 0.04

0.0 0 20 40 60 80

Volume adsorbed, cm3 (STP) g–1

100 120

Empty ETID PAM 160

140 180

0.2 0.4 0.6 0.8 1.0

0.06 0.08 0.10 0.12 0.14 0.16 0.18 0.20 0.22 0.10 0.12 0.14 0.16 0.18 0.20 0.22 0.24 0.26 0.28

Fig. 8: Adsorption studies for dried gels. Specific surface areas for “empty”, ETID- and PAM-loaded dried gels are also shown.

0 0 20 40 60 80

100 ETID

C3BPC5BP

10 20 30

Time (h)

% BP Release

40 50

Fig. 9: Controlled release of BPs with aliphatic, hydrophobic side chain (ETID, C3BP, and C5BP).

a dense packing resulting in very stable crystal lattice, which is difficult to break with water. In contrast, in the structure of ETID [39] such dense packing is absent, therefore the dissolution effect of water is enhanced.

Unfortunately, crystal structures of C3BP and C5BP have not yet been published.

0 0 20 40 60 80

100 PAM

ALEC4NBP NER

10 20 30

Time (h)

% BP Release

40 50

Fig. 10: Controlled release of BPs with hydrophilic, amine-containing side chain (PAM, ALE, C4NBP, and NER).

0 100 200 300 400 500 600 700 800

0 1 2 3 4 5 6 7 8 9 10 11 12

0 1 2 3 4 5 6 7 8 9 10 11 12

Number of side chain C atoms

0 1 2 3 4 5 6 7 8 9 10 11

Number of side chain C atoms

Solubility of amino-BPs with polar aminoalkyl side chain (g amino-BP/L at 21 °C)

y = 254.92x^–2.848 R² = 0.939

Solubility of BPs with non-polar side chain (g BP/L at 21 °C) y = 10.292x² – 189.41x + 873.68

R² = 0.9961

Fig. 11: Correlation between water solubility and the side-chain length in alkyl-BPs (with non-polar alkyl side-chain, upper) and in aminoalkyl-BPs (with polar, aminoalkyl side-chain, lower). Solubility data for the alkyl-BPs [24] and the aminoalkyl-BPs [35]

were taken from the literature.

Step-wise experiments were designed and carried out to explore the reasons why the equilibrium reached after 24 h, and whether the remaining BP can be quantitatively delivered. The supernatant fluid was replaced with “fresh” aqueous medium after 48 h. The results for three step-wise release stages for the ETID- and PAM-loaded gels strongly support the conclusion that after equilibrium is “reset”, BPs’ release continues eventually reaching a final, essentially quantitative value. Significantly, there is no detectable BP entrapped in potentially inaccessible hydrogel pores, as indicated by the quantitative final release within 144 h.

Reloadability of hydrogels was tested by comparing release curves of freshly prepared and externally loaded (with ETID) gels together with a second BP-loaded gel that was exposed in the three step-wise release process (“used” and completely unloaded gel). This parallel set-up was exposed for 48 h to an aqueous super- natant that contained the same amount of BP (0.500 g of ETID) in order to assess whether the hydrogel will be externally loaded. Results indicate that both gels absorbed ~20 % of supernatant’s dissolved ETID. Then the ETID-“externally” loaded gels were subjected to the usual release conditions, delivering ~ 60 % of the externally absorbed ETID (Fig. 12). Release curve profiles in both cases revealed an almost identical to the

“default” silica matrix behavior in terms of release rate and final % BP release. Even though the “re-loadabil- ity” may not seem to be necessary in drug/pharmaceutical applications, the above results prove that both gels (“freshly-prepared” and “used”, after release) are robust and re-loadable. This conclusion adds an interesting property that may be useful for environmental applications e.g. sustainable aquatic anti- pollutant systems.

Hydrogel density is another factor that can assist in a priori “programming” the hydrogels release behav- ior. Initially, it had been observed that the concentration of sodium silicate in gel preparation affects the macroscopic properties of the gel (color, brittleness). Gels with two and four times higher sodium silicate concentration than the initial (default gel) were prepared and BP drug release studies were performed in each case. Figure 13 depicts the clear differences in release rates and final % release between three gels of variable density, i.e. 6.66 %, 13.32 % and 26.64 % in sodium silicate.

The drug release characteristics of drug-loaded dried gels were also assessed. Preliminary studies were performed on dried ETID- and PAM-loaded gels. The prepared hydrogels (see Experimental Section) were placed in an oven overnight at 100 °C. The isolated solids were milled and a manual hydraulic press was used to transform them into pellets (applied pressure 15 tons). Then, the pellets were placed into plastic nets, which were immersed carefully into the top part of the supernatant (50 mL de-ionized water, pre-acidified at pH = 3). Gentle magnetic stirring was applied to ensure homogenous sampling conditions. The same sam- pling protocol was followed for the construction of the release curves, as the one for the hydrogels.

Drug release data from dried gels (pellets) reveal that ETID and PAM demonstrate faster release than that from the “authentic” hydrogels (Fig. 14). Moreover, PAM follows faster release than ETID, in agreement with the studied hydrogels. In both cases a plateau is reached after the 9^th h. It is worth-noting that the dried gel

0 0 5

% BP release 10 15 20 25 30

10 20 30

Time (h)

‘External loading of ETID in a new silica matrix’

‘External loading of ETID in a used silica matrix’

40 50

Fig. 12: Release profiles of re-absorbed ETID from a freshly-prepared gel and from a used (emptied) gel.

systems released the active BP drugs 10 h earlier than the hydrogels. Also, besides fast release, the dried gel systems release almost 100 % of the loaded drug after the plateau is reached.

Finally, the described BP-drug-loaded hydrogels are injectable (Fig. 15). This opens interesting possibili- ties in the technology of drug delivery, a theme that is currently explored in our laboratory.

Controlled release of ETID from APTES-modified silica hydrogels

Silica hydrogels were modified with the grafting agent (3-aminopropyl)triethoxysilane (APTES). The hypoth- esis put forth was that the presence of the –NH₂ moiety (actually protonated at the conditions of the experi- ment) would create additional ionic and/or hydrogen bonding interactions with the anionic phosphonate groups of ETID, and thus further delay the release. The same synthetic procedure was followed as for the ETID- loaded hydrogels, with the difference that APTES was added to the mixture at an APTES:silicate 1:10 molar ratio. These APTES-modified hydrogels were left to mature for 4 days in order to ensure complete grafting.

After 4 days, ¹H NMR spectra of the supernatant aqueous phase did not show any APTES present, a strong indication that grafting was quantitative. The APTES-modified ETID-loaded gels were subjected to the same controlled release experiment as the “authentic” silica hydrogels. Comparative results are shown in Fig. 16.

Initial drug release rates are identical among the two evaluated hydrogels (up to the 5^th h). Afterwards, a distinct difference in release rates is observed, with the APTES-modified silica hydrogel demonstrating a

“delay” effect, and finally a lower % release than the “authentic” silica hydrogel. It is likely that the identical

0 2 4 6 8

ETID in 6.66 %, pH = 7 (default) ETID in 13.32 %, pH = 7 ETID in 26.64 %, pH = 7

10 12

0 10 20 30 40 50 60 70

% BP release

Time (h)

Fig. 13: Comparison of the drug release curves for the “default (6.66 % in sodium silicate) and two more dense gels (13.32 % and 26.64 % in sodium silicate) during the first 12 h of drug release.

00 20 40 60

BP release

80 100

0 20 40 60

BP release

80 100

10 20 30

Time (h)

40 50

ETID hydrogel in 6.6 % pH~7

Dried ETID in 6.6 % pH~7 PAM hydrogel in 6.6 % pH~7

Dried PAM in 6.6 % pH~7

0 10 20 30

Time (h)

40 50

Fig. 14: Comparison of drug release curves of dried gel systems pellets for ETID and PAM with hydrogels.

initial release rates observed are due to the fact that loosely “bound” ETID molecules are released first in both systems. After their release, the remaining ETID molecules are those that interact with the amino group of APTES, and demonstrate a delayed release.

Controlled release of ETID from Ca-ETID tablets

Figure 17 presents a comparison between the controlled release of ETID from “control” and “Ca-ETID” tablets.

ETID release from the “control” tablet (containing “free” ETID and excipients only) is rapid, reaching 100 % in the first 10 h. A plateau is reached, which is maintained until the end of the release experiment (240 h).

ETID release from Ca-ETID-containing tablets is much slower. For comparison, it reaches only ~30 % after 10 h, whereas release from the “control” tablet is quantitative at the same sampling time. Interestingly, a plateau is reached after ~190 h, which corresponds to 90 % total release. It is noteworthy that replacing the supernatant solution with “fresh” solution allows the controlled release to continue and reach 100 %.

Fig. 15: Injectability of BP-loaded hydrogels.

0 10 20 30

Time (h)

40 50

NH₂ O Si

O O APTES

Silica hydrogel with ETID and APTES Silica hydrogel with ETID

0 20 40 60

% BP release 80

100

Fig. 16: Comparison of the ETID drug release curves for the APTES-modified, and “authentic” silica hydrogels.

The profound differences in the rate and the total release between the “free” and Ca-ETID tablets can be rationalized based on the structure of the Ca-ETID compound. The structure of Ca-ETID [27] can be described as 1D chain architecture. The coordination environment of the Ca²⁺ center, the chelation mode of ETID, and the structure of a single 1D chain in Ca-ETID along the a axis are shown in Fig. 18. The Ca²⁺ center is 7-coordi- nated by five phosphonate oxygens from three different ETID ligands, one –OH group and a water molecule.

Fig. 18: Coordination environment of the Ca²⁺ center (upper left), chelation mode of ETID (upper right), and structure of a single 1D chain in Ca-ETID along the a-axis (lower). Color codes: Ca green, C black, P orange, O red. Waters of crystallization are depicted as blue spheres.

0 0 20 40 60

% Release

80 100 120

50 100 150 200

Time (h)

250

Ca: ETID Control ETID

Fig. 17: Controlled release (%) of ETID from “control” (dark blue circles) and “Ca-ETID” (light blue triangles) tablets.

One ETID ligand acts as a tridentate chelate (two phosphonate oxygens and one –OH), one acts as a biden- tate chelate (two phosphonate oxygens) and the third as a monodentate ligand (one phosphonate oxygen).

Each ETID exists as a dianion in the structure (balancing the charge of the Ca²⁺) and coordinates to three Ca²⁺

centers. The bridging ability of ETID creates a “double chain” that runs along the a-axis. There are 2.5 waters of crystallization (depicted as blue spheres in Fig. 18) that hold the chains together via a rich network of hydrogen bonds.

The structure of tetrasodium ETID, used for the release experiments as the “control”, is unknown, but one can assume that the four Na⁺ ions in the lattice are coordinated by the phosphonate oxygens. Once the crystal lattice is exposed to the low pH of the solution surrounding the tablet, a combined action of hydration and hydrolysis of the Na–O bonds occurs. The weak and principally ionic Na-O bonds are then ruptured, thus releasing free ETID. The structure of Ca-ETID, on the other hand, contains strong Ca-O bonds that are much more resistant to hydrolysis, due to the high affinity of ETID for Ca²⁺ cations [40]. We assign the slower release of ETID from Ca-ETID tablets to the resistance of the Ca-ETID dissolution via hydrolysis.

Silica scale inhibition by phosphorus-containing polymers

For this part of the work, we used the cationic polymer PEI and its phosphonomethylated derivative, the zwit- terionic polymer PPEI (Fig. 2) as inhibitors for colloidal silica formation. Their inhibitory efficiency was tested in “long-term” (up to 3 days) and “short term” (8 h) experiments. Long term experiments are an important time frame for scale inhibitors, as it evaluates the ability of a certain inhibitor to prevent scale formation for a prolonged time period. Short term experiments examine the silica polycondensation in its early stages. First, useful comparisons can be made between the non-phosponated PEI and the fully (100 %) phosphonated PPEI, see Fig. 19. It is noted that the inhibitory efficiency of both polymeric inhibitors is dosage-dependent, but in a dramatically different way. For PEI, as inhibitor dosage increases, inhibition drops. In contrast, the inhibitory efficiency of PPEI is directly proportional to dosage increase, but reaches a plateau beyond 60 ppm.

The inhibition superiority of PPEI compared to that of PEI is clearly shown in Fig. 20, where the extra stabilization (PPEI over PEI) is plotted vs. inhibitor dosage. A slight decrease is noted for the 80 ppm dosage.

This is an indication that at high inhibitor concentrations polymer entrapment into the colloidal silica matrix occurs. This has been observed previously for other polymeric inhibitors [12, 13, 41–43].

The results presented in Figs. 21 and 22 establish that by grafting anionic phosphonate moieties onto the PEI backbone, thus creating the zwitterionic PPEI polymer, profoundly increases dramatically the inhibitory activity. However, this is true for the fully grafter PPEI (100 % phosphonomethylation). Hence, in order to evaluate the effect of the extent of phosphonometylation on inhibitory activity, a series of partially grafted PPEI’s were studied. For direct comparison reasons, the 60 ppm dosage was studied for all inhibitors after 8 h

0 25 50 75 100

0 (control) 20 40 60 80

Silicic acid stabilization (%)

Inhibitor dosage (ppm) PEI

PPEI (full)

Fig. 19: Comparison of the dosage-dependent inhibitory % efficiency of PEI and PPEI (full). The graph refers to the 24-h measurements.

0 50 100 150 200 250

0 20 40 60 80 100

PPEI-over PEI-stabilized silica (ppm)

Inhibitor dosage (ppm)

Fig. 20: Difference (in ppm, after control-subtraction) of silicic acid stabilization between PPEI (full) and PEI. The graph refers to the 24-h measurements.

0 10 20 30 40 50 60 70 80 90 100

30 40 50 60 70 80 90 100

Silicic acid stabilization (%)

Phosphonomethylation degree (%)

Fig. 21: Short-term effect of PPEI % phosphonomethylation degree on silicic acid inhibitory activity, at 60 ppm inhibitor concentration, after 8 h.

0 25 50 75 100

Control PPEI (40 %) PPEI (60 %) PPEI (80 %) PPEI (full)

Silicic acid stabilization (%)

Phosphonomethylation degree (%) 24 h48 h

72 h

Fig. 22: Long-term effect of PPEI % phosphonomethylation degree on silicic acid inhibitory activity, at 80 ppm inhibitor concentration, at 24, 48, and 72 h. Lines are added to aid the reader.

of condensation time. These results are shown in Fig. 21. There appears to be an almost linear dependence of inhibitory activity on % grafting degree. The maximum inhibition performance is achieved when PPEI is fully (100 %) substituted, offering 80 % inhibitory efficiency.

The effect of phosphonomethylation degree was also evaluated for longer time periods, up to 3 days.

The results are shown in Fig. 22. It appears that all phosphonomethylated PEIs (from 40 to 100 %) are able to inhibit the formation of colloidal silica for extended polycondensation times. It is obvious that time nega- tively affects inhibitory efficiency, hence, the inhibitors are only able to delay but not stop the silicification process. Nevertheless, silicic acid levels are well above the control, particularly for the PPEI (full) inhibitor.

It demonstrates a ~75 % efficiency at 24 h, ~63 % after 48 h, and ~52 % after 72 h. It is therefore an efficient long-term silica inhibitor.

Conclusion

In this paper we discussed selected aspects of phosphorus chemistry:

(a) The controlled release of BP osteoporosis drugs from several drug delivery systems, such as silica hydro- gels, dried gels, APTES-grafted hydrogels, and metal phosphonate hybrid coordination networks.

(b) The silica inhibitory activity of phosphonomethylated polyethyleneimine (PPEI).

To our knowledge, there are no reports in the literature on BP release from silica-based hydrogels. However, a limited number of other BP controlled release systems have been reported. Mesoporous silica was used for confinement and controlled release of bisphosphonates [44]. Nanocomposite hydrogels based on hya- luronic acid and self-assembled bisphosphonate-magnesium (BP-Mg) nanoparticles were reported [45].

Calcium-Zoledronate complexes were used as a potential nanocarrier for pDNA delivery [46]. ETID release from Na- and Ca-ETID impregnated silicone matrices has been reported [47]. Spherical Zoledronate-Ca@

DOPA composites (DOPA = 1,2-dioleoyl-sn-glycero-3-phosphate monosodium salt) were dispersed in 1,2-dis- tearoyl-sn-glycero-3-phosphoethanolamine-N-[methoxy(polyethyleneglycol)-2000, and the pH-dependent controlled release of Zoledronate was evaluated [48]. The therapeutic effects of Ga- and Gd-alendronate were examined in the rat model of restenosis [49]. First row transition metal complexes (M = Co²⁺, Mn²⁺, Ni²⁺) with ibandronate showed antiparasitic activity [50]. Although a direct comparison of these drug release systems with our systems reported herein is not possible, these reports emphasize the intense interest of the scientific community in the fabrication of effective BP controlled release systems.

Our results reported herein add to the rich milieu of phosphorus chemistry, with specific applications in pharmaceutics and industrial water treatment.

Acknowledgments: KDD thanks the EU for funding the Research Program SILICAMPS-153, under the ERA.

NET-RUS Pilot Joint Call for Collaborative S&T projects.

References

[1] D. E. C. Cordbridge. Phosphorus: Chemistry, Biochemistry and Technology, CRC Press, Boca Raton, USA (2013).

[2] J. Shen, L. Yuan, J. Zhang, H. Li, Z. Bai, X. Chen, W. Zhang, F. Zhang. Plant Physiol. 156, 997 (2011).

[3] J. Kruse, M. Abraham, W. Amelung, C. Baum, R. Bol, O. Kühn, H. Lewandowski, J. Niederberger, Y. Oelmann, C. Rüger, J. Santner, M. Siebers, N. Siebers, M. Spohn, J. Vestergren, A. Vogts, P. Leinweber. J. Plant Nutr. Soil Sci. 178, 43 (1999).

[4] N. Leroy, E. Bres. Eur. Cell Mater. 2, 36 (2001).

[5] M. Bauer, W. E. Klee. Eur. J. Mineral. 5, 307 (1993).

[6] S. V. Dorozhkin. Biomatter 1, 121 (2011).

[7] S. V. Dorozhkin, M. Epple. Angew. Chem. Int. Ed. 41, 3130 (2002).

[8] Z. Amjad, D. Morgan. Phosphorus Res. Bull. 25, 33 (2011).

[9] A. Clearfield, K. D. Demadis. Metal Phosphonate Chemistry: From Synthesis to Applications. Royal Society of Chemistry, London (2012).

[10] S. Monge, B. Canniccioni, A. Graillot, J.-J. Robin. Biomacromolecules 12, 1973 (2011).

[11] K. D. Demadis, A. Tsistraki, A. Popa, G. Ilia, A. Visa. RSC Adv. 2, 631 (2012).

[12] K. D. Demadis, K. Pachis, A. Ketsetzi, A. Stathoulopoulou. Adv. Coll. Interf. Sci. 151, 33 (2009).

[13] K. D. Demadis, A. Ketsetzi, K. Pachis, V. M. Ramos. Biomacromolecules 9, 3288 (2008).

[14] K. E. Poole, J. E. Compston. Br. Med. J. 344, e3211 (2012).

[15] M. Pazianas, B. Abrahamsen, S. Ferrari, R. G. G. Russell. Ther. Clin. Risk Manag. 9, 395 (2013).

[16] K. A. Kennel, M. T. Drake. Mayo Clin. Proc. 84, 632 (2009).

[17] K. E. Papathanasiou, P. Turhanen, S. I. Brückner, E. Brunner, K. D. Demadis. Sci. Rep. 7, Article number: 4743 (2017).

[18] E. Antonogiannakis, E. Tzagkaraki, K. D. Demadis. Int. J. Corros. Scale Inhib. 2, 216 (2013).

[19] E. Antonogiannakis, E. Tzagkaraki, K. D. Demadis. Int. J. Corros. Scale Inhib. 2, 255 (2013).

[20] E. Neofotistou, K. D. Demadis. Int. J. Corros. Scale Inhib. 3, 28 (2014).

[21] A. Spinthaki, A. Stathoulopoulou, K. D. Demadis. Int. J. Corros. Scale Inhib. 4, 85 (2015).

[22] A. Spinthaki, A. Stathoulopoulou, K. D. Demadis. Int. J. Corros. Scale Inhib. 4, 125 (2015).

[23] A. Spinthaki, J. Matheis, W. Hater, K. D. Demadis. Energy Fuels 4, 11749 (2018).

[24] A.-L. Alanne, H. Hyvönen, M. Lahtinen, M. Ylisirniö, P. Turhanen, E. Kolehmainen, S. Peräniemi, J. Vepsäläinen. Molecules 17, 10928 (2012).

[25] G. Kieczykowski, R. Jobson, D. Melillo, D. Reinhold, V. Grenda, I. Shinkai. J. Org. Chem. 60, 8310 (1995).

[26] H. E. Gottlieb, V. Kotlyar, A. Nudelman. J. Org. Chem. 62, 7512 (1997).

[27] F. Niekiel, N. Stock. Cryst. Growth Des. 14, 599 (2014).

[28] R. R. Navarro, S. Wada, K. Tatsumi. J. Hazard. Mater. 123, 203 (2005).

[29] R. R. Navarro, S. Wada, K. Tatsumi. Sep. Sci. Technol. 38, 2327 (2003).

[30] A. Spinthaki, G. Petratos, J. Matheis, W. Hater, K. D. Demadis. Geothermics 74, 172 (2018).

[31] K. D. Demadis, M. Preari. Des. Wat. Treat. 55, 749 (2015).

[32] R. G. G. Russell. Bone 40, S21 (2007).

[33] I. Cukrowski, L. Popović, W. Barnard, S. O. Paul, P. H. van Rooyen, D. C. Liles. Bone 41, 668 (2007).

[34] P. Innocenzi. J. Non-Cryst. Solids 316, 309 (2003).

[35] A.-L. Alanne, M. Tuikka, K. Tõnsuaadu, M. Ylisirniö, L. Hämäläinen, P. Turhanen, J. Vepsäläinen, S. Peräniemi. RSC Adv. 3, 14132 (2013).

[36] L. M. Shkol’nikova, S. S. Sotman, E. G. Afonin. Kristallografiya 35, 1442 (1990).

[37] J. Ohanessian, D. Avenel, D. El Manouni, M. Benramdane. Phosphorus Sulfur Silicon Relat. Elem. 129, 99 (1997).

[38] V. M. Coiro, D. Lamba. Acta Cryst. C C45, 446 (1989).

[39] V. A. Uchtman, R. A. Gloss. J. Phys. Chem. 72, 1298 (1976).

[40] M. L. Lamson, J. L. Fox, W. I. Higuchi. Int. J. Pharmaceutics 21, 143 (1984).

[41] K. D. Demadis, E. Neofotistou. Chem. Mater. 19, 581 (2007).

[42] M. Preari, K. Spinde, J. Lazic, E. Brunner, K. D. Demadis. J. Am. Chem. Soc. 136, 4236 (2014).

[43] K. D. Demadis. J. Chem. Technol. Biotechnol. 80, 630 (2005).

[44] F. Balas, M. Manzano, P. Horcajada, M. Vallet-Regí. J. Am. Chem. Soc. 128, 8116 (2006).

[45] K. Zhang, S. Lin, Q. Feng, C. Dong, Y. Yang, G. Li, L. Bian. Acta Biomater. 64, 389 (2017).

[46] W. Li, X. Xin, S. Jing, X. Zhang, K. Chen, D. Chen, H. Hu. J. Mater. Chem. B 5, 1601 (2017).

[47] G. Golomb, M. Dixon, M. S. Smith, F. J. Schoen, R. J. J. Levy. Pharm. Sci. 76, 271 (1987).

[48] X. Li, Y. W. Naguib, Z. Cui. Int. J. Pharmaceut. 526, 69 (2017).

[49] H. Epstein, V. Berger, I. Levi, G. Eisenberg, N. Koroukhov, J. Gao, G. J. Golomb. J. Control. Release 117, 322 (2007).

[50] B. Demoro, S. Rostán, M. Moncada, Z. H. Li, R. Docampo, C. Olea Azar, J. D. Maya, J. Torres, D. Gambino, L. Otero.

J. Biol. Inorg. Chem. 23, 303 (2018).

Konstantinos E. Papathanasiou, Maria Vassaki, Argyro Spinthaki, Fanouria-Eirini G. Alatzoglou, Eleftherios Tripodianos, Petri Turhanen and Konstantinos D.

Demadis

Phosphorus chemistry: from small mole- cules, to polymers, to pharmaceutical and industrial applications

https://doi.org/10.1515/pac-2018-1012 Pure Appl. Chem. 2018; x(x): xxx–xxx

Conference paper:

Novel drug delivery systems for controlled release of bisphosphonate anti-osteoporosis drugs are based on easy-to-manufacture silica hydrogels, and metal-phosphonate (drug) coordination networks.

Keywords: controlled release; hydrogels;

osteoporosis;

phosphonates; POC- 2018; polymers; scale inhibitors; silica scale;

water treatment.

Graphical abstract