Preparation and solid state NMR characterization of phosphonates encapsulated in raw

and organically modified SBA-15

Thierry Azais

, Daniela Aiello

, Flaviano Testa

, Guillaume Laurent

, and Florence Babonneau

Laboratoire Chimie de la Matière Condensée de Paris, Université Paris 6 – CNRS, UMR 7574

Collège de France, Paris, 75005, France.

Department of Chemical Engineering and Materials, CR-INSTM, University of Calabria,

87036 Arcavacata di Rende (CS), Italy.

ABSTRACT

We present in this communication the preparation and the solid state NMR

characterization of phenyl phosphonic acid encapsulated both in pure and aminopropyl-modified

SBA-15 mesoporous silica materials. The

P and

H MAS studies revealed two radically

different behaviors of the confined molecules. The included phosphonic acid in SBA-15 is

submitted to a confinement effect that implies a weak interaction with the SiO

surface and a

relative mobility at room temperature. On the contrary, phenyl phosphonic acid molecules in the

aminopropyl modified sample possess a strong interaction with the hybrid surface of the

material. This finding is supported by a two dimensional double-quantum

H experiment that

revealed the close proximity between phenyl phosphonic acid and aminopropyl surface groups.

INTRODUCTION

Ordered mesoporous silica matrices are a class of materials highly attractive in reason of

a high porous volume associated to a high specific area (up to 1000 m

-1

). Since 2001, their

textural characteristics are widely used in order to store and release drug molecules [1].Up to

now many studies have been done on different micelles templated silica materials such as MCM-

41 or SBA-15 [2]. In particular, it has been shown that a functionalization of the pore surface can

provide a control of the drug release kinetic [3]. Recently, similar drug delivery systems have

been designed with alendronate, a drug that belongs to the class of bisphosphonates that inhibit

bone resorption and consequently help fight osteoporosis. It has been shown that the

modification of the siliceous pore surface with aminopropyl groups allows to optimize the

amount of the encapsulated drug and to tune drug release kinetic [4]. The kinetic properties of

these drug release systems seem to be linked to the physical state of the confined molecules and

its chemical interaction with the pore surface [5]. Nevertheless, the relationship between release

properties and chemical properties is still not well understood and for this reason we choose to

study phenyl phosphonic acid as a model-molecule for bisphosphonates, confined in pure and

aminopropyl-modified SBA-15. We characterized these materials by means of solid state NMR

which is a perfect tool to differentiate different physical state of a drug and also to probe host-

guest interactions [6]. In particular we studied in details the inclusion of phenyl phosphonic acid

through

P and

H MAS NMR experiments that revealed different physical behavior for the

guest molecules depending on the chemical nature of the porous surface of the host material.

Finally, we showed that host-guest proximities can be efficiently explored using a two-

dimensional

H double quantum experiment recorded at high magnetic field (B

= 16.4 T)

combined to an ultra high fast MAS (65 kHz) in order to recover a good spectral resolution.

EXPERIMENTAL DETAILS

Synthesis of raw and organically modified SBA15 and phenyl phosphonic acid loading

The preparation of the silica-based SBA-15 material was similar to the method described

by Zhao et al. [7]. The template solution was prepared dissolving 4.0 g of Pluronic

P123

(Fluka) in 104 mL of deionized water and 20 mL of 37 wt% HCl with stirring at 35°C. Then,

9.16 mL of TEOS were added into the solution under magnetic stirring for 12 hours at room

temperature. The molar composition of final gel mixture was TEOS/HCl/P123/H

O =

1:6.03:0.017:145. The white precipitate was aged at 100°C for 24 hours, then filtered, washed

with deionised water and dried at 60°C for 12 hours. The free template material (named SBA)

was obtained through a calcination of the surfactant molecules at 550°C in 8 hours in air (heating

rate of 10°C/min). The amine-functionalized material (named SBA-NH) was synthesized

through a post-synthesis grafting. For this purpose 1 g of the free template mesoporous materials,

previously activate in an oven at 120°C for 2 hours, was mixed with 1.33 mL of APTES

((C

Si(CH

)

; Aldrich) in 20 mL of toluene (Fluka) under magnetic stirring. The

mixture was heated under reflux at 110°C overnight and the product was filtered, washed with

toluene and dried at 100°C overnight. The encapsulation process was performed by the incipient

wetness procedure [8], after a activation at 100°C for 12 hours of the free template and the

amine-functionalized samples. Then, 0.500 g of the powders was soaked by a solution of

phenylphosphonic acid ΦP (C

PO(OH)

; Fluka; 0.012 g cm

-3

) in ethanol. This procedure was

repeated successively three times, and after each impregnation, the solvent was removed by

heating at 70°C overnight. After the third impregnation, the samples were washed with 4 ml of

ethanol in order to remove the excess of phenyl phosphonic acid. The samples were then dried in

oven at 70°C overnight. The free template and the amine-functionalized samples loaded with

phenyl phosphonic acid will be referred in the forthcoming text as SBA-ΦP and SBA-NH-ΦP,

respectively, where ΦP stands for phenyl phosphonic acid.

Solid state NMR characterization

H, and

P solid state NMR experiments were performed on a Bruker Avance III 300

spectrometer operating at ν

(

H) = 300.30 MHz, and ν

(

P) = 121.49 MHz (B

= 7 T) using a 4

mm Bruker MAS probe. The spinning rate ν

MAS

was 14 kHz for

H and

P MAS experiments

and 5 kHz for

P CP MAS experiments. The two-dimensional 2D double quantum DQ

experiment was performed on a Bruker Avance III 700 spectrometer (B

= 16.4 T) operating at

(

H) = 700.20 MHz using a 1.3 mm Bruker ultra fast MAS probe (ν

MAS

= 65 kHz).

RESULTS AND DISCUSSION

Loading and mesoporous characteristics

Nitrogen adsorption/desorption isotherms of our samples were found to be of type IV

according to the IUPAC classification, with a H1 type hysteresis loops characteristic of

mesoporous solids. A S

BET

of 766 m

-1

and 355 m

-1

for the free-template SBA sample and

the amine-functionalized SBA-NH sample, respectively, were calculated using the BET model

[9]. Moreover, a narrow distribution of pores was observed with an average pore diameter D

60 Å and 54 Å for SBA and SBA-NH, respectively that were determined with the BJH (Barrett-

Joyner-Halenda) method based on the desorption branch of the isotherms [10]. Once loaded with

phenyl phosphonic acid a substantial reduction of the S

BET

is found for both samples. We

measured a specific surface area of 425 m

-1

and 250 m

-1

for SBA-ΦP and SBA-NH-ΦP,

respectively. These values correspond to a S

BET

decrease of 45 % and 30 %, respectively, when

compared to the initial values of unloaded samples. These data suggests an efficient

encapsulation of ΦP molecules into the mesoporous framework. An amount of 322 g and 380 g

of phenyl phosphonic acid molecules per gram of porous materials for SBA-ΦP and SBA-NH-

ΦP, respectively, was determined both by micro-analysis and ATD-TG. The higher amount

found for SBA-NH-ΦP could be explained by a greater affinity of ΦP molecules with the

aminopropyl group rather than with the raw silica surface of the SBA-15 material. All the

textural characteristics of our samples and the amount of loaded phenyl phosphonic acid are

summarized in Table 1.

BET

/g)

(cm

/g)

(Å)

Isotherm

type

Amount of phosphonic

acid (mg/g)

SBA 766 0.48 60 IV

SBA-ΦP

425 0.43 57 IV 0.322

SBA-NH 355 0.43 54 IV

SBA-NH-ΦP

250 0.31 49 IV 0.380

Table 1: Textural characteristics of loaded and unloaded samples.

Solid state NMR

P MAS NMR spectroscopy was used here to study the inclusion of ΦP in the

mesoporous materials (Figure 1a). The

P MAS spectrum of the bulk phenyl phosphonic acid

displays a single resonance at 21.0 ppm, whereas the

P MAS spectra of the two loaded samples

show two main resonances located at 19.1 and 13.8 ppm for SBA-ΦP and SBA-NH-ΦP,

respectively. It is well known that the condensation of a phosphate or phosphonate moiety leads

to an up-field shift of the

P resonance [11]. Thus, the fact that the phenyl phosphonic acid

resonances of in our loaded samples are up-field shifted could be an indication of a specific

interaction of ΦP with the pores surface organically modified or not. Moreover, we note that the

P resonance in the amino-functionalized sample is highly shifted compared to SBA-ΦP.

Following the previous principle, the ΦP-surface interaction seems to be stronger with the

aminopropyl functionalization rather than with raw SiO

surface. Furthermore, we note that the

line width of the

P resonance of SBA-NH-ΦP is particularly broad and is 10 times larger than

the one of SBA-ΦP (10.4 ppm vs. 1.0 ppm). This is due to a wide distribution of chemical shifts

that could be explained by the fact that the ΦP/aminopropyl interaction leads to a disordered

interface. Finally, we note that a small peak at 9.4 ppm (30% of the main resonance) and a weak

shoulder centered at 6.3 ppm (15 % of the main resonance) are visible on the SBA-ΦP and the

SBA-NH-ΦP spectrum, respectively. These latter signals certainly correspond to two

phosphonate moieties condensed with each other through a P-O-P bond.

The Figure 1b displays the intensity variation of the

P resonances during a variable

contact time

H-

P CP MAS experiment for SBA-ΦP and SBA-NH-ΦP.

Figure 1: a)

P NMR spectra of bulk phenyl phosphonic acid ΦP, SBA-ΦP and SBA-NH-ΦP;

b) intensity of the main

P resonances as a function of a variable contact time during a

H-

CP MAS experiment for SBA-ΦP and SBA-NH-ΦP.

The two curves are fitted with the following equation that expresses the signal intensity as a

function of two time constants [12]:

31P

) = (γ

/γ

31P

(1-λ)

-1

[-exp(-(1-λ)t

)].exp(-t

1ρ

(

H))

where t

corresponds to the contact time, M

corresponds to the

P Zeeman magnetization, λ =

1ρ

(

H) where T

stands for the cross relaxation time constant and T

1ρ

(

H) is the relaxation

time of protons in the rotating frame. We found two radically different CP dynamic behaviors for

the two loaded samples. The

P signal in SBA-NH-ΦP has a faster dynamic than in SBA-ΦP as

we found T

= 0.66 ms and 2.22 ms, respectively. The first value is typical of bulk phenyl

phosphonic acid which has already been determined [13]. The slowest behavior found for SBA-

ΦP can be explained either by reduced source of protons compared to SBA-NH-ΦP or by a

higher mobility of ΦP in SBA-ΦP. The first assumption is unlikely because even if the acid

groups POH in ΦP are deprotonated when encapsulated, the phenyl group is still an efficient

source of proton magnetization. Indeed, previous measurements done on deuterated phenyl

phosphonic acid (C

PO(OD)

) lead to a T

value of 0.8 ms far away from the long 2.2 ms

value found for SBA-ΦP [13]. This long

H-

P cross relaxation time constant can only be

explained by a higher mobility of the ΦP molecules at room temperature that averages the

H-

heteronuclear dipolar interaction. This higher mobility compared to bulk ΦP and ΦP

encapsulated in SBA-NH is certainly due to the so-called confinement effect that acts on small

molecules confined in mesopores [14]. This effect has already been evidenced on carboxylic

acids such as benzoic or lauric acid confined in MCM-41 [15] where the liquid-solid phase

transitions of such molecules were deeply depressed compared to the bulk. Furthermore, we note

that the proton relaxation time in the rotating frame is also very different for the two samples.

We found T

1ρ

(

H) = 1.79 ms for SBA-NH-ΦP whereas no T

1ρ

(

H) effect was detected in the

range 0-10 ms for SBA-ΦP (T

1ρ

(

H) >> 10 ms). The longer T

1ρ

(

H) found for SBA-ΦP is also a

sign of a higher mobility of FP molecules in SBA-ΦP compared to SBA-NH-ΦP [16].

H NMR spectroscopy is also an excellent probe for molecules mobility studies as shown

on Figure 2a.The

H MAS spectrum of bulk ΦP displays a broad signal corresponding mainly to

-10-50510152025303540

ΦΦ

ΦP

SBA -Φ

ΦΦ

ΦP

SBA-NH-Φ

ΦΦ

ΦP

SBA-NH-ΦP

SBA-ΦP

Contact time (ms)

Intensity (a. u.)

(ppm)

0.1

0.2

0.3

0.4

0 2 4 6 8 10 12

the proton resonances of the phenyl ring and the POH groups [17]. The

H MAS spectrum of

SBA-NH-ΦP displays also a broad signal dominated by the proton resonances of encapsulated

ΦP and aminopropyl groups. These two broad spectra are characteristics of rigid compounds

where the

H-

H homonuclear dipolar coupling dominates all other interactions. On the contrary,

the

H MAS spectrum of SBA-ΦP displays thin resonances that correspond to the phenyl ring of

encapsulated ΦP. These unusual narrow resonances are due to the average of the

H-

homonuclear dipolar interaction in reason of the relative mobility of ΦP molecules in SBA-ΦP.

This result confirms the presence of confinement effect for phenyl phosphonic acid confined in

SBA-15. Moreover, we note that the resonances of POH groups (expected around 12 ppm) are

absent on the latter spectrum in reason of a chemical exchange at room temperature [5].

Figure 2: a)

H MAS NMR spectra of ΦP, SBA-NH-ΦP and SBA-ΦP recorded at ν

MAS

= 14

kHz and Β

= 7 Τ b) 2D MAS DQ spectrum of SBA-NH-ΦP recorded at ν

MAS

= 65 kHz and

= 16.4 Τ (the

H MAS NMR spectra of SBA-NH-ΦP recorded at ν

MAS

= 14 kHz and Β

= 7 Τ

is also shown for comparison).

The precise characterization of the interface between mesoporous materials surface and

confined species is usually a hard task. Indeed, when the confined molecule is submitted to a

confinement effect like in the case of ΦP in SBA-ΦP material, its relative mobility averages out

the dipolar couplings that inform on nuclei proximities. On the other hand, when the SiO

surface is organically modified, the interaction between the confined species and the hybrid

surface can be much stronger and lead to a rigid system. The direct consequence is a lack of the

spectral resolution for the

H spectra that became hard to analyze like in the case of SBA-NH-

ΦP. A strategy to enhance this spectral resolution is to increase the static magnetic field B

and

the ν

MAS

. The Figure 2b gives an excellent illustration to that statement. We recorded a 2D MAS

H spectrum of SBA-NH-ΦP at ν

MAS

= 65 kHz and Β

= 16.4 T with the BABA sequence

which is based on the

H-

H dipolar interaction [18]. The gain in resolution is impressive when

compared to the equivalent spectrum recorded at ν

MAS

= 14 kHz and Β

= 7 T. Two distinct

signals are now visible instead of one broad envelope. Three proton resonances centered at 1.5

ppm on the single quantum SQ dimension can be safely assigned to CH

groups of the

aminopropyl function. The resonances centered at 7 ppm correspond to phenyl protons of

included ΦP. Interestingly, strong double quantum cross peaks are clearly evidenced between the

phenyl and the three distinct CH

protons revealing the close proximity between ΦP and the

propyl group. This result strengthens our previous assumption i.e. the existence of a strong

(ppm)

-20-1001020304050

ΦΦ

ΦP

SBA-NH-Φ

ΦΦ

ΦP

SBA-Φ

ΦΦ

ΦP

(ppm)

01020 -10

νν

MAS

= 14 kHz

= 7 T

νν

MAS

= 65 kHz

= 16.4 T

interaction between ΦP and the aminopropyl group of the surface-grafted organic function in

SBA-NH-ΦP. This strong interaction could be present as H-bonding (POH...NH

) or ionic

interaction (PO

...NH

CONCLUSION

H and

P MAS NMR experiments revealed two distinct behaviors for the guest

molecules whether ΦP is encapsulated in pure SBA-15 or in amino-modified SBA-15. In the

first case, ΦP is submitted to a confinement effect that implies a relative mobility at room

temperature and a weak interaction with the siliceous surface. In the second case, ΦP is rigid and

presents a strong interaction with the amino surface groups. Furthermore, we showed that a high

magnetic field (B

= 16.4 T) combined to an ultra high fast MAS (65 kHz) allows an impressive

gain in proton spectral resolution. The close proximity between ΦP and aminopropyl surface

groups is then evidenced by a 2D

H DQ BABA experiment.

REFERENCES

1. Vallet-Regi, M.; Ramila, A.; del Real, R. P.; Pérez-Pariente, J. Chem. Mater. 13, 308 (2001).

2. Vallet-Regi, M.; Balas, F.; Arcos, D. Angew. Chem. Int. Ed. 46, 7548 (2007).

3. Andersson, J.; Rosenholm, J.; Areva, S.; Lindén, M. Chem. Mater. 16, 4160 (2004).

4. F. Balas, M. Manzano, P. Horcajada, and M. Vallet-Regi, J. Am. Chem. Soc. 128, 8116

(2006)

5. Azaïs T.; Tourné-Pétheil C.; Aussenac F.; Baccile N.; Coelho C.; Devoisselle J-M.;

Babonneau F. Chem. Mater. 18, 6382 (2006).

6. Bonhomme, C.; Coelho, C.; Baccile, N.; Gervais, C.; Azaïs, T.; Babonneau, F. Accounts

Chem. Res. 40, 738 (2007).

7. D. Zhao, J. Feng, Q. Huo, N. Melosh, G.H. Fredrickson, B.F. Chmelka, G.D. Stucky, Science

279, 548 (1998).

8. C. Charnay, S. Begu, C. Tourne-Peteilh, L. Nicole, D.A. Lerner, J.-M., Devoisselle, Eur.

Pharm. Biopharm. 57, 533 (2004).

9. S. Brunauer, P.H. Emmett, E.J. Teller, J. Am. Chem. Soc. 60, 309 (1938)

10. P. Barrett, L.G. Joyner, P. Halenda, J. Am. Chem. Soc. 73, 373 (1951)

11. M. M. Crutchfield, C. F. Callis, R. R. Irani, G. C. Roth, Inorg. Chem., 1, 4, 813 (1962).

12. A. Pines, M. Gribby, J. Waugh, J. Chem. Phys. 59, 569 (1973).

13. G. P. Holland, R. Sharma, J. O. Agola, S. Amin, V. C. Solomon, P. Singh, D. A. Buttry and

J. L. Yarger, Chem. Mater. 19, 2519 (2007)

14. Alba-Simionesco, C.; Coasne, B.; Dosseh, G.; Dudziak, G.; Gubbins, K E.; Radhakrishnan

R.; Sliwinska-Bartkowiak, M. J. Phys.: Condens. Matter 18, R15 (2006).

15. Azaïs T.; Hartmeyer, G.; Quignard, S.; Laurent, G.; Tourné-Péteilh, C.; Devoisselle, J-M.;

Babonneau, F. Pure Appl. Chem. 81, 1345 (2009).

16. K. Schmidt-Rohr, H. W. Spiess, « Multinuclear NMR in Solids and Polymers », Acad.

Presse, 1996.

17. C. Gervais, M. Profeta, V. Lafond, C. Bonhomme, T. Azaïs, H. Mutin,C. J. Pickard, F.

Mauri, F. Babonneau, Magn. Reson. Chem. 42, 445 (2004).

18. Schnell, I.; Spiess, H. W. J. Magn. Reson. 151, 153 (2001).