Rapid adsorption and entrapment of benzoic acid molecules onto

Rapid adsorption and entrapment of benzoic acid molecules ontomesoporous silica FSM-16 Keiji Yamamoto Graduate School of Pharmaceutical Sciences, Chiba University, 1-33 Yayoicho, Inage-k

Rapid adsorption and entrapment of benzoic acid molecules onto

mesoporous silica (FSM-16)

Keiji Yamamoto

Graduate School of Pharmaceutical Sciences, Chiba University, 1-33 Yayoicho, Inage-ku, Chiba 263-8522, Japan

Received 14 March 2005; accepted 6 May 2005 Available online 13 June 2005

Abstract

Changes in the molecular state of benzoic acid (BA) in the presence of folded sheet mesoporous material (FSM-16), which has uniformly sized cylindrical mesopores and a large surface area, were assessed with several analyses When BA was blended with FSM-16 for 5 min (BA content= 30%), the X-ray diffraction peaks of BA crystals disappeared, suggesting an amorphous state Fluorescence analysis of the

mixture showed a new fluorescence emission peak for BA at 386 nm after mixing with FSM-16 Fluorescence lifetime analysis of the BA component in the mixture at 386 nm showed a longer lifetime in comparison with that of BA crystals The solid-state13C CP/MAS and PST/MAS NMR spectra of the mixture with FSM-16 showed a significantly different spectral pattern from the mixture with nonporous glass, whose NMR spectra were identical to those of BA crystals These results indicate that BA molecules disperse quickly into the hexagonal channels of FSM-16 by a simple blending procedure and adsorbed BA molecules had clearly different physicochemical properties to BA crystals

2005 Elsevier Inc All rights reserved

Keywords: Porous material; Adsorption; Solid-state NMR; Fluorescence spectra; Benzoic acid; FSM-16

1 Introduction

To improve the dissolution profile of poorly water-soluble

compounds, we have investigated the use of porous

ma-terials as pharmaceutical excipients, and have already

re-ported that medicinal substances could be adsorbed onto

the surface of porous material, resulting in the change of

their physicochemical properties [1] A variety of organic

compounds can be adsorbed onto the surface of porous

materials due to their peculiar adsorptive properties [2,3]

Folded sheet mesoporous material (FSM-16), mesoporous

silica having one-dimensional pores with uniformly sized

mesopores of cylindrical-like nature, has high porosity and

* Corresponding author Fax: +81 43 290 2939.

E-mail address:ytozuka@faculty.chiba-u.jp (Y Tozuka).

1 Department of Pharmacy, University of Yamanashi, 1110 Tamaho-cho,

Nakakoma-gun, Yamanashi 409-3898, Japan.

large surface area (more than 1000 m2/g) FSM-16 is

com-posed of honeycomb-like hexagonal channels and is used

as highly promising model adsorbents for fundamental ad-sorption studies[4] FSM-16 has been shown to maintain its texture at pressures up to 780 kg cm−2[5] and this would allow the material to be used for grinding and tableting, un-like other porous materials These remarkable characteristics give rise to the possibility of FSM-16 for pharmaceutical use

To regulate the quality of drug products in pharmaceuti-cal dosage forms, the molecular states of the medicine and the molecular interactions between the medicine and any ad-ditives in the dispersed system should be investigated from the viewpoint of controlling the stability and safety of medi-cine Assessment using high-sensitivity and high-resolution analyses is necessary for the clarification of the molecular state in the solid dispersion Fluorescence spectroscopy is

a useful tool for detecting molecules even in low concentra-0021-9797/$ – see front matter  2005 Elsevier Inc All rights reserved.

doi:10.1016/j.jcis.2005.05.009

tions due to its high sensitivity[6] Fluorescence analysis has

been applied to solid systems in order to study molecular

in-teractions, such as inclusion phenomena of cyclodextrin and

guest molecules[7], and to study physicochemical changes

of organic compounds adsorbed on solid surfaces[8]

Solid-state13C NMR spectroscopy is also a useful method for the

study of materials in the solid state Combinational

tech-niques of cross-polarization (CP) and magic-angle spinning

(MAS) provide high-resolution13C spectra, showing

molec-ular level information[9] Since solid-state13C NMR

meth-ods simply differentiate between mobile and immobile

con-tributions, these methods provide detailed information about

the mobility[10]

In the present study, we used benzoic acid as a model

compound and FSM-16 (pore width: 2.1 nm) as a model

porous material The molecular state of benzoic acid

ad-sorbed onto the surface of porous material was investigated

by using solid-state fluorescence spectroscopy, fluorescence

lifetime analysis, and solid-state13C NMR spectroscopy

2 Materials and methods

2.1 Materials

Benzoic acid (BA; Nacalai Tesque, Kyoto, Japan) of

reagent grade was used without further purification

Meso-porous silica FSM-16 (mean pore diameter of 2.1 nm,

spe-cific surface area of 1250 m2/g) was kindly supplied from

Toyota Central R&D Labs Inc., Japan FSM-16 was sieved

using a 200-µm aperture size sieve and was used after

dry-ing under a reduced pressure at 110◦C for 3 h Glasperlen, a

nonporous glass powder, from 0.25 to 0.35 mm in radius,

was purchased from B Braun Melsungen AG, Germany

Physical mixtures were prepared by blending BA and

ad-ditives in a glass vial for definite intervals

2.2 Powder X-ray diffractometry

Powder X-ray diffraction was performed using a Rigaku

Miniflex diffractometer (Tokyo, Japan) The measurement

conditions were as follows: target, Cu; filter, Ni; voltage,

30 kV; current, 15 mA; scanning speed, 2◦/min.

2.3 Thermal analysis

A thermogravimetry-differential thermal analysis

(TG-DTA) was carried out using a MAC Science TG-DTA 2000S

(Japan) at a heating rate of 5◦C min−1under a nitrogen gas

flow of 60 ml/min.

2.4 Fourier transform infrared (FT-IR) spectroscopy

Fourier transform infrared spectra were measured by the

KBr disk method at a resolution of 2 cm−1for 32 scans using

a JASCO 230 FT-IR spectrophotometer (Tokyo, Japan)

2.5 Fluorescence spectroscopy

An FP-770F fluorescent spectrometer (Japan Spectros-copy Co., Ltd., Tokyo, Japan) was used for stationary flu-orescence spectroscopy Powder samples were filled into a front-face reflectance cell (FP-1060)

2.6 Determination of the fluorescence lifetime and relative quantum yield

Fluorescence decay profiles were measured by a nanosec-ond time-resolved single-photon counter with a pulse width

of 1.5 ns (Horiba NAES-770, Tokyo, Japan) The excit-ing pulse and emission response functions were measured simultaneously, and the decay parameters were calculated from two or three exponential functions obtained by de-convolution of the excitation pulse profile using nonlinear least-squares fitting The goodness of fit was assessed by

monitoring the value of χ2 and the distribution of

residu-als χ2 values below 1.4 indicate acceptable results for the fluorescence lifetime analysis

2.7 Solid-state NMR spectroscopy

13C NMR spectra were determined on a JNM-LA400 NMR spectrometer (JEOL, Japan) operating at 100.4 MHz with a CP/MAS (cross-polarization/magic-angle spinning) probe The sample (ca 190 mg) was contained in a cylindri-cal rotor made of ceramic materials, and spun at 6000 Hz

A 90◦ pulse width was about 5.5 µs for both 13C and 1H under CP conditions The contact time was 5 ms, and the rep-etition times were 60 and 5 s in the CP/MAS and PST/MAS (pulse saturation transfer/magic-angle spinning), respec-tively In the PST/MAS NMR technique, NOE enhance-ment is used to obtain the13C signal This technique en-hances peak intensity for mobile carbon[11].13C chemical shifts were calibrated indirectly through external adaman-tane (29.5 ppm relative to TMS) The spectral width and the number of data points were 40 kHz and 16,000, respectively The number of accumulations was 2000 in the CP/MAS and PST/MAS experiments at room temperature Experi-mental conditions were as follows: temperature, 25.0◦C;1H decoupling field amplitude, 50 kHz; rf field amplitude for cross-polarization, 50 kHz

3 Results and discussion

The mixture of benzoic acid and FSM-16 (BA content: 30%) was blended in a glass vial for different periods of time The changes in powder X-ray diffraction (XRD) pat-terns of the mixture as a function of mixing time are pre-sented inFig 1 The X-ray diffraction peaks of BA crystals decreased in intensity with duration of mixing time and dis-appeared completely after blending for 5 min, indicating the disappearance of an ordered arrangement of molecules in the

Fig 1 Powder X-ray diffraction patterns of the 30% benzoic acid

(BA)–70% FSM-16 system after mixing for different intervals: (a) 30 s,

(b) 60 s, (c) 300 s.

Fig 2 Solid-state fluorescence emission spectra of the 30% BA–70%

FSM-16 system after storage at 25 ◦C, λex = 262.7 nm (a) 0 min, (b) 3 min,

(c) 15 min, (d) 1 h, (e) 9 h, (f) 72 h.

crystals Since FSM-16 had large specific surface area and

hydrophobic features[12], it was assumed that BA

mole-cules showed a rapid adsorption on mixing with FSM-16

To investigate the adsorption profiles of BA on FSM-16,

BA and FSM-16 (BA content: 30%) were mixed by a

spat-ula for approximately 5 s and then placed in the sample

holder for fluorescence spectroscopy measurement.Fig 2

shows the solid-state fluorescence emission spectra of 30%

BA and 70% FSM-16 system as a function of storage time

Two emission peaks were observed at 317 and 386 nm The

Table 1

Fluorescence lifetime (τ ) and relative quantum yield (Q) of benzoic acid in various systems, λex = 262.7 nm

λobs

(nm)

τ1

(ns)

Q1

(%)

τ2

(ns)

Q2

(%)

χ2

BA crystals 317.0 0.454 15.2 2.58 84.8 1.30

30% BA–70% Glasperlen 317.0 0.120 29.4 2.76 70.6 1.24

30% BA–70% FSM-16 386.0 0.099 44.6 6.57 55.4 1.36

emission peak at 317 nm, which was identical to that in BA crystals, gradually decreased in intensity with an increase

in storage time On the other hand, the intensity of a new emission peak at 386 nm increased with storage time accom-panied by the disappearance of the emission intensity of BA crystals at 317 nm The new emission at 386 nm was not observed when BA crystals were blended with nonporous glass or other pharmaceutical additives, i.e., microcrystalline cellulose, glucose, lactose, mannitol, or sucrose Such an anomalously large Stokes shift might be explained by either the excited state with intramolecular proton transfer or the excimer-like emission of benzene moieties of BA molecules Denisov et al reported that the anomalously large Stokes shift, found in fluorescence emission spectra of molecules like salicylic acid and its derivatives, was characterized by a strong intramolecular hydrogen bond and the low-frequency band attributed to the excited state with intramolecular pro-ton transfer[13] As it is impossible for benzoic acid to form

an intramolecular hydrogen bond due to its structure, the anomalously large Stokes shift is due to another reason Aro-matic compounds like benzene, naphthalene, and pyrene are known to form van der Waals dimers and to show excimer emission with large Stokes shifts[14] For excimer emission,

it was reported that excimer formation occurs more easily

in small pores than that in large pores [15,16] The mean pore width of FSM-16 used was estimated as 2.1 nm, which might be suitable for BA molecules to form parallel or T-shaped van der Waals dimers The new emission peak might

be due to the excimer emission resulting from van der Waals dimer-like contact of adjacent BA molecules on the FSM-16 surface

Fluorescence-decay kinetics was investigated for BA-FSM-16 mixtures of different dispersibility The lifetime and relative quantum yield are listed inTable 1 The component

of a short lifetime (τ1) was estimated as a noise

compo-nent[8] The fluorescence lifetimes of BA observed in the Glasperlen (nonporous glass) system were similar to those observed in BA crystals When the observation wavelength was fixed at 317.0 nm, it was difficult to determine the fluorescence lifetime in the mixture of 30% BA and 70%

FSM-16, as χ2 values could not be converged due to low fluorophore concentration On the other hand, when the ob-servation wavelength was fixed at the new emission peak

at 386 nm, the lifetime of the second component was esti-mated as 6.57 ns, which was long enough compared to that observed in BA crystals This result also supported that BA

Fig 3 FT-IR spectra of the 30% BA–70% FSM-16 system: (a) BA

crys-tals, (b) physical mixture of 30% BA–70% FSM-16 mixed for 5 min,

(c) FSM-16.

molecules change drastically their molecular states during a

simple blending with FSM-16

To investigate the molecular interaction between BA and

FSM-16, IR measurement was performed (Fig 3) The

ab-sorption band observed at 1685 cm−1in the spectrum of BA

crystals was assigned to the carbonyl-stretching vibration,

where BA molecules formed a centrosymmetric

hydrogen-bonded dimer in the crystal[17] Infrared spectrum of the

physical mixture of 30% BA and 70% FSM-16 showed a

new peak at 1702 cm−1, indicating a dramatic change in the

molecular state of BA during a mixing Sinhal et al reported

that BA showed a carbonyl-stretching band at 1712 cm−1

when BA molecules were dispersed and interacted with the

“CH2OH” group of ethyl cellulose[18] We propose that the

centrosymmetric hydrogen-bonded dimer in the BA crystals

breaks and a hydrogen bond may form between the carbonyl

groups of the adsorbed BA molecules and the surface of

FSM-16 The thermogravimetry curves of the 30% BA–70%

Glasperlen and the 30% BA–70% FSM-16 systems are

il-lustrated inFig 4 The weight loss from the mixture of 70%

Glasperlen was observed from 90 to 150◦C, which was due

to the sublimation of BA molecules On the other hand, the

weight loss of BA from the mixture with FSM-16 was

sig-nificantly suppressed The suppression of BA sublimation

indicated that BA molecules must be trapped in the pore of

FSM-16 by the surface interaction between the surface of

FSM-16 and the BA molecules Although strong evidence

was not found for the interaction between the silanol groups

Fig 4 Thermogravimetric curves of (a) 30% BA–70% Glasperlen (non-porous glass) and (b) 30% BA–70% FSM-16.

Fig 5 Solid-state 13C CP/MAS spectra of (a) BA crystals, (b) 30% BA–70% Glasperlen, (c) 30% BA–70% FSM-16.

and carbonyl groups in IR spectra, suppression of BA sub-limation may be due to its interaction with the FSM-16 surface However, it should be pointed out that there must

be many BA molecules that are not interacting with silanol groups It is known that hysteresis is observed in adsorption and desorption isotherms of organic gases on mesoporous material This is based on the affinity of the adsorbate to the surface of the mesoporous material and its surface geome-try[19,20] Thermogravimetry indicates a weight loss of BA from FSM-16 during desorption Therefore, suppression of

BA sublimation might be due to a similar mechanism

Fig 6 Solid-state13C CP/MAS and PST/MAS spectra of BA in the 5-min mixture with FSM-16: (a) BA crystals, (b) 30% BA–70% FSM-16, (c) 10% BA–90% FSM-16.

The solid-state 13C CP/MAS spectra of BA, 30% BA–

70% Glasperlen, and 30% BA–70% FSM-16 are shown in

Fig 5 All NMR peaks were attributed to the carbon atoms

of BA molecules, as both Glasperlen and FSM-16 have no

carbon atoms in their structure The chemical shift observed

at 172 ppm (•) in (a) and (b) was derived from the

car-bonyl carbon of BA molecules Whilst the peak positions

in 30% BA–70% Glasperlen were identical to those of BA

crystals, the resonance shifted to a higher magnetic field

at 165.5 ppm (•) after mixing with FSM-16 Although this

broad resonance is not clear and the magnitude of this shift

was considered to be small, since the relaxation time of

car-bonyl carbon was comparatively long, it might be related to

the interactions of some BA molecules with the FSM-16

sur-face The peaks between 60 and 70 ppm were attributed to

the spinning sidebands of benzene carbons of BA molecules

due to the chemical-shift anisotropy The disappearance of

spinning sidebands in the BA/FSM-16 sample indicate that

BA crystals do not exist in this sample

Fig 6shows the solid-state13C CP/MAS and PST/MAS

spectra of BA–FSM-16 system with different BA contents

Pulse saturation transfer is a technique using a nuclear

Over-hauser effect to obtain the 13C signal, resulting in an

en-hancement of the13C magnetizations by saturated carbons

The PST/MAS method has been reported to be useful for

emphasizing the signal intensity of carbons in the flexible

regions[21] A negligible peak of BA crystals was observed

in the PST/MAS spectrum, because a repetition times of 5 s

was insufficient to relax the carbon atoms in BA crystals

The physical mixtures of BA and FSM-16 clearly showed

the chemical shifts arising from both the carbonyl carbon

at 172 ppm and the benzene carbons at around 130 ppm

The signals of benzene carbons were very sharp and their

spinning side bands were not observed, despite that the spin-ning side bands were generally observed due to the chemical shift anisotropy No spinning side bands in the spectra sug-gested that the chemical-shift anisotropy is being averaged

by high molecular mobility This is similar to the fast

mo-tion of p-nitroaniline molecules after heat treatment with

FSM-type mesoporous silica as reported by Komori and Hayashi[22]

With respect to the signal from the carbonyl carbons, the mixture showed a sharp peak at 172 ppm in the 13C PST/MAS spectrum The magnitude of signal intensity from benzene and carbonyl carbons increased with the increase of the concentration of BA, showing a significant amount of

BA molecules adsorbed on the FSM-16 surface with higher mobility For adsorption profiles of organic compounds on a mesoporous structure, monolayer adsorption, multilayer ad-sorption, and/or capillary condensation can occur depending

on the amount of adsorbent and texture of the mesoporous material Previous work has shown that acetonitrile adsorbed onto mesoporous materials forms hydrogen bonds with the surface hydroxyl groups and further adsorption takes place through physisorption [23] In the case of mixtures with FSM-16, the resonance enhancement observed by PST/MAS might arise from weakly adsorbed BA molecules (i.e., hy-drophobic interaction) or the formation of multiple layers

As described in the fluorescence studies, a mixture of 30%

BA and 70% FSM-16 shows the two emission peaks at 317 and 386 nm The anomalously large Stokes shift of BA ob-served at 386 nm was obob-served in the long lifetime com-ponent of the fluorescence decay curve as well The above new emission and the sharp carbon signals in13C PST/MAS spectra were especially strong for the mixture of 30% BA and 70% FSM-16, indicating that the BA molecules with

high molecular mobility could be related to the BA

mole-cules that showed the new fluorescence emission peak

4 Conclusion

Blending with FSM-16 drastically changes the

molecu-lar state of BA BA molecules adsorbed onto the FSM-16

surface showed high molecular mobility of carbon in terms

of the relaxation time, leading to a sharp carbon resonance

in the 13C PST/MAS spectra The state of adsorbed BA

molecules could be determined by solid-state fluorescence

spectroscopy, fluorescence lifetime analysis, and solid-state

13C NMR spectroscopy For pharmaceutical formulation, it

has been generally recognized that the molecular state of

or-ganic compounds strongly affects the dissolution properties

or a stability of organic compound Hence the detection and

evaluation of molecular states of pharmaceutical active

in-gredients are important for the control of their properties

This study reveals that solid-state13C NMR spectroscopy is

a useful technique for estimating molecular states of

phar-maceutically active ingredients in mesoporous additives

Acknowledgments

The authors thank Toyota Central R&D Labs., Inc.,

Japan, for the kind provision of FSM-16 We also thank

Dr H Seki for her valuable assistance with the solid-state

NMR applications This study was supported in part by a

Grant-in-Aid for Scientific Research from the Ministry of

Education, Culture, Sports, Science, and Technology, Japan

(17790029)

References

[1] Y Tozuka, T Oguchi, K Yamamoto, Pharm Res 20 (2003) 926 [2] P Pendleton, S.H Wu, J Colloid Interface Sci 266 (2003) 245 [3] H Rong, Z Ryu, J Zheng, Y Zhang, J Colloid Interface Sci 266 (2003) 245.

[4] D Gao, Z Shen, J Chen, X Zhang, Micropor Mesopor Mater 67 (2004) 159.

[5] T Ishikawa, M Matsuda, A Yasukawa, K Kandori, S Inagaki, T Fukushima, S Kondo, J Chem Soc Faraday Trans 92 (1996) 1985 [6] J.H Lee, S.W Jung, I.S Kim, Y.I Jeong, Y.H Kim, S.-H Kim, Int J Pharm 251 (2003) 23.

[7] E Junquera, E Aicart, Int J Pharm 1761 (1999) 169.

[8] Y Tozuka, E Yonemochi, T Oguchi, K Yamamoto, Bull Chem Soc Jpn 73 (2000) 1567.

[9] G.V Mooter, M Wuyts, N Blaton, R Busson, P Grobet, P August-ijns, R Kinget, Eur J Pharm Sci 12 (2001) 261.

[10] C Mayer, G Lukowski, Pharm Res 17 (2000) 486.

[11] H Yoshimizu, H Mimura, I Ando, J Mol Struct 246 (1991) 367 [12] A Matsumoto, T Sasaki, N Nishimiya, K Tsutsumi, Colloids Surf.

A Physicochem Eng Aspects 203 (2002) 185.

[13] G.S Denisov, N.S Golubev, V.M Schreiber, S.S Shajakhmedov, A.V Shurukhina, J Mol Struct 381 (1996) 73.

[14] H Hirata, H Ikeda, H Saigusa, J Phys Chem A 95 (1999) 1014 [15] T Fujii, A Ishi, N Takusagawa, H Yamashita, M Anpo, J Pho-tochem Photobiol 86 (1995) 219.

[16] Y Tozuka, E Tashiro, E Yonemochi, T Oguchi, K Yamamoto, J Col-loid Interface Sci 248 (2002) 239.

[17] G.A Sim, J.M Robertson, T.H Goodwin, Acta Crystallogr 8 (1955) 157.

[18] R Singhal, A.K Nagpal, G.N Mathur, J Thermal Anal Cal 58 (1999) 29.

[19] S Inagaki, Y Fukushima, Micropor Mesopor Mater 21 (1998) 667 [20] M McNall, R.L Laurence, W Curtis Conner, Micropor Mesopor Mater 44–45 (2001) 709.

[21] H Yasunaga, I Ando, J Mol Struct 301 (1993) 129.

[22] Y Komori, S Hayashi, Micropor Mesopor Mater 68 (2004) 111 [23] H Tanaka, T Iiyama, N Uekawa, T Suzuki, A Matsumoto, K.K Unger, K Kaneko, Chem Phys Lett 293 (1998) 541.