Using terahertz time-domain spectroscopical technique to monitor cocrystal formation between piracetam and 2,5-dihydroxybenzoic acid

Using terahertz time-domain spectroscopical technique to monitor cocrystal formation between piracetam and 2,5-dihydroxybenzoic acid

Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 111 (2013) 192–195 Contents lists available at SciVerse ScienceDirect Spectrochi...

389KB Sizes 0 Downloads 45 Views

Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 111 (2013) 192–195

Contents lists available at SciVerse ScienceDirect

Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy journal homepage: www.elsevier.com/locate/saa

Using terahertz time-domain spectroscopical technique to monitor cocrystal formation between piracetam and 2,5-dihydroxybenzoic acid Yong Du ⇑, Yi Xia, Huili Zhang, Zhi Hong Centre for THz Research, China Jiliang University, Hangzhou 310018, China

h i g h l i g h t s

g r a p h i c a l a b s t r a c t

 The THz absorption spectra of 40

Absorption coefficient (a.u.)

piracetam and its pharmaceutical cocrystals are reported.  The cocrystal formation is monitored from both time and frequencydomain THz spectra.  THz-TDS is potential to directly characterize the pharmaceutical solid-state reactions.

Physical Mixing 1 min 4 mins 13 mins Piracetam+2,5DHBA 26 mins grinding 43 mins 69 mins Cocrystal 90 mins

1.24

0.84

20

0.72

0 0.6

0.8

1.0

1.2

1.4

Frequency (THz)

a r t i c l e

i n f o

Article history: Received 16 January 2013 Received in revised form 11 March 2013 Accepted 16 March 2013 Available online 2 April 2013 Keywords: Cocrystal Active pharmaceutical ingredient (API) 2,5-Dihydroxybenzonic acid (2,5-DHBA) Piracetam Grinding Terahertz time-domain spectroscopy (THz-TDS)

a b s t r a c t Far-infrared vibrational absorption of cocrystal formation between 2,5-dihydroxybenzoic acid (2,5DHBA) and piracetam compounds under solvent evaporation and grinding methods have been investigated using terahertz time-domain spectroscopy (THz-TDS) at room temperature. The experimental results show large difference among absorption spectra of the formed cocrystals and the involved individual parent molecules in 0.20–1.50 THz region, which probably originated from the intra-molecular and inter-molecular hydrogen bonds due to the presence of two hydroxyl groups in 2,5-DHBA and amide moieties in piracetam compound. The THz absorption spectra of two formed cocrystals with different methods are almost identical. With grinding method, the reaction process can be monitored directly from both time-domain and frequency-domain spectra using THz-TDS technique. The results indicate that THz-TDS technology can absolutely offer us a high potential method to identify and characterize the formed cocrystals, and also provide the rich information about their reaction dynamic process involving two or more molecular crystals in situ to better know the corresponding reaction mechanism in pharmaceutical fields. Ó 2013 Elsevier B.V. All rights reserved.

Introduction Cocrystals, which are formed between an active pharmaceutical ingredient (API) and a cocrystal former, have been gaining increasingly interest and recognized as an attractive alternative for solid drug forms to improve many pharmaceutically relevant performances in the pharmaceutical community [1,2]. Pharmaceutical cocrystals show unique physicochemical properties compared to the single-component crystals, such as solubility, dissolution rate, ⇑ Corresponding author. Tel./fax: +86 571 86875618. E-mail address: [email protected] (Y. Du). 1386-1425/$ - see front matter Ó 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.saa.2013.03.081

hydration stability and bioavailability [3,4]. The API and cocrystal former interact via non-ionic and non-covalent intermolecular interactions such as van der Waals forces and hydrogen bonding. Thus, the presence of unused hydrogen bond donor and acceptor sites is a prerequisite factor in the formation of cocrystals. Piracetam is a nootropic drug which is widely used for the treatment of memory and balance problems [5]. Containing two different amide moieties [6], piracetam is a suitable model pharmaceutical compound for the investigation of cocrystal formation with the compound containing hydrogen bond donor (carboxylic acid or hydroxyl groups) groups, such as 2,5-dihydroxybenzoic acid (2,5DHBA) in this study.

193

Y. Du et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 111 (2013) 192–195

Solid-state grinding and slow solvent-evaporation of an API and a cocrystal former are two mainly important ways for cocrystal preparation. Traditional methods include Fourier transformationinfrared (FT-IR) spectroscopy [6], Raman vibrational spectroscopy [7], X-ray power diffractormetry (XRPD) [6–10] and differential scanning calorimetry (DSC) [6,10] have been employed to characterize the cocrystal formation process. These currently used analytical techniques are sometimes time consuming and also lack in sufficient and direct information about the weak intermolecular interactions. It is of great importance to use specific technique to probe the cocrystal characteristics of APIs in order to directly monitor the reaction dynamic process involving two or more molecular crystals in-line to better know the corresponding cocrystal formation mechanism in pharmaceutical industry. Terahertz time-domain spectroscopy (THz-TDS) is a vibrational spectroscopic technique which uses electromagnetic radiation between millimeter radio and infrared light waves and the range typically covers 0.1–10.0 THz (3–333 cm 1). THz spectral response of molecular system at low frequencies yields rich information about collective molecular vibrations and intermolecular skeletal modes in whole molecular systems [11–15]. Many crystal materials usually have much more distinctive and sharper spectral features in the THz frequency region than in the mid-IR and NIR regions, which makes them easier to be characterized. These low-frequency modes are strongly dependent on the intermolecular interaction of the whole molecular structure, and they provides ‘‘fingerprint’’ information of crystal structures. Another advantage of this technique for analytical purpose is that THz radiation is safe to samples and human-bodies due to the non-ionization nature, which is especially crucial to the biological and pharmaceutical systems. Recently THz-TDS technique has been used to monitor and investigate the solid-state polymorphism [16], crystallinity [17], mechanochemical solid-state reactions [18–20], and so on. THz-TDS is attractive as a means to probe and characterize the weak noncovalent intermolecular interaction [16–21], and has been demonstrated to be a high potential alternative to mainstream analytical tools such as XPRD and DSC in solid-state identification and quantification [17,18], especially process monitoring and control in pharmaceutics-related research fields. In this study, terahertz vibrational absorption of cocrystals obtained between API piracetam and cocrystal former 2,5-DHBA compounds under solvent evaporation and grinding methods have been investigated using THz-TDS technique. The experimental results show large difference among absorption spectra of the formed cocrystals and the involved parent molecules in 0.20– 1.50 THz region, which mainly originated from the intra-molecular and inter-molecular interactions between 2,5-DHBA and piracetam parent molecules. The THz absorption spectra also indicate that two formed cocrystals with different methods are almost identical. With grinding method, the real reaction process of the cocrystal formation could be monitored directly from both time-domain and spectra-domain features using THz-TDS technique. The results indicate that THz-TDS technology can not only give us a new attractive experimental method to identify and characterize the formed cocrystals, but also provide a promising tool for further monitoring the real reaction dynamic involving two or more molecular crystals in-line to better know the corresponding reaction mechanism from molecule-level in pharmaceutical industry.

Experimental methods Chemicals and sample preparation The 2,5-DHBA and piracetam samples were purchased from Sigma–Aldrich (structures shown in Fig. 1) and used without further

O O

OH

NH2 OH O

N HO

Piracetam

2,5-DHBA

Fig. 1. Chemical structures of piracetam and 2,5-dihydroxybenzoic acid (DHBA) molecules.

purification. The API piracetam and cocrystal former 2,5-DHBA were ground before mixing to achieve several micrometer particle size. Physical mixture was obtained by gently mixing two compounds above at a 1:1 molar ratio in a glass vial by using a vortex mixer during 10 mins. Solvent cocrystal was prepared by slow solvent evaporation method. Equimolar piracetam and 2,5-DHBA were dissolved in amount of acetonitrile at ambient temperature. The solution was slowly evaporated at room temperature during several days to produce needle-like crystal. Grinding cocrystal was performed by co-milling piracetam with 2,5-DHBA at 1:1 molar ratio in 25 mL stainless steel milling jars using a planetary ball mill (QM-3SP, gear type, Nanjing University Instrument Plant) with a frequency of 25 Hz at room temperature. A measured amount of product produced in the solid-state reaction at different milling times were taken out for further spectra analysis. All the samples were weighted into 150 mg aliquots and mixed with 150 mg polyethylene (PE) powder. Then the 300 mg mixture was poured into a steel die and subjected to 4 MPa pressure for around 2 mins. The resulting 13 mm in diameter, 1.5 mm thick sample discs were obtained and sealed in plastic before analysis.

Apparatus and procedure The THz beam is produced by a Ti:Sapphire oscillator ultrafast laser with a 75 MHz repetition rate, 780 nm center wavelength and 100 fs pulse duration. THz-TDS measurement was performed using a commercial Zomega-2 time-domain THz spectrometer (Zomega Terahertz Corp., Troy, USA) for which the experimental setup has been described previously in detail [22]. The THz region of the spectrometer was purged with high-purity nitrogen to reduce the radiation absorption from ambient water vapor. A total of three THz spectra representing three complete sets of sample and reference measurements were averaged for each final spectrum. The time-domain of the THz electric field was recording for the reference (without sample holder) and each sample, and then after the fast Fourier transform (FFT) the THz spectral absorption was obtained by dividing the sample frequency response by that of the reference.

Results and discussion The THz absorption spectra of piracetam, 2,5-DHBA, and their physical mixture recorded in the range 0.20–1.50 THz are shown in Fig. 2A. The spectrum of physical mixture is almost identical to the linear combination of the individual absorption spectra for the two involved reactants expressing API piracetam and cocrystal former 2,5-DHBA. The absorption feature of 2,5-DHBA distinguishes at 0.84 THz, while the physical mixture shows the major absorption peak at the same region and exhibits a little broader than that of 2,5-DHBA, which is due to the adding effect of

Y. Du et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 111 (2013) 192–195

Absorption coefficient (a.u.)

0.72

(B)

0.15

1.25

Solvent Cocrystal Grinding Cocrystal

1.24

0.72

2,5DHBA Physical Mixture Piracetam

0.84

reference 1 min 4 mins 26 mins 43 mins 69 mins 90 mins

0.10

Amplitude (a.u.)

194

0.05

0.00

0.84

-0.05

(A)

(A) 4

0.2

0.4

0.6

0.8

1.0

1.2

6

8

10

12

Time Delay (ps)

1.4

Fig. 2. Experimentally obtained THz absorption spectra of (A) 2,5-DHBA, piracetam, physical mixture and (B) cocrystals obtained using solvent evaporation and grinding methods in 0.20–1.50 THz region from THz-TDS measurements.

piracetam although its absorption intensity is very weak in the whole frequency range. In contrast, the spectra of cocrystals obtained from slow solvent evaporation and grinding methods (shown in Fig. 2B) are completely different compared with those of physical mixture, showing peaks which are not observed in the physical mixture and individual reactants. The absorption spectrum of the solvent cocrystal distinguishes at frequencies 0.72 and 1.25 THz, while the grinding one shows peaks at 0.72 and 1.24 THz. The absorption pattern and relative intensity in both spectra are almost similar within the instrumental spectral resolution. This result means that the both cocrystals formed with different methods have the same molecular configuration which differs significantly from the crystal lattices of the single parent components. It is well-known that the cocrystallization occurs as the result of intermolecular hydrogen bonding, nonionic or other nonconvalent interactions between two or more molecules of the different components [18,19,21]. So the experimental results indicate that the vibrational modes observed in these spectra are mostly intermolecular character resulting from intermolecular interactions such as hydrogen bonding effect between the parent 2,5-DHBA and piracetam molecules during the cocrystal formation process. To gain a deeper insight into the transformation of the physical mixture into cocrystal, the solid-state reactions were performed by grinding drily together equimolar piracetam and 2,5DHBA at room temperature. Upon grinding, significant changes in both THz time-domain waveforms (shown in Fig. 3A) and frequency-domain spectra (in Fig. 3B) can be detected. Over grinding time, a gradual shift in position of the terahertz pulse to longer time delay and decrease in peak intensity are observed by analyzing the time-domain waveforms directly (shown in Fig. 3A with arrows). In the frequency-domain result shown in Fig. 3B, the distinctive spectral features from the absorption of cocrystal form can be observed increasedly during grinding process. The intensity of the characteristic absorption peak of the physical mixture (at 0.84 THz) decreases, while that of the cocrystal peaks (at 0.72 and 1.24 THz) increases gradually with the grinding time. There are two clear isosbestic points appearing at 0.77 and 0.94 THz position and the arrows indicate the change of absorption peaks due to physical mixture and cocrystal in the reaction process. It means that the starting parent mixture reactant was consumed while a continuation of the cocrystal formation reaction

Absorption coefficient (a.u.)

Frequency (THz)

Physical Mixing 1 min 4 mins 13 mins 26 mins 43 mins 69 mins 90 mins

40

1.24

0.84

20

0.72

(B) 0 0.6

0.8

1.0

1.2

1.4

Frequency (THz) Fig. 3. Experimentally obtained THz time-domain waveforms (A) and frequencydomain spectra and (B) of cocrystal of 2,5-DHBA and piracetam obtained using grinding method under different reaction times (shown in the legend) from THzTDS measurements.

with high efficiency takes place. After around 90 mins grinding, the absorption spectrum only shows characteristic cocrystal peaks, which indicates a complete solid-state cocrystal conversion occurs finally. Further analysis of the tentative changes in the frequency-domain spectra shown in Fig. 3B may reveal further information as to the kinetics of the cocrystal formation during grinding process. The characteristic absorption peaks in frequency-domain spectra, 0.84 THz for the physical mixture and 1.24 THz for cocrystal, are selected for quantitative analysis during the solid-state conversion process. The normalized change of the relative peak height for these two distinctive features over reaction time is obtained, as plotted in Fig. 4A. Using this approach it is possible to extract the cocrystal composition of the reaction mixture at every grinding time point during the conversion process. From this it appears that the progress of the solid-state reaction from physical mixture to form cocrystal exhibits fast growth within in the first several minutes grinding time. At around 30.7 mins, the relative peak height of the parent physical mixture and cocrystal is almost identical, and then the reaction ratio of cocrystal formation becomes slower after this time scale. From the tentative changes shown in Fig. 4 such solid-state cocrystal transformation reaction and the corresponding kinetics information can be directly quantified and investigated.

Y. Du et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 111 (2013) 192–195

1.0

Peak Height (normalized)

1.0

0.8

0.8

~30.7 mins

0.6

0.6

0.4

0.4

195

Acknowledgments This work was partly supported by the National Natural Science Foundation of China (Grant No. 21205110), financial support from Zhejiang Province selected overseas Chinese Scholar (C00783) and Star-up financial support from China Jiliang University (C00152). References

0.2

0.2

@0.84THz @1.24THz 0.0

0.0 0

20

40

60

80

Grinding Time (mins) Fig. 4. Observed normalized change of the relative peak height for the features at 0.84 THz and 1.24 THz in frequency-domain spectra as the reaction time using grinding method.

Conclusions Using THz-TDS spectroscopy, we have recorded the THz absorption spectra of piracetam, 2,5-DHBA, their physical mixture and also their cocrystals obtained from solvent evaporation and grinding methods, in frequency region between 0.20 and 1.50 THz. The THz-TDS spectral results provide fingerprint information of molecular conformations, which may be useful for the discrimination of these different molecular systems. The experimental results show large difference among absorption spectra of the formed cocrystals and the involved parent molecules in 0.20–1.50 THz region. The solid-state reaction process during grinding can be monitored directly from both time-domain waveforms and frequency-domain spectra using THz-TDS technique. The reported results indicate that THz-TDS technology can not only give a new and high potential experimental method to identify and characterize the cocrystal formation in molecule-level, but also provide a useful and promising tool for further monitoring the reaction dynamic process involving two or more molecular crystals in situ to better know the corresponding reaction mechanism in pharmaceutical industry.

[1] G. Springuel, B. Norberg, K. Robeyns, J. Wouters, T. Leyssens, Cryst. Grewth Des. 12 (2012) 475–484. [2] S.L. James, C.J. Adams, C. Bolm, D. Braga, P. Collier, T. Friscic, et al., Chem. Soc. Rev. 41 (2012) 413–447. [3] T. Friscic, W. Jones, J. Pharm. Pharmacol. 62 (2010) 1547–1559. [4] T. Friscic, L. Fabian, J.C. Burley, D.G. Reid, M.J. Duer, W. Jones, Chem. Comm. (2006) 5009–5011. [5] S. Render, C. Klukkert, K.A.M. Lobmann, C.J. Strachan, A. Sakmann, K. Gordon, et al., Pharmacuetics 3 (2011) 706–722. [6] X. Liao, M. Gautam, A. Grill, H.J. Zhu, J. Pharmacy. Pharmacol. 62 (2010) 1547– 1559. [7] J. Lu, A. Rohani, J. Pharm. Sci. 99 (2010) 4042–4047. [8] D.J. Beryy, C.C. Seaton, W. Clegg, R.W. Harrington, S.J. Coles, P.N. Horton, et al., Cryst. Grewth Des. 8 (2008) 1697–1712. [9] E. Zaini, Y.C. Sumirtapura, S.N. Soewandhi, A. Halim, H. Uekusa, K. Fujii, Asian. J. Pharm. Clinical. Res. 3 (2010) 26–29. [10] N. Chieng, M. Hubert, D. Saville, T. Rades, J. Aaltonen, Cryst. Grewth Des. 9 (2009) 2377–2386. [11] M. Walther, B.M. Fischer, A. Ortner, A. Bitzer, A. Thoman, H. Helm, Anal. Bioanal. Chem. 397 (2010) 1009–1017. [12] M. Walther, B. Plochocka, B.M. Fischer, H. Helm, P.U. Jepsen, Biopolymers 67 (2002) 310–313. [13] A. Markelz, S. Whitmire, J. Hillebrecht, R. Birge, Phys. Med. Biol. 47 (2002) 3797–3805. [14] D.F. Plusquellic, K. Siegrist, E.J. Heilweil, O. Esenturk, Chem. Phys. Chem. 8 (2007) 2412–2431. [15] P.C. Upadhya, Y.C. Shen, A.G. Davies, E.H. Linfield, Vib. Spec. 35 (2004) 139– 143. [16] M.D. King, W.D. Buchanan, T.M. Koeter, Anal. Chem. 83 (2011) 3786–3792. [17] M.D. King, T.M. Korter, J. Phys. Chem. A. 114 (2010) 7127–7138. [18] L.L. Nguyen, T. Friscic, G.M. Day, L.F. Gladden, W. Jones, Nat. Mat. 6 (2007) 206–209. [19] W. Limwikrant, K. Higashi, K. Yamamoto, K. Moribe, Int. J. Pharm. 382 (2009) 50–55. [20] X. Liu, G. Liu, H. Zhao, Z. Zhang, Y. Wei, M. Liu, et al., J. Phys. Chem. Solid. 72 (2011) 1245–1250. [21] E.P.J. Parrott, J.A. Aeitler, T. Friscic, M. Pepper, W. Jones, G.M. Day, et al., Cryst. Grewth Des. 9 (2009) 1452–1460. [22] X.J. Li, Z. Hong, J. He, Y. Chen, Opt. Comm. 283 (2010) 4701–4706.