2021 Volume 69 Issue 3 Pages 271-277
Vibrational spectroscopic imaging has become useful analytical tools for quality control of drug products. In this study, we applied microscopic attenuated total reflection (ATR)-IR and confocal Raman microscopy to elucidate microscopic structure of creams and for the formulation design in the development of semi-solid drug products. The model creams were prepared with prednisolone (PRD) and fluconazole (FLC) as active pharmaceutical ingredients and oily solvents such as mineral oil (MO), isopropyl myristate (IPM), benzyl alcohol (BA) and diethyl sebacate (DES). As a result of microscopic ATR-IR imaging, several domains indicating oily internal phase were observed, which had absorption around 1732 and 1734 cm−1 derived from MO, IPM and DES. In addition, domains of BA around 1009 cm−1 were observed at the complemental or similar position in the formulation with MO or DES, respectively. These results suggested that the creams were oil-in-water type and the distribution of domains would reflect the compatibility of the solvents. The contents of PRD and BA were determined quantitatively in each layer after the intentional separation of the creams and the results agreed well with the imaging analysis. Whereas, confocal Raman imaging allowed to visualize the distribution of the components in depth direction as well as two-dimensional plane. In particular, the Raman imaging would ensure the coexistence of FLC and BA as oily phase in the cream. From these results, the feasibility of spectroscopic imaging techniques was successfully demonstrated for the formulation design of semi-solid dosage forms.
Topical drug products are used to treat skin diseases such as atopic dermatitis and psoriasis vulgaris. Because the topical dosage forms are applied directly to the affected area, they are considered to be safe with regards to the systemic side effects. In particular, oily ointments containing petrolatum as a main base can be applied to both wet and dry diseased part in skin surfaces. Therefore, topical drug products are widely used as most convenient dosage forms for patients with skin diseases. However, there are sometimes decrease in adherence of patients who have troublesome in applying topical drug products. Generally, oily ointments are sticky and difficult to spread, thus some patients prefer creams or lotions that are less sticky and easy to spread.1,2) Consequently, pharmaceutical companies often develop multiple dosage forms for one active pharmaceutical ingredient (API) so that the dosage forms can be used depending on the pathological condition, application part and patient preference, or seasonal condition occasionally.3)
Designing emulsion-type formulations including creams is relatively difficult among topical dosage forms due to thermodynamic instability and complicated interaction with many constituents. A major issue is to ensure the stability of API in dosage form in the development of topical drug products. The requirements are both chemical and pharmaceutical stability, which are to avoid degradation of API and separation of the oil and water phase in the formulation, respectively. Creams including water-in-oil and oil-in-water types, consist of API, oily components, aqueous components, emulsifiers (surfactants) and preservatives and so on. Generally, continuous experimental trial and error is done in the formulation design, and the quality of formulation is improved by considering the factors to affect the total robustness of formulation. There are evaluation methods for formulations such as quantitative assay using a HPLC, observation of emulsified particles using an optical microscope and rheological properties using a rheometer. However, identifying the distribution of individual components in the dosage form could be difficult by using those techniques.
In recent years, microscopic imaging systems using various spectroscopic approaches such as characteristic X-ray, IR, near IR, terahertz, and Raman have become useful analytical tools in the fields of formulation design and quality control.4,5) Although these methods can demonstrate the mixing uniformity and dispersion state of API and excipients in oral dosage forms including powders, granules and tablets,6–10) there are few reports for topical semisolid formulations. We have reported that Almeta® ointment, which is an oil droplet dispersion type ointment containing alclometasone dipropionate, and its generic products were compared of their internal structure by using microscopic attenuated total reflection (ATR)-IR imaging technique.11) As a result, the distribution of API and the excipients were differentiated between the original and generic products. In addition, the confocal Raman spectroscopy was also promising analytical tool.12) In this study, we applied microscopic ATR-IR and confocal Raman microscopy to creams consisting of many components which made it difficult to understand the distribution of API and other excipients in formulations, and studied the feasibility of these approach for the formulation design in the development of semi-solid drug products.
As active ingredients (model drugs), prednisolone (PRD) and fluconazole (FLC), a synthetic adrenocortical hormone drug and an antifungal drug that are widely used in topical drug preparations, were purchased from Tokyo Chemical Industry Co., Ltd. (Tokyo, Japan) and used as received.
Glycerin (GLY) was purchased from New Japan Chemical Co., Ltd. (Osaka, Japan). 1,3-Butylene glycol (BG) was purchased from Daicel Corporation (Osaka, Japan). Propylene glycol (PG) and Dipropylene glycol (DPG) were purchased from Adeka Corporation (Tokyo, Japan). Macrogol 400 (polyethylene glycol (PEG)) was purchased from NOF Corporation (Tokyo, Japan). Propylene carbonate (PC) was purchased from BASF Japan Ltd. (Tokyo, Japan). Ethyleneglycol Salycylate (EGS) was purchased from API Corporation (Tokyo, Japan). Triacetin (TA) and Benzyl alcohol (BA) were purchased from Merck KGaA (Darmstadt, Germany). Diisopropyl adipate (DIA), Diethyl sebacate (DES), Glyceryl triisooctanoate (TIO), Isopropyl myristate (IPM), Glyceryl monostearate (MGS) and Polyoxyethylene cetyl ether (POC) were purchased from Nikko Chemicals Co., Ltd. (Tokyo, Japan). Medium chain fatty acid triglyceride was purchased from Watahan Trading Co., Ltd. (Tokyo, Japan). Liquid paraffin (MO) was purchased from Kaneda Co., Ltd. (Tokyo, Japan). Carboxy vinyl polymer was purchased from FUJIFILM Wako Pure Chemical Corporation (Osaka, Japan). Diisopropanolamine was purchased from Mitsui Fine Chemicals, Inc. (Tokyo, Japan).
Regarding reagents, silver nitrate was purchased form Nacalai Tesque, Inc. (Kyoto, Japan), methanol was purchased from FUJIFILM Wako Pure Chemical Corporation, and methyl parahydroxybenzoate was purchased from Tokyo Chemical Industry Co., Ltd.
Solubility of APIs to Pharmaceutical SolventsThe solubility of PRD and FLC in excipients commonly used as solvent in topical preparations13,14) was determined (Table 1). Approximately 0.4 g of excipient was added to 0.1 g of PRD or FLC, and ultrasonic irradiation was performed for 10 min, and observed visually that the APIs were dissolved (20% dissolution). If there was undissolved residue, 0.5 g of the excipient was further added, and repeated same procedure as described above (10% dissolution). In the case of further undissolved residue, 1, 3 and 5 g of the excipients were added and treated as 5, 2 and 1% dissolution, respectively.
| Pharmaceutical solvents | PRD | FLC |
|---|---|---|
| PG | 2% | — |
| BG | 2% | — |
| PEG | 1% | — |
| DPG | 2% | — |
| EGS | 5% | 10% |
| BA | 5% | >20% |
| PC | <1% | — |
| TA | <1% | — |
| DIA | <1% | — |
| DES | <1% | — |
| TIO | <1% | — |
| MCT | <1% | — |
| IPM | <1% | — |
—: Not tested.
The compatibility of BA, which is commonly used as dissolving agent, and the other pharmaceutical solvents selected in above section was studied. The solvents and BA were mixed at a weight ratio of 1 : 1 and the compatibility was visually observed (Table 2).
| Pharmaceutical solvents | BA |
|---|---|
| Purified Warer | × |
| GLY | 〇 |
| BG | 〇 |
| PG | 〇 |
| PEG | 〇 |
| DPG | 〇 |
| PC | — |
| TA | 〇 |
| DIA | — |
| DES | — |
| TIO | — |
| MCT | 〇 |
| IPM | 〇 |
| MO | × |
〇: Compatible, ×: Not compatible, —: Not tested.
IR spectrum of the APIs, excipients and their mixtures were measured as standard spectrum using an IR spectrometer (FT-IR-4100, JASCO Corporation, Tokyo, Japan) equipped with an ATR unit (DuraSamplIR II™ Smiths Detection Group Ltd., Watford, U.K.). The background was measured in air, and sample scans were recorded at a 4 cm−1 spectral resolution with 16 scans across the range of 4000–600 cm−1.
Preparation of CreamsThe formulations of the creams prepared in this study are shown in Table 3. The APIs were dissolved in the dissolving agents (IPM, DES and BA). The ingredients excluding APIs and its dissolving agent were heated at 70 °C. The dissolved APIs were added to the ingredients keeping at 70 °C. The mixture was stirred to room temperature to obtain a cream sample.
| Cream 01 | Cream 02 | Cream 03 | Cream 04 | Cream 05 | |
|---|---|---|---|---|---|
| PRD | 0.10 | — | 0.10 | — | — |
| FLC | — | 0.10 | — | 0.10 | 2 |
| MO | 20 | 20 | — | — | — |
| IPM | 5 | 5 | — | — | — |
| DES | — | — | 10 | 5 | 10 |
| BA | 2 | 2 | 10 | 5 | 10 |
| MGS | 2 | 2 | 2 | 2 | — |
| POC | 1 | 1 | 1 | 1 | 1 |
| Gly | 5 | 5 | — | 5 | — |
| Others | ad 100 | ad 100 | ad 100 | ad 100 | ad 100 |
Spotlight 400 (PerkinElmer, Inc. Japan, Kanagawa, Japan), a line chemical imaging system equipped with an MCT linear array detector and a Ge ATR accessory, was used to collect the IR spectra of the creams. Each spectrum came from a 1.56 × 1.56 µm square pixel. The background was measured in air, and sample scans were recorded at 8 cm−1 spectral resolution in the range of 4000–720 cm−1. Data were analyzed using Isys chemical imaging software (version 5.0; Malvern Instruments Ltd., Worcestershire, U.K.). Offset data from the creams were used to generate chemical images using a peak height method based on the characteristic peaks of the pure components. The microscopic features of the products were also characterized by applying a small amount of the sample onto a microscope slide. A cover slip was placed over the sample, which was observed under an E-600-Pol polarizing microscope (Nikon Corporation, Tokyo, Japan) in reflection mode at 200 × magnification.
Confocal Raman Microscopy MethodThe confocal Raman microscope LabRam HR Evolution (Horiba, Kyoto, Japan) was used to acquire the Raman spectra of prepared model creams and reference materials. The samples were excited with a 532 nm laser through a 60 × objective. A 1.25 µm step size was used with exposure time of 0.04 s scanning through a 60 × 60 × 25 (X, Y, Z) µm area. The Raman signatures of the samples were acquired in the fingerprint region (3300–300 cm−1). LabSpec 6.0 software was used for spectral collection, spectral preprocessing and data set processing. The obtained spectra were preprocessed to remove influence of fluorescence, and then the spectra were used to generate chemical images using the classical least squares (CLS) algorithm.
Quantitative Assay for PRD in BA and Water ContentsHimac CF7D2 (Hitachi, Ltd., Tokyo, Japan), a centrifugal machine, was used to separate between oil and water phase of creams. Thirty grams of the cream was placed in a centrifuge tube and centrifuged at 5000 rpm for 10 min. If not separated enough, 0.3 g of silver nitrate was added to the sample, and the mixture was centrifuged again under the same conditions.
The water content was determined for each upper and lower layer separated by the procedure as described above according to the Karl Fisher method in JP17.15) The experiment was conducted via the coulometric titration method at room temperature. A sample (2–3 mg) was weighed on a parafilm cut to 1 cm2 and inserted into the reactor chamber of a KF aquameter (Aquacounter AQ-7, Hiranuma Sangyo Co., Ltd., Ibaraki, Japan) filled with KF reagent, generator electrolyte (Aqualyte RO, Hiranuma Sangyo Co., Ltd.), and counterelectrolyte (Aqualyte CN, Hiranuma Sangyo Co., Ltd.) at a titration rate of 0.6 mg H2O/min. Measurements were performed in triplicate.
HPLC system (EXTREMA and ChromNAV V2-02, JASCO Corporation, Tokyo, Japan) was used to assay PRD and BA in the creams. Approximately 0.5 g of sample was weighed, and added exactly 2.5 mL of internal standard solution and 2.5 mL of silver nitrate solution and 15 mL of mobile phase. After mixing and centrifuging at 6000 rpm for 10 min, this solution was filtered through a 0.45 µm membrane filter. The first 3 mL of filtrate was removed, and the next filtrate was used as the sample solution.
Separately, 25 mg of PRD, previously dried in a desiccator at 105 °C for 3 h, was accurately weighed, and dissolved in the mobile phase to make exactly 50 mL. Pipetted 1 mL of this solution, and added exactly 2.5 mL of the internal standard solution. Further added mobile phase to make 20 mL and used this solution as the standard solution 1. Furthermore, accurately 50 mg of BA was weighed, and the mobile phase was added to make exactly 25 mL. Pipetted 5 mL of this solution, and added exactly 2.5 mL of the internal standard solution. Further added mobile phase to make 20 mL and used this solution as the standard solution 2. The test with each 10 µL of the sample solution, the standard solution 1 and the standard solution 2 were performed as directed under Liquid Chromatography <2.01> in JP17 according to the following conditions. The ratios, QT, QS1 and QS2, of the peak area of PRD or BA to that of internal standard were calculated as followings;
 |
MS1: Amount (mg) of PRD taken, MS2: Amount (mg) of BA taken, MT: Amount (g) of sample taken, Internal standard solution–A solution of methyl para-hydroxybenzoate in methanol (1 in 2000)
Operation conditions; Detector: An UV absorption photometer (wavelength 254 nm), Column: Inertsil ODS-3 (4.6 mm i.d. × 150 mm, 5 µm, GL Sciences Inc., Tokyo, Japan), Column temperature: 40 °C, Mobile phase: H2O : CH3CN : CH3OH = 3 : 1 : 1 (v/v), Flow rate: Adjust so that the retention time of prednisolone is about 11 min.16)
Table 1 shows the solubility of the APIs in various pharmaceutical solvents. PRD and FLC were readily soluble in BA and EGS, in particular BA was able to dissolve more than 20% FLC. In case that the creams were prepared using PRD or FLC as APIs and BA as dissolving agent, the APIs were considered to be in liquid state in BA.
Table 2 shows the results of compatibility between BA, which is a dissolving agent, and various excipients. BA was not compatible with water and MO, but compatible with other excipients (GLY, BG, PG, PEG, DPG, TA, MCT and IPM). BA and APIs dissolved in BA were estimated to present in either oil or water phase, otherwise as independent third phase in creams. It is considered extremely difficult to predict their distribution in creams.
Standard IR Spectra of Each Components in CreamsPrior to microscopic ATR-IR imaging, the IR spectrum of the APIs and excipients contained in the creams were measured (Fig. 1). Characteristic absorption peaks derived from PRD and FLC were observed at 1611 and 1652 cm−1, and 1502 and 1618 cm−1, respectively. In addition, the characteristic peak derived from BA were observed at 1009 cm−1. Furthermore, these specific peaks were not observed on other excipients in the creams, and it was considered to be possible to distinguish the APIs and dissolution agent from the other excipients.
(Color figure can be accessed in the online version.)
Microscopic ATR-IR imaging of the creams containing MO and IPM (Cream 01 and 02) as oil components were conducted. Firstly, the absorption peaks around 1732, 2852 and 2920 cm−1, which are characteristic for the oil components such as MO and IPM, were selected and microscopic ATR-IR imaging was performed (Figs. 2a, c for 1732 cm−1, Fig. S1 for 2852 and 2920 cm−1). All red domains were observed in the internal phase of creams. In addition, the distribution of red domains in the images on 1732 cm−1, characteristic for IPM, was similar to those in images on 2852 and 2920 cm−1, characteristic for MO, therefore, IPM and MO would be coexisting. Based on the above, the creams were considered to be oil-in-water type containing oil phase consisted of IPM and MO. Conventionally, the type of creams has been determined by the dye method which aqueous or oily dye was added to cream samples and observed stained area visually.17,18) On the other hand, this microscopic ATR-IR imaging was component-specific and non-destructive method, thus this method was considered to be useful for discriminating the type of creams. Next, the absorption peak near 1009 cm−1 which was characteristic for BA were selected and the imaging analysis was performed (Figs. 2b, d). The red regions were observed on the external water phase of creams, thus BA could mainly distribute on the external water phase. In addition, these images would reflect the incompatibility of BA and MO as shown in Table 2, since the distribution of red domains in the images on 1732 cm−1 (Figs. 2a, c) and that of the blue domains in the images on 1009 cm−1 (Figs. 2b, d) coincided well, respectively.
(Color figure can be accessed in the online version.)
Microscopic ATR-IR imaging of the creams containing DES (Creams 03 and 04) used as the oily component was conducted. The absorption peak around 1734 cm−1, which was characteristic of the oil component DES, was selected for imaging analysis (Figs. 3a, c). The red domains were observed in the internal phase, thus the creams containing DES were considered to be also oil-in-water type creams. The absorption peak near 1009 cm−1 characteristic for BA was selected and the red domains indicated the internal phase same as mentioned above (Figs. 3b, d). Because the polarity of DES was much higher than that of MO and IPM and closer to that of BA, DES and BA were compatible and considered to form the oil phase together.
(Color figure can be accessed in the online version.)
From these results, the distribution of excipients was clearly identified and then the feasibility of imaging analysis with ATR-IR was successfully demonstrated. However, there was no domain derived from APIs observed, the sensitivity for APIs might be insufficient in this experimental condition. Thus, the concentration of PRD in water and oil phase of cream 03 was determined by HPLC measurement after intentional separation of the cream and extraction of the PRD and BA. Table 4 shows the content of components in separated layer of cream 03. Although the separation of cream 03 was incomplete, the upper layer would be water phase from the appearance and the water content. The both contents of PRD and BA were greater in oil phase compared with those of water phase. The estimated PRD content was approximately 1.2% dissolved in BA of oil phase separated from cream 03. In the case that the DES coexisted with BA as oily droplets in the cream, the concentration of API was diluted more.
| Separated layer of cream 03 | Appearance of layer | Water content (%) | Estimated phase | PRD content (%) | BA content (%) | Estimated content of PRD/BA (%) |
|---|---|---|---|---|---|---|
| Upper layer | Liquid | 54.8 | Water phase | 0.03 | 4.4 | 0.68 |
| Lower layer | Semi-solid | 28.1 | Oil phase | 0.25 | 21.4 | 1.17 |
According to the previous report, the contents of APIs in liquid droplets were estimated to be 5 to 3.0% and around 2.0% in alclometasone dipropionate11) and tacrolimus ointments,19) respectively. The slight characteristic peaks of API were detected in the case of alclometasone dipropionate. The IR spectra of PRD dissolved in equivalent mixture of BA and DES are shown in Fig. 4. The slight peak characteristic for PRD appeared around 1652 cm−1 more than 2% of PRD (Figs. 4f, g, h). Therefore, low content API was difficult to be observed in imaging analysis with ATR-IR. Although cream 05 was prepared as a model containing high concentration of API owing to high solubility of FLC into BA, no signal from FLC was detected in ATR-IR imaging.
(Color figure can be accessed in the online version.)
Confocal Raman microscopy was applied to the cream 05 containing high concentration of FLC and oily components DES and BA. The Raman spectra of interest components in the cream 05 were measured ahead of the imaging analysis (Fig. S2). Characteristic scattering peaks derived from FLC and BA were observed around 1250 to 1350 and at 1000 cm−1, respectively. In addition, there was almost no scattering with water, thus it was likely possible to distinct these components. Figure 5 shows the distribution of water, BA and FLC, moreover their overlay image. The coexistence of FLC and BA excluding water was clearly demonstrated. In addition, a depth profiling was also possible in confocal Raman imaging, up to around 25 µm in depth. The great advantage of Raman spectroscopy is less sensitive to water,20) thus it seems to be quite suitable for cream sample among semi-solid dosage forms. Therefore, confocal Raman microscopy would be promising analytical tool for the stage of formulation design.
(Color figure can be accessed in the online version.)
In this study, we attempted to apply vibrational spectroscopic imaging such as microscopic ATR-IR and confocal Raman to evaluate the cream formulations. It was challenging research task from following reasons; (i) mixed system consisting of many kinds of components, (ii) complicated interaction among water, oil and surfactant at their interface and (iii) microstructure of emulsion state which require high resolution analysis. We had successfully demonstrated the visualization for the internal structure of commercially available drug products. Namely, the distribution of API and excipients has been able to be observed in various dosage forms. This technique would be useful for formulation study as well as quality control of drug products, which was certified by the results for the model creams evaluated in this study. The feasibility of microscopic ATR-IR was ensured in particular oily components, also the confocal Raman would be essential tool for drug development.
This work was supported in part by JSPS KAKENHI, a Grant-in-Aid for the Scientific Research (C), Grant Number 17K08253 (to T.F.). The authors thank Tomoko Numata and Yutaro Hirose in Horiba, Ltd. for distinguished technical assistance and great support for the confocal Raman microscopy experiments.
The authors declare no conflict of interest.
The online version of this article contains supplementary materials.
