Introduction

Mirtazapine (MRT) is an antidepressant included in the tetracyclic piperazinoazepine pharmaceutical class that presents with an atypical action mechanism. Being a presynaptic α2-adrenoreceptor antagonist, it acts by increasing central serotonin and norepinephrine release. Furthermore, it is able to potently antagonize histamine H1 receptors determining a calming, sedative effect, and serotonin 5-HT2A, 5-HT2C and 5-HT3 receptors leaving an increased concentration of serotonin to interact with the 5HT1 receptor and determine the antidepressant effect [1, 2]. Its main therapeutic indication is the treatment of major depressive disorder, but due to its sedative, anxiolytic, antiemetic and appetite stimulant properties, clinicians often prescribe it for the management of insomnia, post-traumatic stress disorder, anxiety or panic disorder, obsessive–compulsive disorder and migraines [3,4,5].

MRT presents as a white, slightly hygroscopic powder, with a molecular mass of 265.3529 g/mol and a melting point of 114–116 °C [3, 6]. From a structural point of view, MRT (IUPAC name 5-methyl-2,5,19-triazatetracyclo[13.4.0.02,7.08,13]nonadeca-1(15),8,10,12,16,18-hexaene) has a four-cyclic skeleton (Fig. 1), each cycle having different characteristics: the A cycle is a pyridine, B is a seven-membered azepine heterocycle, C is a phenyl moiety, while D belongs to the piperazine class [7].

Fig. 1
figure 1

The chemical structure of mirtazapine (MRT)

Full size image

As for its solubility, there seem to be some discrepancies in scientific literature. While, the online Biopharmaceutical classification system (BCS) database puts MRT in the high solubility, high permeability BCS class I [8], scientific articles mention its positive log P coefficient (2.9 [3, 9, 10] and 3.3 [11]) and place MRT in the low solubility, high permeability BCS class II [12]. Considering, the log P coefficient and the reported high solubility in solid lipids [10], a high-water solubility seems unlikely and MRT’s inclusion in the BCS class II appears more plausible.

In pharmaceutical formulations found on the market, MRT is incorporated mainly as its hemihydrate form (MRTHH). Sarma et.al. conclude in their study about MRTHH’s crystal structure that a single unit cell must contain four molecules of mirtazapine and two molecules of water in order to achieve the hemihydrate stoichiometry [6].

In the pharmaceutical field an active pharmaceutical ingredient (API) is almost never administered to a patient as such, the journey from active molecule to actual drug involving the presence of several excipients or additives [13]. Although excipients were first believed to be just a mean to an end, i.e., to make sure that the final pharmaceutical formulation has the proper consistency, mass and volume in order to be efficiently administered, nowadays it has been proved that their individual characteristics may directly affect the API [14, 15].

Drug-excipient compatibility studies represent a very important step in the drug development process, as incompatibilities may lead to the inactivation of the API, its conversion to an undesired form or can affect its stability and bioavailability [16]. Nonetheless, given the advances in the pharmaceutical technology field, an incompatibility discovered between an API and an excipient in the pre-formulation studies does not necessarily mean that the excipient can never be used together with said API. Special technological procedures, such as separate granulation, multilayered formulation or the addition of desiccants can be used to minimize or even completely eliminate undesired chemical interactions [17].

As such, compatibility studies are no longer used just to identify possible interactions between APIs and excipients, but also to give indications regarding the series of technological manufacturing procedures needed to obtain a stable, safe and therapeutically efficient formulation. There are a great number of instrumental methods of analysis employed, spectroscopic and thermal methods being two of the most frequently used techniques. Furthermore, thermal analysis is used in order to evaluate and predict the shelf-life of various pharmaceutical formulations [18, 19].

The aim of the present study was to evaluate the compatibility between MRTHH and ten different widely used pharmaceutical excipients, namely α-lactose monohydrate, starch, D-mannitol, sorbitol, stearic acid, magnesium stearate, calcium lactate, magnesium citrate, polyvinylpyrrolidone K30 and aerosil. The presence of incompatibilities between the API and the selected excipients was evaluated using Fourier transform infrared spectroscopy (FTIR) performed on all pure samples and prepared mixtures at room temperature (23 ± 2 °C), as well as a complete thermal stress evaluation (TG—thermogravimetric/DTG—derivative thermogravimetric/DSC—differential scanning calorimetry).

Materials and methods

Samples and preparation

Mirtazapine hemihydrate (MRTHH) was selected as the active pharmaceutical ingredient (API) and was used as such (lot no. 00191348). It was acquired from Pharmaffiliates, India as an API standard (purity ≥ 97%) used without further purification and stored according to the supplier’s specifications.

α-lactose monohydrate (Lact) was acquired from Sigma Life Science, Germany (lot. no. #SLBK4809V), being obtained from bovine milk and presenting with a total lactose content of 99% (determined, according to the supplier, by Gas-Chromatography).

Starch (Sth) was a product commercialized by Sigma-Aldrich, Germany (lot. no. #SZBF1670V), with a low impurity content of metal elements (puriss. p.a.), being obtained from potato and having the characteristics required by the European Pharmacopoeia.

D-mannitol (Man) was purchased from Sigma-Aldrich, Germany (lot no. #BCBM1675V) as an ACS reagent in powder form. Sorbitol (Sorb) was acquired from Fisher Scientific, UK (lot no. 1205417) presenting as a laboratory grade pharmaceutical excipient.

Stearic acid (StA) was commercialized by Merck, Germany (lot no. S7080573540), purity ≥ 97%, while magnesium stearate (MgSt) was acquired from Sigma-Aldrich, Germany (lot. no. #SZBF2590V) purity ≥ 90%.

Calcium lactate (CaLact), PURACAL® DC, is the highly soluble calcium derivative of natural L-lactic acid and was received from Corbion, Netherlands (code 1.02102). Magnesium citrate (MgCit) was purchased from Jost Chemical, US (code 2549) as the tribasic anhydrous granular form, having US Pharmacopoeia certification.

Polyvinylpyrrolidone K30 (PVP) was acquired from Sigma-Aldrich, Germany (lot. no. #BCBV7579) in powder form as well as Aerosil 200 (SiO2), with a 200 nm particle size (code 914,878).

All excipients were used as received, without any additional purification.

Preparation of mirtazapine-excipient binary mixtures

The drug-excipient binary mixtures were prepared by weighing equal masses of the corresponding two components for each MRTHH + excipient mixture. The compounds were triturated in an agate mortar in the presence of absolute ethanol for 1 h, the latter being added as needed as to maintain a paste-like consistency. Trituration was continued until the complete evaporation of the solvent and, after spraying, the samples were placed in an oven in open containers for 12 h at 30 °C to ensure the complete evaporation of the ethanol. After drying, the samples were placed in airtight vials and stored at room temperature until the instrumental evaluations were performed. The same trituration protocol in the presence of ethanol was followed for the pure MRTHH in order to ensure the same sample treatment and to be able to accurately compare the obtained results.

FTIR study

An ATR-FTIR (Attenuated Total Reflection Fourier Transform Infrared Spectroscopy) evaluation was performed at room temperature (23 ± 2 °C) under normal atmospheric conditions on all pure compounds and on the prepared binary mixtures using a Shimadzu Fourier transform infrared spectrophotometer IR tracer 100 fitted with an atr single reflection accessory. The spectral range 4000–400 cm−1 was selected and a resolution of 4 cm−1 was used for the determinations. Each presented spectrum represents a superposition of 20 individual scans.

Thermogravimetric study

The thermogravimetric (TG), derivative thermogravimetric (DTG) and differential scanning calorimetry (DSC) evaluation was performed for the pure API, the pure excipients, as well as for each of the prepared mixtures using the thermal analysis system TGA/DSC 3 + (acquired from Mettler Toledo), after the system has been calibrated with In (ME-119442), Zn (ME-119441), Al (ME-5119701), Au (ME-51140816) and Pd (ME-51140817). The TG/DTG data were obtained after the samples were weighted and placed in open 40 μL aluminum crucibles. For the DSC evaluation the same procedure was followed with the addition of a piercing lid that was used to seal each crucible. During the study, a synthetic air atmosphere was used, with a flow rate of 20 mL min−1. A single heating rate was set for all experiments (β = 10 °C min−1), the compounds being evaluated in the temperature range 30–400 °C in order to reveal their behavior when subjected to thermal stress. The collected DSC data was expressed as mW mg−1. The experimental procedures were repeated twice for each sample and the TG/DTG and DSC obtained data were basically unchanged.

Results and discussion

The current study presents a preliminary compatibility evaluation between MRTHH and ten commercially used pharmaceutical excipients performed using two types of instrumental techniques, namely the nonthermal spectrophotometric FTIR analysis and the TG/DTG/DSC thermal stress evaluation.

It is well known that infrared spectra are obtained, because chemical compounds can absorb energy when irradiated with an electromagnetic radiation in the IR frequency region. As a consequence of energy absorption, multiple molecular vibrational and rotational states determine characteristic signals in the registered spectrum [20]. Fourier transform infrared spectroscopy (FTIR) has been an indispensable tool in the chemical and pharmaceutical areas, since it requires easily available instrumentation, small samples that are not destructed during the analysis and it is a fast, low-cost and no-reagents required technique that provides significant information regarding the structure of a wide variety of molecules. In the case of organic molecules, the characterization and detection of covalent bonds and functional groups is mostly performed in the mid-IR spectral region (4000–400 cm−1) [21].

The applicability of FTIR in drug-excipient compatibility studies has been proved over the years in scientific literature [22,23,24,25]. From a practical approach, the FTIR spectra of the drug and excipient are compared with those obtained for the drug-excipient mixture. Spectral changes, such as the disappearance of individual bands, the appearance of new peaks or a significant peak intensity reduction are considered clear indications that the investigated API and excipient interact with each other [26]. In any final pharmaceutical formulation unwanted drug-excipient interactions must be avoided as much as possible since they may influence the stability, overall bioavailability and final therapeutic effect of the drug [13, 16, 27, and 28].

Thermogravimetric analysis (TGA) has been used successfully in order to compare the behaviors of the pure API and excipients with the ones of the mixtures prepared for the interaction study when subjected to thermal stress [29,30,31,32]. TGA coupled with DSC provides information regarding thermally-induced chemical degradations as well as physical transformations [14, 16, and 30]. It is generally considered that when the obtained thermal characteristics of the mixtures represent the sum of the individual properties of the APIs and excipients, the compounds are considered entirely compatible with one another. Changes in thermal profiles and especially in the DSC results raise the issue of possible incompatibilities [14, 30]. It is documented in scientific literature that the suspicion of drug-excipient interactions arises from significant shifts of the DSC melting signal as well as the modification of the enthalpy value associated with the melting process of the API [33,34,35]. A definite conclusion cannot be drawn however from the sole interpretation of the DSC curves, an incompatibility assessment requiring correlations between the results obtained using multiple techniques, among others thermal analysis and IR spectroscopy [36].

The data obtained during the FTIR analysis performed for MRTHH and each of the ten drug-excipient mixtures are shown in Table 1. For the binary mixtures, the spectral positions of the peaks associated with the excipient are presented with an Italic font and the new bands unseen for either pure MRTHH or the excipient are marked with a Bold font. The ATR-FTIR spectra are plotted on the 4000–400 cm−1 spectral range (Fig. 2a–e), the results being presented comparatively between pure MRTHH, each pure excipient and the analyzed binary mixtures.

Table 1 The FTIR spectral data obtained for MRTHH and the prepared binary mixtures
Full size table
Fig. 2
figure 2figure 2

The ATR-FTIR spectra of MRTHH, pure excipients and respective binary mixtures (50% w/w API and 50% w/w excipient): a—Lact, MRTHH + Lact, Sth, MRTHH + Sth, b—Man, MRTHH + Man, Sorb, MRTHH + Sorb, c—StA, MRTHH + StA, MgSt, MRTHH + MgSt, d—CaLact, MRTHH + CaLact, MgCit, MRTHH + MgCit and e—PVP, MRTHH + PVP, SiO2, MRTHH + SiO2

Full size image

The results obtained during the thermal protocol applied at a heating rate (β) of 10 °C min−1 in a synthetic air atmosphere on the temperature range 30–400 °C can be seen in Table 2, while the TG, DTG and DSC curves are visible in Fig. 3a, b, c, d, e, f, respectively.

Table 2 The thermoanalytical data (TG/DTG/DSC) obtained for MRTHH and the prepared binary mixtures at β = 10 °C min−1
Full size table
Fig. 3
figure 3figure 3

The overlapping of the a, b TG (thermogravimetric), c, d DTG (derivative thermogravimetric) and e, f DSC (Differential scanning calorimetry) curves obtained for the pure API and excipients, as well as for each prepared mixture (50% w/w API and 50% w/w excipient), as follows: (A) MRTHH, (B) Lact, (C) MRTHH + Lact, (D) Sth, (E) MRTHH + Sth, (F) Man, (G) MRTHH + Man, (H) Sorb, (I) MRTHH + Sorb, (J) StA, (K) MRTHH + StA, (L) MgSt, (M) MRTHH + MgSt, (N) CaLact, (O) MRTHH + CaLact, (P) MgCit, (Q) MRTHH + MgCit, (R) PVP, (S) MRTHH + PVP, (T) SiO2 and (U) MRTHH + SiO2 in a synthetic air atmosphere on the temperature range 30–400 °C at β = 10 °C min.−1

Full size image

Mirtazapine hemihydrate (pure API)

The FTIR spectrum of MRTHH (Fig. 2) confirmed the structural identity of the chemical compound, the overall shape and individual peaks (Table 1) being almost identical to the literature available data [6, 7, 10].

Although the structure of the compound contains only three types of atoms (C, H and N), their arrangement in a four-cyclic structure determines the appearance of multiple signals that can be observed on MRTHH’s spectrum. The presence of the crystallization water is confirmed by the wide band seen in the 3500–3100 cm−1 spectral range determined by the stretching vibration of the O–H bond. While, the signals observed in the 3060–3000 cm−1 spectra range are determined by the aryl-type stretching vibration of the C–H bonds from the phenyl (C) and pyridine (A) rings, the presence of the C–H bonds from the B and D heterocycles, as well as from the free methyl group are firstly revealed by the asymmetric and symmetric stretching vibrations seen in the 3000–2700 cm−1. The stretching vibrations of the C–C and C=C bonds can be associated with the signals seen at 1585, 1566 and 1497 cm−1, in agreement with data presented by Sarma B. et. al. and Sagdinc S. et. al. [6, 7]. The in-plane bending vibrations of the –CH2 and –CH3 groups can be correlated with the signals seen at 1466 and 1442 cm−1. The 1390–1047 cm−1 spectral range reveals peaks determined by the in-plane bending vibration of the C–H bonds, but also by the stretching vibration modes of the C–N bonds (signals at 1306, 1265 and 1223 cm−1 in correlation with literature data [7]), while the out-of-plane bending of the C–H bonds can be found in the 937–812 cm−1 range. The lower end of the spectra reveals multiple twisting vibrations for the C–C–C–C, H–C–C–H, C–C–N–C and H–C–C–C moieties found in MRTHH’s structure, the out-of-plane deformation vibrations of the A, B, C and D rings, as well as the rocking vibration of the methylene groups found in B and D rings (peaks in Table 1).

The thermal analysis data performed for pure MRTHH revealed a two-stage mass loss (see Fig. 3 and Table 2). The TG curve of the pure API (Fig. 3a, b(A)) reveals a mass loss process starting from 74 °C, with a small Δm value (0.25%) up until the 110–135 °C temperature range in which the mass loss becomes more obvious and the process can be associated with a single DTG peak seen at 116 °C (Fig. 3c, d(A)). Although this stage should be correlated with the loss of the crystallization water of the hemihydrate form of the API, literature data mentions that hot stage microscopy revealed a sublimation process associated with MRT base starting from 65 °C [6] and so, an accurate calculation of the water content of MRTHH is difficult to perform since the sublimation begins at a much lower temperature value than the beginning of the process revealed on the DTG curve. After this, a relatively quick decomposition of MRT takes place in the 135–265 °C temperature range, a single DTG peak being revealed at 254 °C, the degradation being almost complete before 270 °C. The melting process of the API is revealed on the DSC curve (Fig. 3e, f(A)) by the endothermic signal identified at 117 °C (in agreement with literature data [3, 6]) correlated with an enthalpy value of 105.5 J g−1. An exothermic event follows, with DSCpeak at 300 °C, associated with the degradation of the organic skeleton of the API.

Mirtazapine hemihydrate and α-lactose monohydrate

The FTIR spectrum of the MRTHH + Lact mixture (Fig. 2a) does not reveal the presence of new signals that cannot be attributed to either MRTHH or pure Lact. In addition to the signals of the API’s, a few peaks specific to the excipient [37] are worth mentioning, such as the sharp signal determined by the stretch of the O–H bonds (3522 cm−1), as well as the intense band seen at 1031 cm−1 associated with the stretching vibration of the C–O–C bonds.

The thermoanalytical profile of the mixture (Fig. 3a, c(C)) reveals three events determined by mass loss (better revealed on the DTG curve). The first two processes (DTGpeak = 116 and 149 °C) can easily be attributed to the dehydration of MRT and Lact, respectively. At higher temperatures however, the degradation of the API and that of the excipient are not seen separately, a single mass loss being identified up to 400 °C. The global experimental mass loss is slightly lower than the theoretical calculated one (83.6% vs. 89.1%) the difference being observed on the 150–400 °C temperature range. The DSC curve (Fig. 3e(C)) reveals the presence of the solid–liquid transition corresponding to MRTHH (DSCpeak = 115 °C). The calculated ΔmH value of 49.9 J g−1 was approximately equal to the theoretical expected value (the mixture contains MRTHH and Lact in a 1:1 mass ratio). Considering all the analyzed data, we can affirm that no incompatibilities were detected for the MRTHH + Lact mixture, at least up until 150 °C.

Mirtazapine hemihydrate and starch

Studying the FTIR spectrum of MRTHH + Sth (Fig. 2a), very few differences were revealed from MTRHH’s spectrum, those being easily attributed to the presence of Sth in the mixture [38, 39]. The wide band (3600–3070 cm−1) correlated to the stretching vibration of the O–H bonds is much more distinctive and has an increased intensity compared to the one seen for the API due to the presence of the excipient, and the characteristic α-1,4-glycosidic bond found in the structure of Sth determines a specific spectral signal seen at 997 cm−1 on the mixture’s spectrum. Thermal analysis confirmed the findings of the FTIR study, no interactions being revealed up until 150 °C. The TG/DTG profile of the MRTHH + Sth mixture (Fig. 3a, c(E)) contains the same degradation steps seen for the individual ingredients, a single DTGpeak shift being revealed from 254 °C (MRTHH) to 223 °C (MRTHH + Sth), suggesting that the presence of the excipient lowers the maximum temperature of the decomposition process. The residual mass of the sample (see Table 2) is slightly higher than the expected calculated value. The melting process of MRTHH is seen at 116 °C (Fig. 3e(E)), basically unchanged from the DSC curve of the pure API. The analysis of the endothermic signal revealed a ΔmH value of 48.8 J g−1, the slight decrease from the exact halved value seen for pure MRTHH can be attributed to the presence of the excipient in the sample (that can be seen as an impurity).

Mirtazapine hemihydrate and D-mannitol

The obtained MRTHH + Man FTIR spectrum (Fig. 2b) reveals a few aspects worth mentioning. As such, for the most part, the peaks associated with the API are revealed on the spectrum of the mixture, as well as several signals characteristic to the excipient [40] (shown in Table 1 with an Italic font), i.e., the stretching vibration of the hydroxyl groups (wide band with peaks at 3390 and 3285 cm−1), the deformation vibration of the C–H bond from –CH2 (1420 cm−1) and O–H bond from –CH2OH (1315 cm−1), as well as the out-of-phase stretching vibration of the C–C–O bonds from the primary and secondary alcohol moieties (1078 and 1018 cm−1). Some unexpected modifications were revealed however, such as the disappearance of MRTHH’s bands from 995, 648, 621, 584 and 426 cm−1 and the presence of peaks at 1560, 1541, 1508 and 1456 cm−1, the latter ones not corresponding to either pure MRTHH or Man.

During the TG/DTG evaluation (see Fig. 3a, c(G)), the degradation of MRTHH from the mixture occurred in a different manner than the one seen for the pure compound, namely the 125–250 °C thermal range associated with the decomposition of MRTHH is associated, in the case of the mixture, with three DTG peaks (223, 230 and 234 °C) in comparison to the single one seen at 254 °C for the pure API, indicating the possible presence of intermediary degradation products. In addition, the experimental mass loss is almost 100%, slightly elevated from the expected 98.4%, giving further indication of the formation of intermediary products that present a different degradation behavior than the pure components.

The DSC curve (Fig. 3e(G)) reveals both endothermic processes associated with the melting processes of the individual components, while revealing a small intensity exothermic event at 333 °C previously unseen for either mixture component. In addition, an increase in the melting enthalpy value of MRTHH was detected (see Table 2), potentially due to the formation of H-bonds between the excipient and the residual water molecules that have not yet left the hemihydrate API (since the melting point of the drug is relatively low and close to the dehydration process and according to literature a stable crystal structure contains two water molecules and four molecules of MRT [6]). The results of the performed thermal analysis coupled with the FTIR evaluation suggests the possibility of an interaction between MRTHH and Man that does not completely damage the integrity of the API but indicates the need for further evaluations on this mixture in order to confirm a definite incompatibility.

Mirtazapine hemihydrate and sorbitol

Considering the structural similarities between Man and Sorb, it was expected that the FTIR spectrum obtained for the MRTHH + Sorb mixture to present similar characteristics to the one discussed previously for MRTHH + Man. However, the overall characteristics of the MRTHH + Sorb spectrum were more indicative of a lack of interactions between the components of the mixture. The only notable aspect would be the appearance of peaks at 1560, 1541, 1508 and 1491 cm−1 not detected on the spectra of the pure compounds.

The overall degradation process revealed on the TG/DTG curves for the MRTH + Sorb mixture (Fig. 3a, c(I)) seems to be, at a first glance, the sum of the degradation steps of the pure components. However, both DTGpeak values for the decomposition of pure MRTHH and Sorb are shifted to lower temperatures, and an additional mass loss is observed with a corresponding DTGpeak at 196 °C, unaccounted by either the API or the excipient. Moreover, a difference regarding the experimental and theoretical mass loss was detected, the measured residual mass being 5.4%, a higher value than the expected 1.1%. The melting processes of the components of the mixture are seen well individualized and localized on the DSC curve of the mixture (Fig. 3e(H)). The melting enthalpy value calculated for the solid–liquid transition of MRTHH (Table 2) corresponds to the expected value. All in all, the obtained data indicate that at low temperature values the API remains approximately unperturbed by the presence of the excipient, but at temperatures exceeding 150 °C some interactions begin to appear, their exact nature requiring an additional, more detailed evaluation.

Mirtazapine hemihydrate and stearic acid

The FTIR spectrum of the MRTHH + StA mixture (Fig. 2c) reveals the definite presence of the excipient in the analyzed mixture, especially from the high intensity peaks seen at 2916 and 2849 cm−1 characteristic to the asymmetric and symmetric stretching vibration of the multiple C–H bonds present in the excipient’s structure, as well as from the sharp signal seen at 1717 cm−1 characteristic to the stretching vibration of the C = O bond from the carboxyl moiety. The wide band characterizing the stretch of the O–H bond of the excipient is no longer visible on the spectrum of the mixture, the spectrum presenting however with an additional spectral signal at 746 cm−1.

This observation is supported by the results of the thermal analysis (Fig. 3a, c, e(K)) that revealed a single degradation process on the TG curve with a corresponding DTGpeak situated at 261 °C, above that seen for MRTHH (254 °C) and the one observed for pure StA (256 °C). Furthermore, the DSC curve of the mixture does not display either the melting point of MRTHH (117 °C) or the one of StA (71 °C), the latter one being instead observed as an endothermic event at 59 °C. This DSCpeak can be explained as a shift of the melting process of StA due to the presence of MRTHH in the mixture, since StA has a 0.6 molar fraction in the studied binary mixture. The disappearance of the melting process of MRTHH indicates however that an interaction occurs between the components even at low temperatures.

Mirtazapine hemihydrate and magnesium stearate

In the case of the MRTHH + MgSt mixture, the FTIR spectrum (Fig. 2c) displays the characteristics of the components, the most intense spectral signals being correlated to the presence of the excipient [37]. These include the stretching vibrations of the C–H bonds (2916 and 2851 cm−1), as well as the vibrations associated with the carboxylate ion group (1574 and 1456 cm−1). No new bands appear on the spectrum, indicating the lack of interactions between MRTHH and MgSt at room temperature.

The thermoanalytical profile of the MRTHH + MgSt mixture (Fig. 3b, d, f(M)) follows a complex path, with multiple signals detected on the DTG curve, especially above 260 °C, most degradation steps being correlated with the decomposition of the excipient. Although MRTHH seems to completely degrade up until that temperature (aspect indicated by the mass loss value), the DTGpeak associated with the process shifts to lower values indicating that the presence of the excipient may lower the temperature at which the decomposition of the API takes place. The unperturbed presence of MRTHH in the mixture at low temperature values is supported by the DSCpeak seen at 115 °C correlated with a melting enthalpy value of 53,7 J g−1.

Mirtazapine hemihydrate and calcium lactate

The FTIR spectrum of the MRT + CaLact mixture (Fig. 2d) did not reveal any new bands that could be correlated with a possible room temperature interaction. The thermal behavior of the mixture (Fig. 3b, d, f(O)) corresponded with the ones seen for the pure ingredients, with two exceptions: the DTGpeak associated with the decomposition of MRTHH shifted from 254 °C (pure API) to 228 °C (binary mixture), as well as the DSC exothermic signal from 300 °C (pure MRTHH) to 284 °C (mixture). The melting process of MRTHH was revealed at 116 °C with a calculated enthalpy value of 48.7 J g−1. As in previously discussed cases, these observations may indicate that at temperature values above 130 °C the decomposition of MRTHH begins at lower temperatures in the presence of CaLact than the ones seen for the pure API. Because of these findings further studies are required for this mixture if, during the manufacturing or conditioning stages of pharmaceutical formulations containing MRTHH and CaLact, temperatures exceed 130 °C.

Mirtazapine hemihydrate and magnesium citrate

Figure 2d reveals the FTIR spectrum of the MRTHH + MgCit mixture. The spectral positions of the identified peaks (Table 1) did not unveil signals that could not be associated with either the API or the excipient and no significant disappearance of bands was observed. Figure 3b, d, f(Q) reveals that a DTGpeak shift can be observed in this case (from 254 °C for the API to 230 °C for the mixture), as previously seen for MRTHH + CaLact, and accompanied by a similar DSCpeak variation (from 300 °C for the API to 283 °C for the mixture). However, the overall mass loss corresponds to the calculated theoretical value (the residual mass of 33.9% being expected considering that on the 25–400 °C temperature range MgCit loses only 66.1% of its mass, while MRTHH suffers an almost complete degradation). The solid–liquid transition of MRTHH form the analyzed mixture is revealed on the DSC curve, as expected at 116 °C (melting enthalpy 53.1 J g−1) showing no signs of drug-excipient interactions.

Mirtazapine hemihydrate and PVP

The spectral analysis performed on the MRTHH + PVP mixture (Fig. 2e) did not raise any concerns regarding drug-excipient interactions, the identified bands belonging either to MRTHH or PVP. However, when subjected to thermal stress, the mixture presented an interesting behavior (Fig. 3b, d, f(S)), a single main decomposition being observed in the temperature range 150–350 °C, with a corresponding DTG signal at 208 °C. The melting point of pure MRTHH (seen for the API at 117 °C) shifted to a lower temperature value (114 °C) and presented with a reduced intensity (melting enthalpy value 38.6 J g−1), indicating that drug-excipient interactions are possible even at lower-than-expected temperatures. In addition, the overall mass loss of the mixture should have been almost 10% higher if the decomposition of the API and excipient would have occurred without any interactions, aspect that suggests that the degradation of MRTHH is influenced by the presence of PVP.

Mirtazapine hemihydrate and aerosil

The last studied binary mixture, MRTHH + SiO2, revealed a significantly different FTIR spectrum (Fig. 2e) than the ones seen previously for the discussed mixtures, an expected occurrence considering the structural particularities of aerosil. As such, the most intense spectral signals can be attributed to the excipient and are determined by the presence of the Si–O bond [36] (max. at 1074 cm−1). Due to the increased intensity of the excipient’s signals, MRTHH’s spectral characteristics are difficult to observe and a compatibility assessment cannot be made based solely on these data.

While, the overall progress of MRTHH’s degradation when incorporated in the binary mixture (Fig. 3b, d, f(U)) was not altered by the presence of the excipient (the extremely high thermal stability of aerosil has been well documented in literature [41]), MRTHH’s DTGpeak is shifted at 205 °C from the 254 °C value observed in the case of the pure API. The high percentage residual mass of the mixture (49.7%) is to be expected considering the 1:1 mass ratio used in the preparation and the total degradation of the individual components (calculated residual mass being 1.8% for MRTHH and 95.9% for aerosil). Because of the inorganic, high stability property of the excipient, no interactions were expected in theory. However, the DTGmax shift coupled with the barely visible DSCpeak determined by the melting process of the API (enthalpy value 26.9 J g−1), as well as its shift to 107 °C on the DSC curve of the mixture, gave reason to observe that the thermal behavior of MRTHH is altered by the presence of this thermally inert excipient.

Conclusions

The present study brings to light the compatibility evaluation between the atypical antidepressant mirtazapine and ten different pharmaceutical excipients. The results of the evaluation performed at room temperature for the binary mixtures MRTHH + excipient were correlated with the thermal analysis study (TG/DTG/DSC) and by comparison with the data obtained for each pure mixture component, several observations can be drawn. For the MRTHH + Lact, MRTHH + St, MRTHH + Sorb, MRTHH + MgSt, MRTHH + CaLact and MRTHH + MgCit mixtures no interactions were observed at room temperature, but the thermogravimetric analysis revealed that increased temperatures determine the alteration of the thermal behavior of MRTHH when found in association with the mentioned excipients. The presence of the excipient seems to interfere with the degradation path of the pure API at higher temperatures, as seen in the cases of MRTHH + Lact mixture (temperature values over 150 °C) and MRTHH + CaLact (temperature values over 130 °C). In the case of MRTHH + St, the excipient seems to determine an acceleration of the degradation process, lowering the temperature value of the main decomposition step. The thermal analysis of the MRTHH + Man mixture revealed that temperature values above 120 °C seem to determine the appearance of interactions between the components and intermediary degradation products may form, while the FTIR evaluation indicated that some changes may occur even at room temperature, without completely damaging the API’s integrity. A similar situation was revealed for the MRTHH + Sorb mixture, in the case of which, temperature values over 150 °C seem to determine the formation of intermediary degradation compounds that alter the thermal behavior of pure MRTHH. For the MRTHH + StA mixture, a slight indication of incompatibility was revealed by the results of the FTIR analysis and the presence of interactions was confirmed during thermal analysis even at low temperature values. The association of MRTHH and PVP revealed to be safe at room temperatures, however, over 100 °C precautions may need to be considered. The most unexpected result was obtained during the study of the MRTHH + SiO2 mixture, the thermal behavior of the API being altered in the presence of the excipient during the heating protocol. This was an unforeseen event since the excipient is well known for its thermal inertia.

In conclusion, the employed compatibility study has brought to light that precautions need to be considered during the manufacturing processes of drugs containing mirtazapine as an active ingredient since high temperatures may have an unexpected detrimental effect on the stability of the compound and its following therapeutic activity when associated with several pharmaceutic excipients.