As cocrystal formation cannot yet be accurately predicted by computer modeling and the selection of a suitable cocrystal former (CCF) from thousands of substances presents a challenge in crystal engineering. Therefore, methods to enhance the efficiency of cocrystal screening have received considerable attention. Among them, thermal screening methods, including differential scanning calorimetry (DSC) and Hot-Stage Microscope (HSM), have been proposed as rapid solvent free methods in cocrystal screening. Cocrystal may exhibit stoichiometric diversity, i.e. an active pharmaceutical ingredient (API) can cocrystallize with CCF under two or more stoichiometric compositions, which may exhibit different physicochemical properties and therefore influence product quality and pharmaceutical performance, including oral bioavailability. However, screening methods capable of determining or predicting stoichiometrically diverse co-crystals are still limited. The aim of the present study is to establish a thermal method to screen and determine cocrystals with stoichiometric diversity. We proposed that cocrystals can be formed from the eutectic melt, and thermal method is suitable for screening stoichiometrically diverse cocrystals. To test our hypothesis, Salicylic acid (SAA), which is a monohydroxybenzoic acid, is selected as a model API in this study. Eighteen API-CCF binary mixtures were tested on DSC. If the two components are miscible, a eutectic melting event will appear below the melting endotherms of SAA and CCF. Otherwise, only individual endothermic melting peaks of SAA and CCF will be observed. In the miscible system, if a cocrystal is formed, one or more endotherms will appear after eutectic melting point. Otherwise, only a single eutectic melting peak will be observed. Heating rate during DSC experiments influences the rate and extent of cocrystals formation. We observed that the resolution of three endothermic peaks attributed to eutectic and two cocrystals melting could be enhanced by using a lower heating rate (e.g., 1℃/min vs. 10 ℃/min). To support above results generated from DSC, hot-stage microscope (HSM) were utilized where samples were prepared by Kofler mixed fusion method. The behavior of binary mixture on HSM was consistent with DSC results. Thermal screening methods showed that among of eighteen API-CCF binary mixtures, five systems may form new cocrystals. SAA-3-nitrobenzamide (3NBA), SAA-benzamide (BA), and SAA-isonicotinamdie (INA) potentially exhibited stoichiometric diversities which had not been reported previously. Cocrystals with different molar ratios of SAA-3NBA, SAA-BA and SAA-INA systems were synthesized by liquid-assisted grinding method (LAG), and further characterized by DSC, powder X-ray diffraction (PXRD), and Fourier transform infrared spectroscopy (FTIR). DSC profile shows a single endotherm for each cocrystal system, corresponding to the melting point of individual synthesized cocrystal. The endothermic peaks of two cocrystals with different stoichiometric ratios prepared by LAG are consistence with the endotherms observed on DSC curves of responding binary systems. PXRD results showed unique peaks in SAA-3NBA cocrystal at 1:1 molar ratio, SAA-BA cocrystals at 1:2 and 1:1 molar ratios, and SAA-INA cocrystals at 1:1 and 2:1 molar ratios. Variable-temperature powder X-ray diffraction (VT-PXRD) confirmed that cocrystals with different molar ratios can be formed after eutectic melting. FTIR results showed that there were broad stretches around 1900-2500 cm-1 , suggesting that OH…N hydrogen bond has been formed in different cocrystals. Single crystal structures of three synthesized new cocrystals including SAA-3NBA (1:1), SAA-BA (1:2) and SAA-INA (2:1) has been elucidated. Taken previous reported two cocrystals of SAA-BA and SAA-INA at 1:1 molar ratio into consideration, a thermal method using DSC has been successfully developed for screening and determination of stoichiometric diverse cocrystals using SAA as model API, BA and INA as CCFs.

由於計算機不能有效地預測藥物共晶的形成，在成千上萬的共晶形成物 (CCF)中篩選出合適的 CCF，是一項非常具有挑戰性的工作。因此，提高共晶篩 選的效率成為共晶研究的熱點。熱篩選方法，包括差示掃描量熱法(DSC)和熱台 顯微鏡(HSM)，優點在於快速和不需要使用有機溶劑。然而，共晶也可能表現出 化學計量多樣性，大部份不同摩爾比的共晶都是通過不同結晶方法意外地被發現。 這類共晶會可能表現出不同的理化性質，影響藥物質量和口服生物利用度。可是 到目前為止仍然沒有一個有效的方法可以篩選共晶的化學計量多樣性。本研究的目的是建立一個熱力學篩選的方法，確定活性藥用成分(API)與 CCF 是否存在多個摩爾比。為了驗証上述的假設，在這項研究中以水楊酸(SAA) 作為模型藥物。18 種 API 與 CCF 的物理性混合物在 DSC 測試中，發現可以混 溶的 API 與 CCF 會形成共熔點，而且共熔點都會是低於 API，CCF 及共晶。相 反，如果兩種成份不混溶，在熱力曲線中可以同時觀察到 API 與 CCF 的吸熱峰， 沒有共熔點的形成。可見, API 與 CCF 可以混溶是形成共晶的基本條件。能夠形 成共晶的 API 與 CCF，在 DSC 篩選中，兩個或以上的吸熱峰會在共熔點後出現。 否則，將只有單一共熔點的吸熱峰形成。研究結果還表明，在 DSC 篩選過程中， 加熱速率會影響共晶形成的速度和程度。低加熱速率(如 1℃/min)條件會提高吸 熱峰之間的解析度，成功分離出三個吸熱峰，分別是共熔點和兩種共晶熔點所引 致，而這兩種共晶可能就是不同摩爾比的共晶。為了驗證上述的結果，利用熱台 顯微鏡(HSM)進一步分析。樣品以 Kolfer 熱熔混合方式來製備。在 HSM 下以相 同的升溫速度觀測共晶的熱力學性能.結果顯示與 DSC 的結果是一致的，成功找 出共熔點和兩種共晶熔點。在整個篩選方法中，發現十八種 API 與 CCF 的混合 物中，其中五種屬於新共晶，如：SAA-3-硝基苯甲醯胺(3NBA)，另外兩個兩種 混合物，SAA與苯甲醯胺(BA)和SAA 與異煙醯胺(INA)存在著化學計量多樣性。 以下部分將側重於 SAA-3NBA，SAA-BA 和 SAA-INA 進行詳細研究。首先不同 摩爾比同的 API 與 CCF 利用液態輔助磨法(LAG)合成共晶，並進一步通過 DSC， 粉末 X 射線衍射法(PXRD)，傅裡葉紅外光譜(FTIR)來進行表徵。制備的共晶都 顯現出單一的吸熱峰，而且可以與篩選時發現的吸熱峰相對應。另外，PXRD 結 果表明，1:1 摩爾比的 SAA-3NBA，1:2 和 1:1 摩爾比的 SAA-BA，及 1:1 和 2:1 摩爾比的SAA-INA共晶出現新的特徵峰。在變溫粉末X射線衍射法（VT-PXRD） 對三種物理性混合物進行檢測，可確定不同摩爾比的共晶同時在共熔點後形成， 當到達其中一種共晶熔點後，會轉變成其他摩爾比的共晶。 FTIR 結果表明，在 1900 至 2500 cm -1 的波長之間，有廣闊綿延的峰出現，表明共晶間具有 OH…N的氫鍵鏈接。共晶的單晶體結構解析結果进一步證明确实有三種新共晶形成。 本研究中以水楊酸作為模型藥物，成功地利用熱力學方法包括差示掃描量熱 法和熱台顯微鏡法建立了化學計量比例多樣性共晶的篩選。 關鍵詞： 熱力學方法；篩選；差示掃描量熱法；熱台顯微鏡；化學計量多 樣性