# Links between chapters: Solid state stability

Pairs are written once. The direction runs from the finding that needs the other, or whose papers are later, to the finding that supplies its basis, data or caution. The four chapters are [Crystallization tendency](crystallization-tendency/index.md), [Miscibility and drug-polymer interactions](miscibility-and-drug-polymer-interactions/index.md), [PEG dispersions](peg-dispersions/index.md) and [Physical stability](physical-stability/index.md). Relations: builds on, explains, extends, same mechanism, conditions differ, contradicts. Citation keys are in [sources.md](sources.md). Reasoning that goes beyond the cited chapter findings is marked `(synthesis)`.

| from | to | relation | why |
|---|---|---|---|
| [Crystallization tendency F8](crystallization-tendency/findings.md#f8) | [Miscibility and drug-polymer interactions F1](miscibility-and-drug-polymer-interactions/findings.md#f1) | builds on | Both read the indomethacin carboxylic acid by IR: in solution a dimer band grows with concentration even below saturation [towler2007 R13, R17], and in the PVP dispersions the dimer band falls as the acid OH bonds to the PVP carbonyl [taylor1997 R10, C35]. |
| [Crystallization tendency F19](crystallization-tendency/findings.md#f19) | [Miscibility and drug-polymer interactions F10](miscibility-and-drug-polymer-interactions/findings.md#f10) | builds on | The felodipine polymer rows of the similar-factor finding are the 3% growth ratios of the polymer-chemistry study, PVP 10.0, PVP/VA 6.5, HPMCAS 4.4 and PVAc 1.2 at 70 °C [kestur2010a R6]. |
| [Crystallization tendency F19](crystallization-tendency/findings.md#f19) | [Miscibility and drug-polymer interactions F13](miscibility-and-drug-polymer-interactions/findings.md#f13) | conditions differ | PVP lowered the growth of five donor drugs by a similar 3–14-fold [trasi2014a R6, X6], but for the acceptor-only bifonazole PAA gave about 15-fold at 18 °C against about 5.9 for PVP [kestur2011 R7]; a similar factor holds for one polymer across donor drugs, not across polymer chemistries. |
| [Miscibility and drug-polymer interactions F9](miscibility-and-drug-polymer-interactions/findings.md#f9) | [Crystallization tendency F6](crystallization-tendency/findings.md#f6) | same mechanism | Mixing with PVP changed the Hoffman free energy of crystallization of nifedipine only from about −9.0 to −7.95 kJ/mol [marsac2006b R12], and the larger driving force of the vemurafenib α form did not decide which polymorph grew [lu2016 C52]; both read the differences in crystallization as not set by the driving force. |
| [Miscibility and drug-polymer interactions F9](miscibility-and-drug-polymer-interactions/findings.md#f9) | [Crystallization tendency F17](crystallization-tendency/findings.md#f17) | extends | Dispersions are mostly supersaturated and their stability is inferred to be kinetic [marsac2009 C34]; the neat-drug class, a kinetic property of the melt, is the baseline against which that protection is read [baird2010 C46] (synthesis). |
| [Miscibility and drug-polymer interactions F10](miscibility-and-drug-polymer-interactions/findings.md#f10) | [Crystallization tendency F2](crystallization-tendency/findings.md#f2) | builds on | The felodipine form II growth maximum near 105 °C [kestur2010a R5] is where the polymer ratios fall to about 1.1 at 110 °C for all but PVP (2.9) [kestur2010a R6, X2]; the inhibition is largest at the supercooling where the neat rate is lowest. |
| [Miscibility and drug-polymer interactions F11](miscibility-and-drug-polymer-interactions/findings.md#f11) | [Crystallization tendency F12](crystallization-tendency/findings.md#f12) | same mechanism | Neither the dispersion Tg (42.7–44.1 against 43.3 °C) nor the pure-polymer Tg explained the polymer order [kestur2010a R1, C31, C34], and Tg/Tm did not separate the melt classes either [baird2010 R14, C37]. |
| [Miscibility and drug-polymer interactions F11](miscibility-and-drug-polymer-interactions/findings.md#f11) | [Crystallization tendency F19](crystallization-tendency/findings.md#f19) | same mechanism | PVP raised the Tg of the five donor drugs by only about 2–3 °C while lowering growth 3–14-fold [trasi2014a R6, R7], and 3 wt% polymer left the felodipine Tg at 42.7–44.1 °C against 43.3 °C [kestur2010a R1]. |
| [Miscibility and drug-polymer interactions F12](miscibility-and-drug-polymer-interactions/findings.md#f12) | [Crystallization tendency F19](crystallization-tendency/findings.md#f19) | same mechanism | Both attribute the polymer effect on growth to hydrogen bonding of the polymer to the drug donor, measured in spin-coated films and not at growth conditions: felodipine NH at 3294 cm⁻¹ with PVP [kestur2010a R9, C26], nilutamide with PVP against PAA [trasi2014b C19]. |
| [Miscibility and drug-polymer interactions F12](miscibility-and-drug-polymer-interactions/findings.md#f12) | [Physical stability F14](physical-stability/findings.md#f14) | builds on | The felodipine NH band at 3294 cm⁻¹ with PVP [kestur2010a R9] is the drug–PVP hydrogen bond seen in the dry felodipine–PVP films, whose NH band persisted from 0 to 180 °C [marsac2010 R2, R4]. |
| [Miscibility and drug-polymer interactions F14](miscibility-and-drug-polymer-interactions/findings.md#f14) | [Crystallization tendency F19](crystallization-tendency/findings.md#f19) | explains | Nimesulide holds its NH in an intramolecular bond that diminishes drug–polymer bonding and flattens the differences between polymers [kestur2011 C27], which is why its polymer factors in the similar-factor comparison are only 2–10-fold [kestur2011 R9]. |
| [Miscibility and drug-polymer interactions F15](miscibility-and-drug-polymer-interactions/findings.md#f15) | [Crystallization tendency F14](crystallization-tendency/findings.md#f14) | builds on | Both score spin-coated films by polarized light over 7 days of dry storage: 200 µL spun for 20 s gave films about 1 µm thick [vaneerdenbrugh2010a P2, P8], and the polymer screen applied the method to drug–polymer solutions at seven ratios [vaneerdenbrugh2010b P1, P2]. |
| [Miscibility and drug-polymer interactions F15](miscibility-and-drug-polymer-interactions/findings.md#f15) | [Crystallization tendency F17](crystallization-tendency/findings.md#f17) | extends | Class I drugs had the lowest mean amorphous index across seven polymers (47–59%) [vaneerdenbrugh2010b R4], but no polymer was best for every drug: PAA and PSSA, lowest overall, were best for the amides and worst for the acids [vaneerdenbrugh2010b R3, C20]. |
| [Miscibility and drug-polymer interactions F16](miscibility-and-drug-polymer-interactions/findings.md#f16) | [Crystallization tendency F14](crystallization-tendency/findings.md#f14) | same mechanism | Slower solvent removal crystallizes more: slow evaporation left 22 of 31 class III combinations fully crystalline [vaneerdenbrugh2010a R12, R13], and 50/50 rotary-evaporated powders showed crystallization with PAA that the films did not [vaneerdenbrugh2010b C43]. |
| [Miscibility and drug-polymer interactions F17](miscibility-and-drug-polymer-interactions/findings.md#f17) | [Crystallization tendency F17](crystallization-tendency/findings.md#f17) | extends | Complementary groups say which polymer protects, the neat-drug class how much protection is needed (synthesis): a complementary polymer did not guarantee stability, since the 50% lumefantrine–HPMCAS dispersion crystallized within a week [hiew2022a C63], and the class is likewise only a first flag [baird2010 C46]. |
| [Miscibility and drug-polymer interactions F17](miscibility-and-drug-polymer-interactions/findings.md#f17) | [Crystallization tendency F19](crystallization-tendency/findings.md#f19) | conditions differ | PVP slowed five donor drugs by a similar factor [trasi2014a R6, C10], while the order of polymers reversed with the drug, PAA being best for the acceptor-only bifonazole [kestur2011 R7, C27]. |
| [Miscibility and drug-polymer interactions F18](miscibility-and-drug-polymer-interactions/findings.md#f18) | [Physical stability F14](physical-stability/findings.md#f14) | builds on | Heating the itraconazole–HPMC films for 2 min at 165 °C mixed the domains [purohit2015a R1, R4, R12], after the felodipine–PVP films whose water-induced separation was only partly reversed by 1500 min at 125 °C [marsac2010 R17, C47]. |
| [Miscibility and drug-polymer interactions F21](miscibility-and-drug-polymer-interactions/findings.md#f21) | [Crystallization tendency F20](crystallization-tendency/findings.md#f20) | same mechanism | In both an ionic contact of a base was proposed but not measured: lumefantrine was 19–55% protonated by four enteric polymers [hiew2022a R20, C47], and for propranolol and nicardipine salts crystallization and moisture uptake did not follow counterion pKa [towler2008 R12, C34, C25]. |
| [Miscibility and drug-polymer interactions F23](miscibility-and-drug-polymer-interactions/findings.md#f23) | [Crystallization tendency F17](crystallization-tendency/findings.md#f17) | conditions differ | Class III glasses showed only edge crystallization after 3 months at 0% RH [baird2010 R25], whereas lumefantrine with PVPVA crystallized after 1 to 8 weeks at 40 °C/75% RH [hiew2022a R6]. |
| [Miscibility and drug-polymer interactions F24](miscibility-and-drug-polymer-interactions/findings.md#f24) | [Crystallization tendency F17](crystallization-tendency/findings.md#f17) | builds on | The 2011 screen chose 15 readily crystallizing drugs [vaneerdenbrugh2011a C16], extending the 8-drug screen in which class I drugs had the lowest amorphous index [vaneerdenbrugh2010b R4]. |
| [Miscibility and drug-polymer interactions F24](miscibility-and-drug-polymer-interactions/findings.md#f24) | [Crystallization tendency F19](crystallization-tendency/findings.md#f19) | same mechanism | Acceptor-only polymers (E100, PVP, PVPVA) gave amorphicity indices of about 54–100 for the donor drugs [vaneerdenbrugh2011a R3], and PVP slowed five donor drugs 3–14-fold in the melt [trasi2014a R6, X6]. |
| [Miscibility and drug-polymer interactions F26](miscibility-and-drug-polymer-interactions/findings.md#f26) | [Crystallization tendency F9](crystallization-tendency/findings.md#f9) | builds on | Lumefantrine is class III in the heat–cool–heat cycle [hiew2022a R3, trasi2020 R5], and its PVPVA dispersions still crystallized after 1 to 8 weeks [hiew2022a R6]. |
| [PEG dispersions F1](peg-dispersions/findings.md#f1) | [Miscibility and drug-polymer interactions F17](miscibility-and-drug-polymer-interactions/findings.md#f17) | same mechanism | Drugs with hydrogen-bond donors had the highest eutectic compositions in PEG 3350, 30–40 wt% for chlorpropamide, aceclofenac, benzocaine and ibuprofen [baird2011 R1; zhu2011 R5.b, R5.d], as the polymer whose acceptor matched the drug donor protected best in amorphous polymers [kestur2010a R6]. |
| [PEG dispersions F2](peg-dispersions/findings.md#f2) | [Miscibility and drug-polymer interactions F5](miscibility-and-drug-polymer-interactions/findings.md#f5) | builds on | The athermal Flory–Huggins model (no interaction energy) is the entropy-only case of the nifedipine–PVP lattice [marsac2006b E1, R1]; it predicted PEG eutectics of 12.8 and 1.5 wt% for chlorpropamide and aceclofenac against 35 and 30 measured [baird2011 R1]. |
| [PEG dispersions F2](peg-dispersions/findings.md#f2) | [Miscibility and drug-polymer interactions F6](miscibility-and-drug-polymer-interactions/findings.md#f6) | same mechanism | A model without specific interactions failed for drugs with donor groups: Fedors χ = 2.0 for nifedipine–PVP against −3.8 from melting point depression [marsac2006b R4, R9], athermal PEG eutectics 2.7 and 20 times too low (synthesis: 35/12.8 and 30/1.5) [baird2011 R1, C14]. |
| [PEG dispersions F3](peg-dispersions/findings.md#f3) | [Miscibility and drug-polymer interactions F1](miscibility-and-drug-polymer-interactions/findings.md#f1) | same mechanism | Both read a drug–polymer bond from a spectrum compared with the physical or additive mixture: loss of the indomethacin dimer band and a 1636 cm⁻¹ band absent from physical mixes [taylor1997 R10, R12], and chlorpropamide carbonyl at 1717 against 1663 cm⁻¹ in the additive spectrum [baird2011 R10.a]. |
| [PEG dispersions F3](peg-dispersions/findings.md#f3) | [Miscibility and drug-polymer interactions F2](miscibility-and-drug-polymer-interactions/findings.md#f2) | same mechanism | The PVP carbonyl shift ranked donor drugs (45 cm⁻¹ indomethacin, 25 felodipine, 0 ketoconazole) [marsac2009 R12, R14], and the drug carbonyl band in molten PEG moved by 16 cm⁻¹ for aceclofenac (1744 against 1728) and 54 cm⁻¹ for chlorpropamide (1717 against 1663) [baird2011 R10.a, R10.c] (synthesis: 1744 − 1728 and 1717 − 1663). |
| [PEG dispersions F3](peg-dispersions/findings.md#f3) | [Miscibility and drug-polymer interactions F4](miscibility-and-drug-polymer-interactions/findings.md#f4) | conditions differ | Ketoconazole shows no PVP carbonyl shift and χ near zero [marsac2009 R3, R12.h], yet its activity coefficient in PEG 400 is 0.09 while its interaction parameter in PEG 3350 is positive [marsac2008b R14.f; zhu2011 R5.f]. |
| [PEG dispersions F3](peg-dispersions/findings.md#f3) | [Miscibility and drug-polymer interactions F6](miscibility-and-drug-polymer-interactions/findings.md#f6) | same mechanism | Different routes to the interaction give different answers: a group-contribution estimate gave χ = 2.0 for nifedipine–PVP against −3.8 from melting point depression [marsac2006b R4, R9], and the PEG readings agree in sign but not in size, chlorpropamide giving −2.53 and −0.50 [baird2011 R9.a; zhu2011 R5.c]. |
| [PEG dispersions F3](peg-dispersions/findings.md#f3) | [Miscibility and drug-polymer interactions F7](miscibility-and-drug-polymer-interactions/findings.md#f7) | builds on | The melting point depression route to χ, developed for drugs in PVP [marsac2006b R9], was applied to drugs in PEG 3350 [zhu2011 E1]. |
| [PEG dispersions F3](peg-dispersions/findings.md#f3) | [Miscibility and drug-polymer interactions F8](miscibility-and-drug-polymer-interactions/findings.md#f8) | same mechanism | Routes agree in sign but not in size: PVP χ from melting and from solubility in EP differ by up to 1.01 [marsac2009 R7, R9, X7], and chlorpropamide in PEG gave −2.53 and −0.50 in two papers [baird2011 R9.a; zhu2011 R5.c]. |
| [PEG dispersions F3](peg-dispersions/findings.md#f3) | [Miscibility and drug-polymer interactions F17](miscibility-and-drug-polymer-interactions/findings.md#f17) | same mechanism | Four readings of interaction with the PEG ether oxygen singled out the drugs with hydrogen-bond donors, chlorpropamide, aceclofenac, benzocaine and ibuprofen [baird2011 R9, R10; zhu2011 R5, R20], as donor drugs paired with acceptor polymers in the melt-growth designs [kestur2010a R6]. |
| [PEG dispersions F3](peg-dispersions/findings.md#f3) | [Physical stability F12](physical-stability/findings.md#f12) | builds on | The additive-spectrum procedure for the drug carbonyl band in molten PEG [baird2011 P4] is the calculated physical-mixture comparison of the moisture-induced immiscibility study [rumondor2009c P4, E7]. |
| [PEG dispersions F4](peg-dispersions/findings.md#f4) | [Miscibility and drug-polymer interactions F9](miscibility-and-drug-polymer-interactions/findings.md#f9) | same mechanism | An estimated solubility in the polymer ranks drugs but does not say whether the dispersion stays amorphous: 0.010–0.134 weight fraction in PVP with stability inferred to be kinetic [marsac2009 R9, C34], and benzocaine and ibuprofen below the PEG estimate crystallizing within hours [zhu2012 R23.a, X13]. |
| [PEG dispersions F6](peg-dispersions/findings.md#f6) | [Crystallization tendency F9](crystallization-tendency/findings.md#f9) | builds on | Haloperidol (class I-A) crystallized at once and ketoconazole (class III) stayed amorphous [zhu2011 R9, R17], the opposite ends of the neat-drug classes [baird2010 R1, M17, M42] (synthesis: the interaction, not the class, decided incorporation). |
| [PEG dispersions F6](peg-dispersions/findings.md#f6) | [Miscibility and drug-polymer interactions F11](miscibility-and-drug-polymer-interactions/findings.md#f11) | same mechanism | A drug Tg up to 44 °C did not change the PEG pattern [zhu2011 C28, C31], and dispersion Tg and pure-polymer Tg did not explain the polymer order for felodipine [kestur2010a R1, C34]. |
| [PEG dispersions F8](peg-dispersions/findings.md#f8) | [Crystallization tendency F9](crystallization-tendency/findings.md#f9) | builds on | Neat haloperidol and benzocaine crystallized within seconds and neat ibuprofen and fenofibrate in only 10 and 17 of 50 samples after 3 weeks [zhu2012 R1, R2, R3], the order of the DSC classes [baird2010 R1, R3] (synthesis). |
| [PEG dispersions F8](peg-dispersions/findings.md#f8) | [Crystallization tendency F17](crystallization-tendency/findings.md#f17) | conditions differ | The neat-drug order (seconds against weeks) did not carry into the 20/80 comelts, where ibuprofen and fenofibrate began after about 100 and 120–180 min [zhu2012 R23.a], and the authors state that the link between neat and dispersion crystallization breaks down [zhu2012 C41]. |
| [PEG dispersions F9](peg-dispersions/findings.md#f9) | [Crystallization tendency F17](crystallization-tendency/findings.md#f17) | explains | Crystalline PEG acts as a heterogeneous nucleation site for ibuprofen and fenofibrate, a crystal growing at the PEG interface within minutes [zhu2012 R15, R19, C4], which is the reason given for the broken link between class and dispersion [zhu2012 C41]. |
| [PEG dispersions F9](peg-dispersions/findings.md#f9) | [Crystallization tendency F21](crystallization-tendency/findings.md#f21) | same mechanism | A crystalline interface starts crystallization of a slow crystallizer [zhu2012 R15, R19], as a free surface speeds growth of felodipine about 6-fold at 18 °C [kestur2013a R8, R9] (synthesis: both are interfaces that bypass the bulk barrier). |
| [PEG dispersions F10](peg-dispersions/findings.md#f10) | [Miscibility and drug-polymer interactions F12](miscibility-and-drug-polymer-interactions/findings.md#f12) | same mechanism | Hydrogen-bonding drugs slowed PEG spherulite growth most at 40 °C, 1 µm/s with benzocaine and 3 µm/s with ibuprofen against 10 µm/s without drug [zhu2012 R22], the same ordering by hydrogen bonding that ranked polymers for felodipine [kestur2010a R6, R9]. |
| [Physical stability F1](physical-stability/findings.md#f1) | [Crystallization tendency F3](crystallization-tendency/findings.md#f3) | builds on | Nucleation windows are separate from growth windows in neat drugs, acetaminophen nucleating only at 50–60 °C while its growth was measured at 70–110 °C [trasi2012a C22, R6], and a few percent polymer acts on nucleation, 3% lowering felodipine nucleation from about 1.3 × 10⁵ to 2.3–3.0 × 10³ m⁻³ s⁻¹ [konno2006 R4]. |
| [Physical stability F1](physical-stability/findings.md#f1) | [Crystallization tendency F19](crystallization-tendency/findings.md#f19) | extends | A polymer lowers nucleation by about two orders of magnitude at a few percent [marsac2006a R16, C18], where it lowers melt growth by about 3–14-fold [trasi2014a R6, X6]. |
| [Physical stability F1](physical-stability/findings.md#f1) | [Miscibility and drug-polymer interactions F10](miscibility-and-drug-polymer-interactions/findings.md#f10) | conditions differ | Dry nucleation of felodipine was equal for PVP, HPMC and HPMCAS within error [konno2006 C14], whereas melt growth was ordered PVP > PVP/VA > HPMCAS > PVAc [kestur2010a R6]. |
| [Physical stability F1](physical-stability/findings.md#f1) | [Miscibility and drug-polymer interactions F15](miscibility-and-drug-polymer-interactions/findings.md#f15) | conditions differ | Dry nucleation was equal for the three polymers [konno2006 C14], while the spin-coated screen gave different mean amorphicity indices (PVP and PVPVA 70, HPMCAS 62, PAA 49) [vaneerdenbrugh2010b R5]. |
| [Physical stability F1](physical-stability/findings.md#f1) | [PEG dispersions F9](peg-dispersions/findings.md#f9) | conditions differ | Amorphous polymers lowered the nucleation of felodipine by 1.5 orders of magnitude at 3% [konno2006 C13, C37], while crystalline PEG nucleated ibuprofen and fenofibrate at its interface [zhu2012 R15, R19]. |
| [Physical stability F2](physical-stability/findings.md#f2) | [Crystallization tendency F6](crystallization-tendency/findings.md#f6) | same mechanism | Nifedipine has the larger enthalpic driving force than felodipine (9.6 against 6.2 kJ/mol at 25 °C) and nucleates faster [marsac2006a R4.k, C15], while for indomethacin against felodipine and the vemurafenib polymorphs the driving force pointed the wrong way [lu2016 C52]. |
| [Physical stability F2](physical-stability/findings.md#f2) | [Crystallization tendency F17](crystallization-tendency/findings.md#f17) | builds on | The nucleation order in PVP, nifedipine before felodipine [marsac2006a R16, C19], is the order of the quenched glasses stored at 0% RH, nifedipine complete by 3 months and felodipine only at the edges [baird2010 R24, R25]. |
| [Physical stability F2](physical-stability/findings.md#f2) | [Crystallization tendency F18](crystallization-tendency/findings.md#f18) | same mechanism | The ranking of nifedipine above felodipine carried into solution, where nifedipine crystallized just below the coexistence curve and felodipine underwent liquid–liquid phase separation [raina2014a R20, C31], and into water-containing PVP, where nifedipine was reasoned to be more supersaturated [marsac2008a C49] (synthesis). |
| [Physical stability F2](physical-stability/findings.md#f2) | [Miscibility and drug-polymer interactions F9](miscibility-and-drug-polymer-interactions/findings.md#f9) | builds on | The Hoffman estimate of the crystallization free energy used for nifedipine and felodipine in PVP [marsac2006a R4.k] is the one applied to the 30 and 40% PVP mixtures [marsac2006b R12]. |
| [Physical stability F3](physical-stability/findings.md#f3) | [Crystallization tendency F1](crystallization-tendency/findings.md#f1) | builds on | The steep fall of felodipine growth on cooling, 3.5 × 10⁻⁹ m/s at 70 °C against 1.7 × 10⁻⁷ at 105 °C [kestur2010b R7], is the single-drug case of the steep temperature dependence seen across drugs [kestur2010a R5.a, C10]. |
| [Physical stability F3](physical-stability/findings.md#f3) | [Crystallization tendency F2](crystallization-tendency/findings.md#f2) | builds on | The growth maximum of pure felodipine near 105 °C [kestur2010b R7] is the reference against which the PVP inhibition widens toward 70 °C [kestur2010b R8]. |
| [Physical stability F3](physical-stability/findings.md#f3) | [Crystallization tendency F19](crystallization-tendency/findings.md#f19) | extends | PVP K-29/32 lowered felodipine growth log-linearly with concentration up to 4.5 wt% and more at lower temperature [kestur2010b R14, R15], quantifying the polymer factor that was a single ratio per drug for five drugs [trasi2014a R6]. |
| [Physical stability F3](physical-stability/findings.md#f3) | [Miscibility and drug-polymer interactions F10](miscibility-and-drug-polymer-interactions/findings.md#f10) | extends | The same felodipine–PVP system, with three PVP grades and 0.5–4.5 wt% [kestur2010b R8, R14], adds the concentration law and the chain-length effect to the polymer order at 3% [kestur2010a R6]. |
| [Physical stability F3](physical-stability/findings.md#f3) | [Miscibility and drug-polymer interactions F12](miscibility-and-drug-polymer-interactions/findings.md#f12) | same mechanism | Both explain the weaker inhibition at high temperature by drug–PVP hydrogen bonds that weaken on heating [kestur2010b C35, C37; kestur2010a C43, C44]. |
| [Physical stability F4](physical-stability/findings.md#f4) | [Crystallization tendency F6](crystallization-tendency/findings.md#f6) | same mechanism | The Flory–Huggins activity of nifedipine and felodipine in PVP stayed at about 0.96 at a polymer volume fraction of 0.1 [marsac2006a R19, C47], and the driving force explained only part of the differences between drugs [trasi2014a R11]. |
| [Physical stability F4](physical-stability/findings.md#f4) | [Miscibility and drug-polymer interactions F9](miscibility-and-drug-polymer-interactions/findings.md#f9) | same mechanism | Both estimate that a few percent PVP lowers the driving force very little (nifedipine −9.0 to −8.4 kJ/mol with 30% PVP) [marsac2006b R12] and read the inhibition as kinetic [marsac2006a C47]. |
| [Physical stability F4](physical-stability/findings.md#f4) | [Miscibility and drug-polymer interactions F12](miscibility-and-drug-polymer-interactions/findings.md#f12) | extends | The kinetic route through drug–PVP hydrogen bonds that lower drug mobility [kestur2010b C35, C42] is tied to the hydrogen-bond order PVP > PVP/VA > HPMCAS ≥ PVAc [kestur2010a C26, C42]. |
| [Physical stability F5](physical-stability/findings.md#f5) | [Crystallization tendency F5](crystallization-tendency/findings.md#f5) | same mechanism | Melt viscosity rose 10–20-fold with PVP K-90 without matching inhibition [kestur2010b C29], and within classes viscosity did not rank compounds [baird2012 C11, C16]. |
| [Physical stability F5](physical-stability/findings.md#f5) | [Crystallization tendency F12](crystallization-tendency/findings.md#f12) | same mechanism | Tg-scaled and bulk mobility measures failed to rank polymers [konno2006 R6, R4, C26], as Tg/Tm failed to rank the melt classes [baird2010 R14, C37]. |
| [Physical stability F5](physical-stability/findings.md#f5) | [Crystallization tendency F21](crystallization-tendency/findings.md#f21) | same mechanism | Felodipine crystal growth below Tg outpaced bulk mobility, which the authors attribute to growth at the free surface of uncovered films [rumondor2010 C28, C29], the surface effect of the surface–bulk finding [kestur2013a R8, R9]. |
| [Physical stability F5](physical-stability/findings.md#f5) | [Miscibility and drug-polymer interactions F11](miscibility-and-drug-polymer-interactions/findings.md#f11) | same mechanism | Dispersion Tg at 3 wt% polymer and at up to 25% did not rank polymers [kestur2010a R1, C31; konno2006 R6, R4]. |
| [Physical stability F5](physical-stability/findings.md#f5) | [PEG dispersions F6](peg-dispersions/findings.md#f6) | same mechanism | Neither the Tg of a PEG-dispersed drug (33–44 °C) nor the dispersion Tg of PVP dispersions explained the crystallization outcome [zhu2011 C28, C31; konno2006 R6, R4]. |
| [Physical stability F6](physical-stability/findings.md#f6) | [PEG dispersions F11](peg-dispersions/findings.md#f11) | same mechanism | Plasticization by water is tied to the hygroscopicity of the polymer [konno2008 C20], and in PEO compacts water speeds the loss of drug crystallinity by plasticizing the amorphous PEO [marsac2008b C24, C25, C26]. |
| [Physical stability F6](physical-stability/findings.md#f6) | [PEG dispersions F12](peg-dispersions/findings.md#f12) | extends | Pure PVP held about 33 g water per 100 g at 75% RH [konno2008 R8], and PEO held about 10 wt% water at 94% RH after hours but about 37 wt% after 5 months [marsac2008b R11]. |
| [Physical stability F7](physical-stability/findings.md#f7) | [Miscibility and drug-polymer interactions F1](miscibility-and-drug-polymer-interactions/findings.md#f1) | builds on | Sub-additive water uptake (indomethacin–PVP with 25% PVP, 3.6 measured against 14.0 g/100 g predicted) is explained by drug donors binding the single acceptor of the PVP repeat unit so that those sites are lost to water [rumondor2010a R12.h, C11, C12], the bond first shown by IR for indomethacin–PVP [taylor1997 R10, C35]. |
| [Physical stability F7](physical-stability/findings.md#f7) | [Miscibility and drug-polymer interactions F3](miscibility-and-drug-polymer-interactions/findings.md#f3) | builds on | The blends that separated sat near additive uptake and those that stayed mixed far below it [rumondor2009c R22, C24, C25], with the phase state read from IR spectra against calculated physical mixtures as in the IR/PCA method [rumondor2009a R6, R8]. |
| [Physical stability F8](physical-stability/findings.md#f8) | [Miscibility and drug-polymer interactions F7](miscibility-and-drug-polymer-interactions/findings.md#f7) | same mechanism | Both give χ that depends on the route: the ternary water-sorption fit gave χ₂₃ from −9.6 to 26.9 for felodipine–PVP [rumondor2010a R14], and one melting endotherm gave −6.16, −1.94 and −0.82 [marsac2009 R6]. |
| [Physical stability F8](physical-stability/findings.md#f8) | [Miscibility and drug-polymer interactions F8](miscibility-and-drug-polymer-interactions/findings.md#f8) | extends | A third route, water sorption of the blend, was added to melting point depression and solubility in a liquid analogue, and could not recover the drug–polymer χ [rumondor2010a R13, C30; marsac2009 R7, R9]. |
| [Physical stability F8](physical-stability/findings.md#f8) | [PEG dispersions F3](peg-dispersions/findings.md#f3) | same mechanism | The drug–polymer parameter was not recoverable from the water-sorption fit [rumondor2010a R13, R14], as the PEG readings agreed in sign but not in size [baird2011 R9.a; zhu2011 R5.c]. |
| [Physical stability F9](physical-stability/findings.md#f9) | [Crystallization tendency F17](crystallization-tendency/findings.md#f17) | builds on | The storage ranking of the neat classes was run at 0% RH only [baird2010 X16]; sorbed water raised the nucleation of amorphous felodipine about 20-fold at 75% RH [konno2008 R1, C11], so the dry order says little about the humid risk (synthesis). |
| [Physical stability F9](physical-stability/findings.md#f9) | [Crystallization tendency F21](crystallization-tendency/findings.md#f21) | builds on | Growth of neat felodipine films jumped 20–40-fold above about 80% RH in uncovered thin films, which the authors take to report surface growth [rumondor2010 R1, C28], the surface effect of the surface–bulk finding [kestur2013a R8, R9]. |
| [Physical stability F9](physical-stability/findings.md#f9) | [PEG dispersions F11](peg-dispersions/findings.md#f11) | conditions differ | Water sped nucleation and growth of the amorphous drug (nucleation 1.4 × 10⁵ to 2.9 × 10⁶ m⁻³ s⁻¹ from 0 to 75% RH) [konno2008 R1, R3], whereas in PEO compacts water made crystalline drugs lose crystallinity (acetaminophen in about 14–17 days) [marsac2008b R2, R10]. |
| [Physical stability F10](physical-stability/findings.md#f10) | [Crystallization tendency F17](crystallization-tendency/findings.md#f17) | extends | A polymer kept the nucleation of felodipine below that of the neat drug at every humidity, 2.3 × 10³ against 2.9 × 10⁶ m⁻³ s⁻¹ at 75% RH with 25% PVP [konno2008 R3, C13], so the neat-drug class, a first flag only [baird2010 C46], describes the case without a polymer (synthesis). |
| [Physical stability F10](physical-stability/findings.md#f10) | [Miscibility and drug-polymer interactions F9](miscibility-and-drug-polymer-interactions/findings.md#f9) | extends | The kinetic stabilization inferred for dry dispersions [marsac2006b C42] persisted with sorbed water but needed enough polymer: 3% PVP left nucleation only about threefold below the drug alone at 75% RH [konno2008 R3.b, C12]. |
| [Physical stability F11](physical-stability/findings.md#f11) | [Miscibility and drug-polymer interactions F17](miscibility-and-drug-polymer-interactions/findings.md#f17) | conditions differ | The dry ranking by hydrogen-bond strength, PVP best for felodipine [kestur2010a R6, R9], reversed in humid storage, where PVP was the weaker inhibitor from about 64–75% RH [rumondor2009b R9, C13]. |
| [Physical stability F11](physical-stability/findings.md#f11) | [Miscibility and drug-polymer interactions F23](miscibility-and-drug-polymer-interactions/findings.md#f23) | conditions differ | PVP dispersions of felodipine and celecoxib crystallized sooner than HPMCAS ones in humid storage [rumondor2009b R9; xie2016b R23], whereas naringenin and quercetin with PVP showed no crystals for more than 200 to more than 450 days where HPMCAS crystallized in 1–3 days [wegiel2013a R9]. |
| [Physical stability F11](physical-stability/findings.md#f11) | [Miscibility and drug-polymer interactions F26](miscibility-and-drug-polymer-interactions/findings.md#f26) | same mechanism | The more hygroscopic neutral polymer lost ground in humid storage: PVPVA dispersions of lumefantrine crystallized first at 40 °C/75% RH [hiew2022a R6, C10] as PVP did for celecoxib at 94% RH [xie2016b R23] (synthesis). |
| [Physical stability F12](physical-stability/findings.md#f12) | [Miscibility and drug-polymer interactions F3](miscibility-and-drug-polymer-interactions/findings.md#f3) | builds on | The comparison of an IR spectrum with a calculated physical-mixture spectrum, which showed mixing in the dry dispersions [rumondor2009a R6, R8], was used to show unmixing after humid storage [rumondor2009c R7, C17]. |
| [Physical stability F12](physical-stability/findings.md#f12) | [Miscibility and drug-polymer interactions F18](miscibility-and-drug-polymer-interactions/findings.md#f18) | same mechanism | Drug-rich domains appeared after water in dispersions that were mixed when dry (pimozide–PVP, two Tgs wet and dried) [rumondor2009c R17] and in solvent-cast itraconazole–HPMC films whose powder gave one Tg [purohit2015a R1, R7]; in both the history of the sample, not its dry composition alone, set the state (synthesis). |
| [Physical stability F13](physical-stability/findings.md#f13) | [Crystallization tendency F13](crystallization-tendency/findings.md#f13) | same mechanism | No single descriptor decided: log P, Tm, heat of fusion and the Hoffman driving force showed no trend with humid separation [rumondor2009c C27, C32], and molecular descriptors separated the crystallization classes only partly [baird2010 R21, C63]. |
| [Physical stability F13](physical-stability/findings.md#f13) | [Miscibility and drug-polymer interactions F1](miscibility-and-drug-polymer-interactions/findings.md#f1) | builds on | The PVP carbonyl red shift as a measure of drug–PVP bond strength [rumondor2009c R29, C48] follows the band assignments for indomethacin–PVP [taylor1997 R11, R12]. |
| [Physical stability F13](physical-stability/findings.md#f13) | [Miscibility and drug-polymer interactions F2](miscibility-and-drug-polymer-interactions/findings.md#f2) | builds on | The shift scale (45 cm⁻¹ indomethacin, 25 felodipine, 0 ketoconazole) [marsac2009 R12] is extended to 20 systems, acids 41–44 and OH and NH donors 15–24 cm⁻¹ [rumondor2011 R15]. |
| [Physical stability F13](physical-stability/findings.md#f13) | [Miscibility and drug-polymer interactions F7](miscibility-and-drug-polymer-interactions/findings.md#f7) | same mechanism | The χ of nifedipine–PVP is cited as 0.0 in the humidity paper [rumondor2009c R26] against −3.8 from the melting onset [marsac2006b R6, R9], the reading dependence of melting point depression [marsac2009 R6]. |
| [Physical stability F13](physical-stability/findings.md#f13) | [Miscibility and drug-polymer interactions F17](miscibility-and-drug-polymer-interactions/findings.md#f17) | same mechanism | The donor group of the drug and the strength of its bond to PVP decided which dispersions separated [rumondor2009c C46, C48], as complementary groups decided which polymer protected [kestur2010a R6, R9]. |
| [Physical stability F13](physical-stability/findings.md#f13) | [Miscibility and drug-polymer interactions F23](miscibility-and-drug-polymer-interactions/findings.md#f23) | conditions differ | Carboxylic acids with strong PVP shifts (42–44 cm⁻¹) resisted humid separation [rumondor2009c R29, C48], whereas PVP dispersions of the polyphenols naringenin and quercetin stayed amorphous for more than 200 to more than 450 days at humid conditions [wegiel2013a R9] (synthesis: separation was not measured there). |
| [Physical stability F13](physical-stability/findings.md#f13) | [Miscibility and drug-polymer interactions F25](miscibility-and-drug-polymer-interactions/findings.md#f25) | conditions differ | The PVP carbonyl shift of naringenin and quercetin (21 and 24 cm⁻¹) [wegiel2013a R4] lies in the 15–24 cm⁻¹ range of the OH and NH donors that separated in PVP [rumondor2011 R15], yet those dispersions stayed amorphous for more than 200–450 days; separation was not measured in the polyphenol study (synthesis). |
| [Physical stability F13](physical-stability/findings.md#f13) | [PEG dispersions F11](peg-dispersions/findings.md#f11) | same mechanism | The drug–water interaction separated the outcomes in both carriers: drug–water χ 2.84–3.58 for drugs that separated from PVP against 2.50–2.71 [rumondor2009c R24], and the activity coefficient in water from about 10¹ (complete loss) to 10⁵ (no loss) in PEO [marsac2008b R14, R16] (synthesis). |
| [Physical stability F14](physical-stability/findings.md#f14) | [Crystallization tendency F14](crystallization-tendency/findings.md#f14) | extends | The route sets the result as the drying rate did for spin-coated drugs [vaneerdenbrugh2010a R12, R13]: water in the casting solvent or humid casting air prevented drug–polymer mixing from the start [marsac2010 R13, C16; saboo2017 R6, R8]. |
| [Physical stability F15](physical-stability/findings.md#f15) | [Miscibility and drug-polymer interactions F9](miscibility-and-drug-polymer-interactions/findings.md#f9) | builds on | After separation the drug-rich phase holds little polymer, so crystallization is faster [rumondor2009c C5, C6], the opposite of the kinetic stabilization of a mixed dispersion [marsac2009 C34]. |
| [Physical stability F15](physical-stability/findings.md#f15) | [Miscibility and drug-polymer interactions F19](miscibility-and-drug-polymer-interactions/findings.md#f19) | same mechanism | Separated dispersions crystallized within 22 h to 6 days at 94% RH [rumondor2009c R7, R9, R12], and telaprevir films carried drug-rich domains from 10% drug with HPMC and PVPVA [li2016c R4, R6]; the crystallization of such domains was not measured for telaprevir (synthesis). |
| [Physical stability F16](physical-stability/findings.md#f16) | [Crystallization tendency F17](crystallization-tendency/findings.md#f17) | extends | The storage ranking of the neat classes was run at 0% RH only [baird2010 X16], whereas only 75% RH and above separated PVP from HPMCAS in felodipine dispersions [rumondor2009b R19, R9]. |
| [Physical stability F16](physical-stability/findings.md#f16) | [PEG dispersions F12](peg-dispersions/findings.md#f12) | same mechanism | Accelerated humid conditions can change the system: the route of crystallization may be altered at high RH [rumondor2009c C52], and partially crystalline PEO may defeat Arrhenius extrapolation [marsac2008b C40]. |
| [Physical stability F17](physical-stability/findings.md#f17) | [Crystallization tendency F3](crystallization-tendency/findings.md#f3) | builds on | Nucleation and growth respond separately to a polymer, as they occupy separate windows in neat drugs [trasi2012a C22, C35; trasi2014b R5]: PAA gave the largest growth inhibition and the most nuclei in acetaminophen [trasi2012a R6, R12]. |
| [Physical stability F17](physical-stability/findings.md#f17) | [Crystallization tendency F16](crystallization-tendency/findings.md#f16) | builds on | Nuclei counted after one cooling–reheating cycle [trasi2012a P6, R12] use the logic of nuclei that form on cooling and grow on reheating [baird2013 C36, C37]. |
| [Physical stability F17](physical-stability/findings.md#f17) | [Miscibility and drug-polymer interactions F10](miscibility-and-drug-polymer-interactions/findings.md#f10) | conditions differ | Growth was ordered PVP > PVP/VA > HPMCAS > PVAc for felodipine [kestur2010a R6], while for acetaminophen the nuclei counts ran in a different order from growth, HPMCAS 2.5 and PVP 5.2 against PAA 26 [trasi2012a R12]. |
| [Physical stability F18](physical-stability/findings.md#f18) | [Crystallization tendency F19](crystallization-tendency/findings.md#f19) | builds on | PVP slowed flutamide about tenfold at 10 mol% [trasi2012b R3, C7], the value that the five-drug comparison repeats (9–14-fold) [trasi2014a R6, X6]. |
| [Physical stability F18](physical-stability/findings.md#f18) | [Miscibility and drug-polymer interactions F12](miscibility-and-drug-polymer-interactions/findings.md#f12) | extends | The hydrogen-bond order was extended from felodipine [kestur2010a R6, R9] to flutamide, acetaminophen and celecoxib, where the order of IR shifts matched the order of growth inhibition [trasi2012b R14, C7; xie2016b R3, R4, R5, R21]. |
| [Physical stability F18](physical-stability/findings.md#f18) | [Miscibility and drug-polymer interactions F13](miscibility-and-drug-polymer-interactions/findings.md#f13) | same mechanism | Donor polymers worked for an acceptor, PAA for the acetaminophen carbonyl [trasi2012a R6, C31] as for the bifonazole acceptor [kestur2011 R7]. |
| [Physical stability F18](physical-stability/findings.md#f18) | [Miscibility and drug-polymer interactions F14](miscibility-and-drug-polymer-interactions/findings.md#f14) | same mechanism | An intramolecular bond made the nominal partner less available: flutamide's carbonyl in a CH···O contact [trasi2012b C35, C37], nimesulide's NH [kestur2011 C27]. |
| [Physical stability F18](physical-stability/findings.md#f18) | [Miscibility and drug-polymer interactions F17](miscibility-and-drug-polymer-interactions/findings.md#f17) | extends | Acceptor polymers slowed the donor flutamide and the donor polymer PAA slowed acetaminophen, whose amide carbonyl is free [trasi2012b R3, C7; trasi2012a R6, C31], adding two drugs to the melt-growth designs of the complementarity pattern [kestur2010a R6; kestur2011 R7]. |
| [Physical stability F19](physical-stability/findings.md#f19) | [Crystallization tendency F7](crystallization-tendency/findings.md#f7) | builds on | Bulk form I of felodipine grew faster than form II at every temperature [kestur2013a R8, C4], the polymorph kinetics of the supercooled liquid treated in the polymorph finding [lu2016 C42, C52]. |
| [Physical stability F19](physical-stability/findings.md#f19) | [Crystallization tendency F19](crystallization-tendency/findings.md#f19) | same mechanism | 3% PVP slowed the two felodipine polymorphs by a similar factor (about 16–19 at 60 °C) [kestur2013a R15], as it slowed fast and slow growers of five drugs [trasi2014a R6, C10], and both conclude that faster growers need more polymer [kestur2013a C26; trasi2014a C11]. |
| [Physical stability F19](physical-stability/findings.md#f19) | [Crystallization tendency F21](crystallization-tendency/findings.md#f21) | builds on | Form I grew about 6 times faster at the free surface than in the bulk at 18 °C [kestur2013a R8, R9]. |
| [Physical stability F19](physical-stability/findings.md#f19) | [Miscibility and drug-polymer interactions F10](miscibility-and-drug-polymer-interactions/findings.md#f10) | builds on | The polymorph study used the same felodipine form II melt system: 3% PVP K-29/32 lowered form II growth 18.7-fold at 60 °C [kestur2013a R15], and PVP K-12 10.0-fold at 70 °C [kestur2010a R6]. |
| [Physical stability F20](physical-stability/findings.md#f20) | [Crystallization tendency F7](crystallization-tendency/findings.md#f7) | extends | A polymer or surfactant can change which polymorph crystallizes (curcumin form II from HPMC, HPMCAS and CMCAB) [wegiel2014 R15, R17], extending polymorph selection from temperature alone [lu2016 C42, C52]. |
| [Physical stability F21](physical-stability/findings.md#f21) | [Crystallization tendency F21](crystallization-tendency/findings.md#f21) | extends | 3% PVP slowed bulk growth about 7-fold and surface growth about 2-fold, and smaller particles crystallized sooner [kestur2013b R12, C32, R3.e], extending the surface effect from films to powders [kestur2013a R9]. |
| [Physical stability F22](physical-stability/findings.md#f22) | [Crystallization tendency F19](crystallization-tendency/findings.md#f19) | extends | Surfactants that mix and lower Tg accelerated growth of celecoxib, TPGS about 85-fold at 70 °C [mosqueragiraldo2014 R1, C9], the accelerators noted beside the polymer factor [mosqueragiraldo2014 C9; sato2016 R11]. |
| [Physical stability F23](physical-stability/findings.md#f23) | [Miscibility and drug-polymer interactions F17](miscibility-and-drug-polymer-interactions/findings.md#f17) | conditions differ | Humid separation followed the water the polymer brings in, PVP > PVPVA > HPMCAS [rumondor2010c R18, C20, C34], while the complementarity rule ranks polymers by the groups they offer [kestur2010a R6, R9]. |
| [Physical stability F24](physical-stability/findings.md#f24) | [Crystallization tendency F14](crystallization-tendency/findings.md#f14) | builds on | Naringenin and quercetin could not be made amorphous by rotary evaporation without a polymer [wegiel2013a R7], and resveratrol was classed a rapid crystallizer [wegiel2013b C6]. |
| [Physical stability F24](physical-stability/findings.md#f24) | [Miscibility and drug-polymer interactions F14](miscibility-and-drug-polymer-interactions/findings.md#f14) | same mechanism | An intramolecular hydrogen bond weakened the protection: curcumin keeps its keto–enol bond [wegiel2014 C11, C33], as nimesulide keeps its NH···O bond [kestur2011 C27]. |
| [Physical stability F24](physical-stability/findings.md#f24) | [Miscibility and drug-polymer interactions F17](miscibility-and-drug-polymer-interactions/findings.md#f17) | extends | Polymers that accept strong hydrogen bonds or pair with the drug as a base gave the most stable resveratrol and curcumin dispersions [wegiel2013b R12, C29; wegiel2014 R13]. |
| [Physical stability F24](physical-stability/findings.md#f24) | [Miscibility and drug-polymer interactions F21](miscibility-and-drug-polymer-interactions/findings.md#f21) | same mechanism | The E100 amine bands disappeared as the polyphenol was added, read as an acid–base reaction [wegiel2013b R7, C15; wegiel2013a R6, C17]. |
| [Physical stability F24](physical-stability/findings.md#f24) | [Miscibility and drug-polymer interactions F23](miscibility-and-drug-polymer-interactions/findings.md#f23) | builds on | Resveratrol dispersions with PVP and E100 stayed amorphous for more than 400 days at 84% RH [wegiel2013b R12], as naringenin and quercetin did for more than 200 to more than 450 days [wegiel2013a R9]. |
| [Physical stability F24](physical-stability/findings.md#f24) | [Miscibility and drug-polymer interactions F24](miscibility-and-drug-polymer-interactions/findings.md#f24) | builds on | The donor and acceptor classification was applied to the polyphenols (acceptor strength pKBHX 2.38 for PVP, 1.98 for the E100 amine) [wegiel2013b R17], giving the E100 > PVP > HPMC order [wegiel2013b C21]; for the screen drugs, acceptor-only polymers worked for donors [vaneerdenbrugh2011a R3, C20]. |
| [Physical stability F24](physical-stability/findings.md#f24) | [Miscibility and drug-polymer interactions F25](miscibility-and-drug-polymer-interactions/findings.md#f25) | same mechanism | The IR ranking did not follow the storage ranking in either paper [wegiel2013b X10; wegiel2014 X7], and in the naringenin and quercetin paper HPMCAS, third by IR, gave the least stable dry dispersion [wegiel2013a X4, R9]. |
| [Physical stability F25](physical-stability/findings.md#f25) | [Crystallization tendency F17](crystallization-tendency/findings.md#f17) | extends | The storage test of the neat classes used one condition [baird2010 X16]; curcumin dispersions ranked polymers only at 50 °C/0% RH, 25 °C/84% RH and 40 °C/75% RH [wegiel2014 R13, C23]. |
| [Physical stability F25](physical-stability/findings.md#f25) | [Miscibility and drug-polymer interactions F23](miscibility-and-drug-polymer-interactions/findings.md#f23) | builds on | The five-condition storage design of the polyphenol paper [wegiel2013a P1, P2, P3] was extended to curcumin and resveratrol with dry heat, where humidity and heat gave different orders [wegiel2013b R12, R13, C39]. |
| [Physical stability F25](physical-stability/findings.md#f25) | [PEG dispersions F12](peg-dispersions/findings.md#f12) | same mechanism | Both authors warn that accelerated conditions are hard to choose: the different rate and extent at 50 °C highlights the difficulty of predicting lower temperatures [kestur2013b C37, C40], and partially crystalline polymers may defeat Arrhenius extrapolation [marsac2008b C40]. |
| [Physical stability F27](physical-stability/findings.md#f27) | [Miscibility and drug-polymer interactions F12](miscibility-and-drug-polymer-interactions/findings.md#f12) | builds on | The enantiomer pair is the cleanest test of the hydrogen-bond reading: with HPMCAS the enantiomer that lost more of its free NH2 band was slowed about twice as much [sato2015 R13, R14, R5], and the felodipine IR order matched growth inhibition [kestur2010a R6, R9]. |
| [Physical stability F27](physical-stability/findings.md#f27) | [Miscibility and drug-polymer interactions F17](miscibility-and-drug-polymer-interactions/findings.md#f17) | extends | Achiral PVP slowed both enantiomers alike and the chiral grades discriminated by about twofold [sato2015 R5, C10], showing that the polymer–compound contact, not bulk properties, sets inhibition [sato2015 C27]. |
| [Physical stability F28](physical-stability/findings.md#f28) | [Crystallization tendency F2](crystallization-tendency/findings.md#f2) | builds on | The temperatures of fastest growth of the PHB study follow the 0.94 Tm rule [sato2016 E2], as for vemurafenib and indomethacin [lu2016 R12; tian2017 E3]. |
| [Physical stability F28](physical-stability/findings.md#f28) | [Crystallization tendency F19](crystallization-tendency/findings.md#f19) | extends | PHB sped three high-Tg compounds up to about tenfold (griseofulvin about 0.015 to 0.15 µm/s at 120 °C) [sato2016 R4, R6], extending the accelerators of the polymer factor [sato2016 R11]. |
| [Physical stability F29](physical-stability/findings.md#f29) | [Crystallization tendency F2](crystallization-tendency/findings.md#f2) | builds on | Indomethacin δ, α and γ peaked at 110, 135 and 135 °C [tian2017 R7, E3], and HPMC moved the α maximum from 135 to 100 °C [tian2017 R15, C34]. |
| [Physical stability F29](physical-stability/findings.md#f29) | [Crystallization tendency F7](crystallization-tendency/findings.md#f7) | builds on | Polymorph-specific polymer effects extend the polymorph kinetics of the supercooled liquid: at 120 °C α was slowed most and γ least by each of three polymers [tian2017 R12.i, C37; lu2016 C42, C52]. |
| [Physical stability F29](physical-stability/findings.md#f29) | [Crystallization tendency F19](crystallization-tendency/findings.md#f19) | conditions differ | Indomethacin polymorphs were slowed by factors from 2.9 (γ) to 2900 (α) with HPMC at 120 °C [tian2017 R12.i], unlike the similar factors for five drugs with PVP [trasi2014a R6, X6]. |
| [Physical stability F29](physical-stability/findings.md#f29) | [Miscibility and drug-polymer interactions F10](miscibility-and-drug-polymer-interactions/findings.md#f10) | extends | HPMC, HPMCAS and PVP lowered indomethacin growth 5–25-, 2–12- and 2–3-fold at 80–100 °C [tian2017 R12, C28], adding polymorphs and temperatures above the growth maximum to the felodipine polymer order [kestur2010a R6]. |
| [Physical stability F30](physical-stability/findings.md#f30) | [Miscibility and drug-polymer interactions F18](miscibility-and-drug-polymer-interactions/findings.md#f18) | builds on | The film methods that found drug-rich domains where DSC gave one Tg [purohit2015a R1, R7] were turned into a storage prediction for indomethacin, with one Tg throughout [tian2016 R13, R2.f]. |
| [Physical stability F30](physical-stability/findings.md#f30) | [Miscibility and drug-polymer interactions F19](miscibility-and-drug-polymer-interactions/findings.md#f19) | same mechanism | Onset loadings of drug-rich domains differed by polymer, 10% for HPMC and PVPVA and about 40% for HPMCAS in telaprevir [li2016c R4, R5, R6] and 50–60%, 20–30% and 70–80% in indomethacin [tian2016 R2.f, R3.c, R4.h]. |
| [Physical stability F30](physical-stability/findings.md#f30) | [Miscibility and drug-polymer interactions F23](miscibility-and-drug-polymer-interactions/findings.md#f23) | extends | Dry storage of indomethacin dispersions followed the miscibility limit [tian2016 C53, R17], adding a phase-state reading to the contact reading of storage stability [wegiel2013a C35]. |
| [Physical stability F33](physical-stability/findings.md#f33) | [Crystallization tendency F17](crystallization-tendency/findings.md#f17) | conditions differ | The neat amorphous tacrolimus lasted 2 weeks at 40 °C/75% RH, while one generic crystallized faster than the pure drug and the other five products did not [trasi2017 R16, C28, R6, R7]; the class is only a first flag. |
| [Physical stability F34](physical-stability/findings.md#f34) | [Crystallization tendency F9](crystallization-tendency/findings.md#f9) | builds on | Lumefantrine is class III by the heat–cool–heat cycle [trasi2020 R5, R6, C7], and its dispersions crystallized in 1 week with PVPVA but not in 3 months with enteric polymers [trasi2020 R3, C6]. |
| [Physical stability F34](physical-stability/findings.md#f34) | [Crystallization tendency F14](crystallization-tendency/findings.md#f14) | extends | In powders precipitated into water, residual water and drying time decided (CAP with 9% water crystallized from day 1, Eudragit L100 dried for 4 days was crystalline as made) [bhujbal2021 R1, R4, R8, R10, R11], as slower solvent removal crystallized slow crystallizers [vaneerdenbrugh2010a R12, R13]. |
| [Physical stability F34](physical-stability/findings.md#f34) | [Miscibility and drug-polymer interactions F21](miscibility-and-drug-polymer-interactions/findings.md#f21) | builds on | The XPS protonation of lumefantrine by the acidic polymers in 40:60 rotary-evaporated dispersions (16–61%) [trasi2020 R11, C14] and in 25% compacts (19–55%) [hiew2022a R20] is the marker on which the stability correlation rests. |
| [Physical stability F34](physical-stability/findings.md#f34) | [Miscibility and drug-polymer interactions F23](miscibility-and-drug-polymer-interactions/findings.md#f23) | builds on | Enteric polymers kept 40% lumefantrine amorphous at 40 °C/75% RH for 3 months where PVPVA crystallized in a week [trasi2020 R3, C6], the same direction as the loading series [hiew2022a R6]. |
| [Physical stability F34](physical-stability/findings.md#f34) | [Miscibility and drug-polymer interactions F26](miscibility-and-drug-polymer-interactions/findings.md#f26) | builds on | The stability result of the rotary-evaporated dispersions [trasi2020 R3] is the storage side of the stability–release trade-off [hiew2022a R6, C65]. |

<!-- end -->