JDDST D 23 02773 - R1

Journal of Drug Delivery Science and Technology
A Multifaceted Approach for Grading of Polymers for the Development of Stable

Amorphous Solid Dispersion of Riluzole
--Manuscript Draft--
Manuscript Number: JDDST-D-23-02773R1
Article Type: Research Paper
Keywords: Amorphous solid dispersion; Drug-Polymer Interactions; Density functional theory;

Computational study; Miscibility; Flory-Huggins theory
Corresponding Author: Brahmeshwar Mishra, Ph.D.

Indian Institute of Technology BHU Varanasi
Varanasi, INDIA
First Author: Kanchan Bharti
Order of Authors: Kanchan Bharti
Gurudutt Dubey
Manish Kumar
Abhishek Jha
Manjit Manjit
Mansi Upadhyay
Pramod S Mali
Ashutosh Kumar
Prasad V. Bharatam
Brahmeshwar Mishra, Ph.D.
Abstract: The current work involves grading three widely used polymers for preparing a
kinetically and thermodynamically stable amorphous solid dispersion (ASD) of a
neuroprotective drug Riluzole (RLZ) by evaluating the drug-polymer interactions.
Polymers were screened based on their chemical interaction with RLZ and their ability
to inhibit the crystallization of the drug. Detailed computational studies were performed
to quantify the non-covalent interactions between RLZ and the modeled structures of
polymers; polyacrylic acid (PAA), polyvinylpyrrolidone vinyl acetate (PVP VA), and
hydroxypropyl methyl cellulose acetate succinate (HPMC AS) for calculating the
interaction energies of drug-polymer complexes. Experimental characterization of drug-
polymer complexes through analytical techniques further validated the formation of the
drug-polymer complexes. The results indicated that RLZ interacts with the polymers in
the order PAA > PVP VA > HPMC AS, giving an insight into the stability of drug-
polymer complexes. In vitro dissolution study also showed higher dissolution profile of
RLZ with PAA. Further, the drug-polymer miscibility was determined by employing
Flory-Huggins theory and stability in accordance with Gibbs free energy of mixing and
phase diagram. This work presented a multi-technique approach for the grading of
polymers in search of a stable ASD of the poorly water-soluble drug RLZ.
Suggested Reviewers: Krishna Kumar Patel, Ph.D.

The University of Iowa College of Pharmacy
krishnakumar-patel@uiowa.ed
Expert of pharmaceutics
Rahul Tripathi, Ph.D.

Senior scientist
rahul.tripathi@medpharm.co.uk
Expert of Pharmaceutics
Kanchan Kohli
Jamia Hamdard Faculty of Pharmacy
kanchankohli@gmail.com
Powered by Editorial Manager® and ProduXion Manager® from Aries Systems Corporation
Expert of Pharmaceutics
Nalini Shastri
National Institute of Pharmaceutical Education and Research Hyderabad
nalini.niperhyd@gov.in
Expert in solid-state pharmaceutics
Opposed Reviewers:
Response to Reviewers: The authors are thankful to the editor and reviewers for evaluating the manuscript
critically and giving their insightful comments to enhance the quality of the work.
Reviewer #1: This manuscript described the use of multiple approaches to determine
the strength of drug-polymer interaction between RLZ and 3 polymers, PAA, PVPVA
and HPMCAS. The manuscript is well written. I suggest to accept it for publication with
minor changes.
Comment 1. Suggest showing the molecular structure of the model drug RLZ in Fig 1
together with the 3 polymers, and describe the procedure of how the 3D structure of
the drug molecule was prepared for MD simulations.
Response 1. As per the suggestion, the molecular structure of RLZ has been added in
Fig 1. The procedure for obtaining the 3D structure of RLZ has been described in the
section 2.1.1 in the revised manuscript. The structure was obtained from the
experimentally reported crystal structure of RLZ in the Cambridge Crystallographic
Data Centre.
Comment 2. Describe the physicochemical properties of RLZ such as solubility in the
dissolution medium.
Response 2. The reported solubility of drug at neutral and acidic pH and the
dissociation constant has been mentioned and cited in the in vitro dissolution study
section (section 3.8) of the manuscript. It has been mentioned that riluzole is a weakly
acidic molecule with a solubility of 0.3 mg/ml at neutral pH and approximately 12 mg/ml
at acidic pH. It has a pKa of 3.47, and it shows pH dependent solubility. The purpose of
selecting pH 6.8 as dissolution medium has also been mentioned in the revised
manuscript.
Comment 3. Dissolution: Describe how the dissolution sink condition was selected.
Explain if the less differentiation in dissolution rate in Fig S7 is related to higher sink
conditions.
Response 3. The authors appreciate the suggestion of the reviewer and have included
the explanation in section 3.8 in the revised manuscript. The method for dissolution
was adopted from Yadav et al. 2018 and is cited as Reference no. 31.
Yes, it is true that the small difference in dissolution between different formulations is
attributed to higher sink conditions. The ASD formulation is a supersaturated drug
delivery system. These systems show higher dissolution of drug of poorly soluble
drugs due to supersaturation, which can lead to drug precipitation. For this reason,
higher sink condition was maintained. The dissolution profile figure is now figure S8 in
the revised supporting information.
Comment 4. In MD simulations, the molar ratio of polymer to drug is probably close to
1:1. Please describe the molar ratio of polymer to drug in the ASDs (i.e. 50% drug load
at w/w%). Please explain why the interaction determined by MD simulations at 1:1
molar ratio is correlated to the 50% drug loading ASDs.
Response 4. The molar ratio of polymer to drug in the 3 ASDs vary for 50% drug load
at w/w% because of the variable molecular weights of the three polymers. For PAA-
RLZ ASD, the polymer to drug molar ratio is 1923:1, for PVP VA-RLZ ASD, it is 218:1
and for HPMC AS-RLZ ASD, it is 77:1, in experimental condition, which is mentioned in
section 3.1.1 section of the revised manuscript. In silico studies (especially DFT
calculation) are best performed at the small model system of large polymers to save
the calculation time and computational cost which allow more detailed and precise
structural investigations. The DFT calculations of the large polymers collapse because
of high degree of freedom and many local minima energy points. In the current work,
we have taken the small model system comprising of around 20 monomeric units of
the polymer, which can be handled during in silico calculations and the length of the
model system also provides sufficient flexibility to the polymers. We have taken 1:1
molar ratio to precisely estimate the sites and type of molecular interactions between
RLZ and the polymers in the small model system. When the system is extended to the
actual length, the similar interaction are expected with the other RLZ molecules with
the long polymer chain in ASDs. The choice of small model system is justified and
molar ratio are mentioned in the revised manuscript in the last paragraph of section
3.1.1.
Comment 5. Physical stability of ASDs is usually shown by testing the ASDs under
stressed conditions (for example, 40°C/75% RH). You did not show the real stability
data. I suggest you add stability testing in your future studies.
Response 5. The authors are greatly thankful to the reviewer for their insightful
suggestion on stability studies. We are already including real time stability studies
under stressed condition of 40°C/75% RH as an extension of this work in another
manuscript. This suggestion could have impact on the quality of our future work. A
statement mentioning the importance of real time stability study as a future scope of
this work has also been added in the conclusion section of the revised manuscript.
Comment 6. At page 14 where the MD simulation data were discussed, I have a
trouble to understand why interaction with MeOH need to be calculated and reported.
During the ASD preparation, MeOH was removed. MeOH does not play any role in the
stability of the ASDs. Please explain your rationale of including MeOH in the
calculations.
Response 6. While performing the calculations for interaction energies between RLZ
and polymers, the gas phase (absence of solvent) and solvent phase conditions were
considered. We agree with the reviewer’s point that MeOH is not playing any role in the
interaction during ASD preparation. In these in silico studies, we have attempted to
quantify the interactions between drug and the polymer using model systems rather
than simulating the whole amorphous dispersion. To validate the order of interaction
energies of RLZ with the three polymers, the calculation are performed in both gas and
solvent phases. Usually presence of polar solvent may affect the binding energies
because of its high dielectric constant which is added to the wavefunction of energy
calculation. Since, methanol was used as the polar protic solvent to dissolve the drug
and polymers, it was the choice of solvent to validate the order of interaction energies
obtained in the gas phase. During our studies, we have seen that the order of
interaction energies is maintained in the solvent phase as well. The rationale for
choosing methanol for solvent phase DFT study has been mentioned in section 2.1.4
in the revised manuscript.
Reviewer #2: The manuscript describes the process of using a few different techniques
available to inform the formulation decision making process. Both in silico and
experimental based modelling techniques are presented for a single model compound.
These techniques have been performed for other drugs in the past and even reviewed
in review articles recently (Walden DM, Bundey Y, Jagarapu A, Antontsev V,
Chakravarty K, Varshney J. Molecular Simulation and Statistical Learning Methods
toward Predicting Drug-Polymer Amorphous Solid Dispersion Miscibility, Stability, and
Formulation Design. Molecules. 2021; 26(1):182.
https://doi.org/10.3390/molecules26010182) for instance.
Comment 1. As such, the manuscript requires greater depth and commentary on each
technique to be of use to a reader. For example, a couple of in silico methods were
used that resulted in different values. The discussion should go into more detail about
the differences and rationale for the selection of the tests selected and which
correspond with the true results. The manuscript should describe how the software is
treating each test and then which would be more likely expected to be useful to the
formulating scientist. There is also a lack of discussion regarding the limitations of the
testing.
Response 1. As per the reviewer’s suggestion the discussion on rationale for selection
of the in silico techniques and in silico results are explained in more detail in section
3.1.1 and 3.1.2 in the revised manuscripts. The limitations and usefulness of these
studies to the formulating scientist are also described in the section 3.1.1, 3.1.2 in the
revised manuscript.
Comment 2. The polymer molecules are truncated drastically from their true size in the
in silico experiments. The possible implications of such should be mentioned.
Response 2. The justification for truncating the polymer and choosing smaller model
system is explained in the section 3.1.1. in the revised manuscript.
The choice of the 20 monomeric units in PAA and PVP VA, and 5 monomeric units in
HPMC AS is based on an assumption that the optimized structures of the three
polymers should possess equivalent surface area and flexibility to interact with RLZ
molecule. During molecular dynamics simulation and quantum chemical energy
minimization of drug-polymer complexes, the uniformity in the surface area and volume
of the complexes can minimize the possibility of inconsistencies. DFT calculation are
best performed at the small model system of large polymers to save the calculation
time and computational cost, which allow more detailed and precise structural
investigations. The DFT calculations of the large polymers collapse because of high
degree of freedom and many local minima energy points.
Comment 3. The technique shows the interaction and strength of the drug and the
polymer but doesn't present how this compares to the drugs intermolecular interactions
with itself that guide its propensity to crystallize. This should be further discussed.
Response 3. We agree with the reviewer to compare the drug-polymer interactions with
the drug-drug interaction in order to observe crystallization tendency. A discussion has
been added in the section 3.1.2 of the revised manuscript. The crystallization
propensity of the drug in the simulated ASD can be better demonstrated using large
systems in the MD simulation. The drug-drug interaction energies were calculated for
the dimer of the drug. The drug-drug interaction energy of -10.18 kcal/mol justifies the
requirement of heating (enthalpy) at 50-60 oC to break these drug-drug intermolecular
interactions within the crystal lattice of drug. However, comparison of drug-drug and
drug-polymer interaction energies will lead to the false results, when calculated using
small model systems. In the current work, our aim was to screen the best polymer for
stable ASD formation through quantification of interaction energies. DFT calculation
provide more accurate energy values; however, DFT studies are limited to smaller
model system in order to provide the correct estimation of energy values. The order of
interaction energy values coincide with the order of stable ASDs observed during
experiments.
Wang et al. also describes the similar approach in their work [Wang, B.; Wang, D.;
Zhao, S.; Huang, X.; Zhang, J.; Lv, Y.; Liu, X.; Lv, G.; Ma, X. Eur. J. Pharm. Sci. 2017,
96, 45–52, https://doi.org/10.1016/j.ejps.2016.08.046.]
Cover Letter
To,
Prof. Dr. Florence Siepmann,

The Editor-in-chief,
Journal of Drug Delivery Science and Technology.
Subject: Submission of Original Research Manuscript for Publication
Dear Prof. Siepmann,
We are here submitting our research manuscript entitled “A Multifaceted Approach for
Grading of Polymers for the Development of Stable Amorphous Solid Dispersion of
Riluzole” as original research for publication in your esteemed journal “Journal of Drug
Delivery Science and Technology”.
The manuscript has been read and approved by all the authors and there is no conflict of interest
between the authors. We have provided a unique and multifaceted approach involving
computational studies, analytical techniques and mathematical modelling for selecting most
appropriate polymer with potential to form a stable amorphous solid dispersion. The study is a
mechanistic insight of the drug-polymer interactions. To the best of our knowledge, such
amalgamation of techniques have not been reported yet. We believe that the manuscript would
be an interesting read for wide range of researchers. We also confirm that the manuscript has
not been previously submitted or communicated elsewhere in any form for publication.
Sincerely,
Prof. B. Mishra
Department of Pharmaceutical Engineering & Technology,
Indian Institute of Technology (Banaras Hindu University),
Varanasi -221005 (Uttar Pradesh), INDIA
Phone no: +91 9415811824,
Fax No.: +91 542 2368428,
Email: bmishrabhu@rediffmail.com
Response to Reviewers
Response sheet on submission JDDST-D-23-02773

The authors are thankful to the editor and reviewers for evaluating the manuscript critically
and giving their insightful comments to enhance the quality of the work.
Reviewer #1: This manuscript described the use of multiple approaches to determine the
strength of drug-polymer interaction between RLZ and 3 polymers, PAA, PVPVA and
HPMCAS. The manuscript is well written. I suggest to accept it for publication with minor
changes.
Comment 1. Suggest showing the molecular structure of the model drug RLZ in Fig 1 together
with the 3 polymers, and describe the procedure of how the 3D structure of the drug molecule
was prepared for MD simulations.
Response 1. As per the suggestion, the molecular structure of RLZ has been added in Fig 1.
The procedure for obtaining the 3D structure of RLZ has been described in the section 2.1.1 in
the revised manuscript. The structure was obtained from the experimentally reported crystal
structure of RLZ in the Cambridge Crystallographic Data Centre.
Comment 2. Describe the physicochemical properties of RLZ such as solubility in the

dissolution medium.
Response 2. The reported solubility of drug at neutral and acidic pH and the dissociation
constant has been mentioned and cited in the in vitro dissolution study section (section 3.8) of
the manuscript. It has been mentioned that riluzole is a weakly acidic molecule with a solubility
of 0.3 mg/ml at neutral pH and approximately 12 mg/ml at acidic pH. It has a pKa of 3.47, and
it shows pH dependent solubility. The purpose of selecting pH 6.8 as dissolution medium has
also been mentioned in the revised manuscript.
Comment 3. Dissolution: Describe how the dissolution sink condition was selected. Explain
if the less differentiation in dissolution rate in Fig S7 is related to higher sink conditions.
Response 3. The authors appreciate the suggestion of the reviewer and have included the
explanation in section 3.8 in the revised manuscript. The method for dissolution was adopted
from Yadav et al. 2018 and is cited as Reference no. 31.
Yes, it is true that the small difference in dissolution between different formulations is attributed
to higher sink conditions. The ASD formulation is a supersaturated drug delivery system. These
systems show higher dissolution of drug of poorly soluble drugs due to supersaturation, which
can lead to drug precipitation. For this reason, higher sink condition was maintained. The
dissolution profile figure is now figure S8 in the revised supporting information.
Comment 4. In MD simulations, the molar ratio of polymer to drug is probably close to 1:1.
Please describe the molar ratio of polymer to drug in the ASDs (i.e. 50% drug load at w/w%).
Please explain why the interaction determined by MD simulations at 1:1 molar ratio is
correlated to the 50% drug loading ASDs.
Response 4. The molar ratio of polymer to drug in the 3 ASDs vary for 50% drug load at w/w%
because of the variable molecular weights of the three polymers. For PAA-RLZ ASD, the
polymer to drug molar ratio is 1923:1, for PVP VA-RLZ ASD, it is 218:1 and for HPMC AS-
RLZ ASD, it is 77:1, in experimental condition, which is mentioned in section 3.1.1 section of
the revised manuscript. In silico studies (especially DFT calculation) are best performed at the
small model system of large polymers to save the calculation time and computational cost
which allow more detailed and precise structural investigations. The DFT calculations of the
large polymers collapse because of high degree of freedom and many local minima energy
points. In the current work, we have taken the small model system comprising of around 20
monomeric units of the polymer, which can be handled during in silico calculations and the
length of the model system also provides sufficient flexibility to the polymers. We have taken
1:1 molar ratio to precisely estimate the sites and type of molecular interactions between RLZ
and the polymers in the small model system. When the system is extended to the actual length,
the similar interaction are expected with the other RLZ molecules with the long polymer chain
in ASDs. The choice of small model system is justified and molar ratio are mentioned in the
revised manuscript in the last paragraph of section 3.1.1.
Comment 5. Physical stability of ASDs is usually shown by testing the ASDs under stressed
conditions (for example, 40°C/75% RH). You did not show the real stability data. I suggest you
add stability testing in your future studies.
Response 5. The authors are greatly thankful to the reviewer for their insightful suggestion on
stability studies. We are already including real time stability studies under stressed condition
of 40°C/75% RH as an extension of this work in another manuscript. This suggestion could
have impact on the quality of our future work. A statement mentioning the importance of real
time stability study as a future scope of this work has also been added in the conclusion section
of the revised manuscript.
Comment 6. At page 14 where the MD simulation data were discussed, I have a trouble to
understand why interaction with MeOH need to be calculated and reported. During the ASD
preparation, MeOH was removed. MeOH does not play any role in the stability of the ASDs.
Please explain your rationale of including MeOH in the calculations.
Response 6. While performing the calculations for interaction energies between RLZ and
polymers, the gas phase (absence of solvent) and solvent phase conditions were considered.
We agree with the reviewer’s point that MeOH is not playing any role in the interaction during
ASD preparation. In these in silico studies, we have attempted to quantify the interactions
between drug and the polymer using model systems rather than simulating the whole
amorphous dispersion. To validate the order of interaction energies of RLZ with the three
polymers, the calculation are performed in both gas and solvent phases. Usually presence of
polar solvent may affect the binding energies because of its high dielectric constant which is
added to the wavefunction of energy calculation. Since, methanol was used as the polar protic
solvent to dissolve the drug and polymers, it was the choice of solvent to validate the order of
interaction energies obtained in the gas phase. During our studies, we have seen that the order
of interaction energies is maintained in the solvent phase as well. The rationale for choosing
methanol for solvent phase DFT study has been mentioned in section 2.1.4 in the revised
manuscript.
Reviewer #2: The manuscript describes the process of using a few different techniques
available to inform the formulation decision making process. Both in silico and experimental
based modelling techniques are presented for a single model compound. These techniques have
been performed for other drugs in the past and even reviewed in review articles recently
(Walden DM, Bundey Y, Jagarapu A, Antontsev V, Chakravarty K, Varshney J. Molecular
Simulation and Statistical Learning Methods toward Predicting Drug-Polymer Amorphous
Solid Dispersion Miscibility, Stability, and Formulation Design. Molecules. 2021; 26(1):182.
https://doi.org/10.3390/molecules26010182) for instance.
Comment 1. As such, the manuscript requires greater depth and commentary on each
technique to be of use to a reader. For example, a couple of in silico methods were used that
resulted in different values. The discussion should go into more detail about the differences and
rationale for the selection of the tests selected and which correspond with the true results. The
manuscript should describe how the software is treating each test and then which would be
more likely expected to be useful to the formulating scientist. There is also a lack of discussion
regarding the limitations of the testing.
Response 1. As per the reviewer’s suggestion the discussion on rationale for selection of the
in silico techniques and in silico results are explained in more detail in section 3.1.1 and 3.1.2
in the revised manuscripts. The limitations and usefulness of these studies to the formulating
scientist are also described in the section 3.1.1, 3.1.2 in the revised manuscript.
Comment 2. The polymer molecules are truncated drastically from their true size in the in
silico experiments. The possible implications of such should be mentioned.
Response 2. The justification for truncating the polymer and choosing smaller model system
is explained in the section 3.1.1. in the revised manuscript.
The choice of the 20 monomeric units in PAA and PVP VA, and 5 monomeric units in HPMC
AS is based on an assumption that the optimized structures of the three polymers should possess
equivalent surface area and flexibility to interact with RLZ molecule. During molecular
dynamics simulation and quantum chemical energy minimization of drug-polymer complexes,
the uniformity in the surface area and volume of the complexes can minimize the possibility of
inconsistencies. DFT calculation are best performed at the small model system of large
polymers to save the calculation time and computational cost, which allow more detailed and
precise structural investigations. The DFT calculations of the large polymers collapse because
of high degree of freedom and many local minima energy points.
Comment 3. The technique shows the interaction and strength of the drug and the polymer but
doesn't present how this compares to the drugs intermolecular interactions with itself that guide
its propensity to crystallize. This should be further discussed.
Response 3. We agree with the reviewer to compare the drug-polymer interactions with the
drug-drug interaction in order to observe crystallization tendency. A discussion has been added
in the section 3.1.2 of the revised manuscript. The crystallization propensity of the drug in the
simulated ASD can be better demonstrated using large systems in the MD simulation. The drug-
drug interaction energies were calculated for the dimer of the drug. The drug-drug interaction
energy of -10.18 kcal/mol justifies the requirement of heating (enthalpy) at 50-60 oC to break
these drug-drug intermolecular interactions within the crystal lattice of drug. However,
comparison of drug-drug and drug-polymer interaction energies will lead to the false results,
when calculated using small model systems. In the current work, our aim was to screen the best
polymer for stable ASD formation through quantification of interaction energies. DFT
calculation provide more accurate energy values; however, DFT studies are limited to smaller
model system in order to provide the correct estimation of energy values. The order of
interaction energy values coincide with the order of stable ASDs observed during experiments.
Wang et al. also describes the similar approach in their work [Wang, B.; Wang, D.; Zhao, S.;
Huang, X.; Zhang, J.; Lv, Y.; Liu, X.; Lv, G.; Ma, X. Eur. J. Pharm. Sci. 2017, 96, 45–52,
https://doi.org/10.1016/j.ejps.2016.08.046.]
Revised Manuscript with Changes Marked Click here to access/download;Revised Manuscript with
Changes Marked;3 JDDST-D-23-02773 Manuscript
A Multifaceted Approach for Grading of Polymers for the

Development of Stable Amorphous Solid Dispersion of Riluzole
Kanchan Bharti1, Gurudutt Dubey2, Manish Kumar1, Abhishek Jha1, Manjit1, Mansi
Upadhyay3, Pramod S Mali3, Ashutosh Kumar3, Prasad V. Bharatam2, Brahmeshwar
Mishra1*
1Department of Pharmaceutical Engineering and Technology, Indian Institute of Technology
(Banaras Hindu University), Varanasi, U.P. 221005, India
2Department of Medicinal Chemistry National Institute of Pharmaceutical Education and
Research, S.A.S Nagar, Punjab 160062, India
3Department of Biosciences and Bioengineering, Indian Institute of Technology Bombay,
Powai, Mumbai 400076, India
*Corresponding author at Department of Pharmaceutical Engineering & Technology, Indian
Institute of Technology (Banaras Hindu University), Varanasi- 221005, India
Email addresses:
kanchan.bharti.rs.phe18@itbhu.ac.in (Kanchan Bharti)
dubeygurudutt93@gmail.com (Dr. Gurudutt Dubey)
manishkumar.rs.phe19@itbhu.ac.in (Manish Kumar)
abhishekjha.phe17@itbhu.ac.in (Abhishek Jha)
manjit.rs.phe20@itbhu.ac.in (Manjit)
mansiupadhyay14@gmail.com (Dr. Mansi Upadhyay)
malips01@gmail.com (Pramod S Mali)
ashutoshk@iitb.ac.in (Prof. Ashutosh Kumar)
pvbharatam@niper.ac.in (Prof. Prasad V. Bharatam)
bmishrabhu@rediffmail.com (Prof. Brahmeshwar Mishra)
1
Abstract
The current work involves grading three widely used polymers for preparing a kinetically and
thermodynamically stable amorphous solid dispersion (ASD) of a neuroprotective drug
Riluzole (RLZ) by evaluating the drug-polymer interactions. Polymers were screened based on
their chemical interaction with RLZ and their ability to inhibit the crystallization of the drug.
Detailed computational studies were performed to quantify the non-covalent interactions
between RLZ and the modeled structures of polymers; polyacrylic acid (PAA),
polyvinylpyrrolidone vinyl acetate (PVP VA), and hydroxypropyl methyl cellulose acetate
succinate (HPMC AS) for calculating the interaction energies of drug-polymer complexes.
Experimental characterization of drug-polymer complexes through analytical techniques
further validated the formation of the drug-polymer complexes. The results indicated that RLZ
interacts with the polymers in the order PAA > PVP VA > HPMC AS, giving an insight into
the stability of drug-polymer complexes. In vitro dissolution study also showed higher
dissolution profile of RLZ with PAA. Further, the drug-polymer miscibility was determined
by employing Flory-Huggins theory and stability in accordance with Gibbs free energy of
mixing and phase diagram. This work presented a multi-technique approach for the grading of
polymers in search of a stable ASD of the poorly water-soluble drug RLZ.
Keywords: Amorphous Solid Dispersion, Drug-Polymer Interactions, Density functional
theory, Computational study, Miscibility, Flory-Huggins theory.
1 Introduction
A large number of active pharmaceutical ingredients (API) and marketed drug products give
poor therapeutic results due to their poor aqueous solubility in water of the drug. Poorly soluble
drugs have low bioavailability, compromised therapeutic efficacy, and dose escalation. This
2
problem is a major bottleneck in the development of an effective drug delivery system
especially for orally administered formulations. The amorphous solid dispersions (ASDs) and
co-amorphous systems have higher solubility owing to their high thermodynamic properties
and high internal energy; however, these very properties of amorphous solids render them
highly unstable. Therefore, they need to be transformed into a form that is stable enough. ASDs
with drugs in an amorphous form inand a highly stable state can be a potential tool for both
solubility as well as stability enhancement [1]. Development of such ASDs is a widely
employed technique for drug bioavailability enhancement owing to its enhanced solubility
profile with respect to their crystalline counterpart and high stability [2]. ASD consists of a
polymeric matrix for drug dispersion which acts as a crystallization inhibitor (CI). In the
presence of CI, the amorphous drug can form a supersaturated solution that has a substantially
higher solubility than its counterpart crystalline drug’s saturation solubility. The higher degree
of supersaturation favors drug concentration gradient guided drug transport across the
gastrointestinal lumen, responsible for higher drug bioavailability and thus improved
therapeutic activity [3].
Polymer selection is a vital step in designing a stable ASD. Physical stability is imperative in
the case of ASDs and is governed by the phase behavior during storage. The type of polymer
highly influences the formation of ASD and its stability, which is an interplay of the anti-
plasticizing effect of polymer, increase in the glass transition temperature (T g) of the system,
and reduction in the chemical potential of the drug and its molecular mobility [4]. These cohort
effects demand an improved approach of mechanistic understanding of drug and polymers
interplay in crystallization inhibition [5]. Single-phase amorphous binary systems such as
ASDs and co-amorphous solids exhibit interactions like van der Waal forces, ionic interactions,
and hydrogen bonding that are responsible for the inhibition of crystallization [6], [7].
Therefore, drug-polymer interactions need to be studied well to determine the homogenous
3
single phase of drug and polymer for a stable ASD. The solubility and miscibility study of the
drug in the polymer matrix can be used as an markerindicator for the interaction of drug and -
polymer interaction and ASD formation, as phase separation can occur in an amorphous-
amorphous or amorphous-crystalline phase that depends upon the miscibility of drug into
polymer [4], [8].
Screening of polymers can be done by predicting miscibility through in silico and experimental
methods [9]. Analytical techniques like FT-IR and solid-state NMR (ssNMR) are capable of
identifying the non-covalent interactions between the drug and polymers. However, the
strength of such interactions is also very important. This could be possible by employing in
silico tools. In silico tool has emerged as a powerful technique for the virtual screening of
polymers based on the basis of the degree of interactions between the drug and the polymer in
systems like ASD [10], [11], nano-micelle [12], nanoparticles [13], dendrimers, [14] etc.
The current study involvesed a detailed study to comprehendunderstand the drug-polymer
interactions in the formulation of a kinetically and thermodynamically stable ASD. Riluzole
(RLZ) is a glutamate antagonist and the only drug approved by US FDA for amyotrophic lateral
sclerosis by oral route. It belongs to BCS Class II and thus has low aqueous solubility which
limits its bioavailability. The polymers having Tg lower or equivalent to the melting point
((TmTm)) of RLZ were selected and screened for this study. The initial screening of the suitable Formatted: Subscript
Formatted: Not Superscript/ Subscript
polymers for ASD preparation and understanding the molecular level interactions were carried
out by extensive in silico studies. The experimental validation of the results was done by
powder X-ray diffraction (PXRD), differential scanning calorimetry (DSC), fourier transform
infrared spectroscopy (FT-IR), solid-state NMR (ssNMR), microscopic analysis by using
scanning electron microscopy (SEM) and transmission electron microscopy (TEM).
Mathematical modeling using Flory-Huggins theory was also used to determine the extent of
drug loading and the temperature, up to which an ASD system will uphold its homogenous
4
phase. The studies aided in understanding the mechanistic insight of drug’s interaction with
polymers in terms of Tg, drug crystallization, demixing, drug-polymer miscibility, phase
solubility of drugs in a polymer matrix, physical stability of the drug, and preparation process,
that may significantly affect the physical stability of ASDs.
2 Materials and Method
Crystalline RLZ was received as a gift sample from Alkem Laboratories, Maharashtra, India.
The three polymers PAA, PVP VA and HPMC AS was received as gift samples from Lubrizol
Life Science, Mumbai, Maharashtra India. Phosphorus pentoxide was purchased from Central
Drug House (P) Ltd. Analytical grade anhydrous methanol and dichloromethane were used for
the preparation of ASD.
2.1 Methodology for in silico studies
2.1.1 Preparation of 3D structures of polymers
The 3D structures of polymers were designed using polymer builder tool in Macromodel
module of Schrodinger software package [15]. Chain growth with backbone dihedral option
was set to variable between 120-210 degrees to avoid clashes. The energy minimization of the
models was performed on the single polymer chains using OPLS3 force field in vacuum using
Powell-Reeves conjugate gradient (PRCG) method. The resultant optimized geometries of
polymers were subjected to 20 ns of molecular dynamics simulations using Desmond module
[16] of Schrodinger software package. The most stable forms of polymer chains were taken for
further study. To design the polymers: PAA, PVP VA and HPMC AS, following monomeric
compositions were considered based on the literature reports:
For polyacrylic acid (PAA): 19 units of acrylic acid and one unit of acrylate allyl
pentaerythreitol were used in a linear combination (AAAAAAAAAABAAAAAAAAA) to
5
obtain a polymer of length equal to 20 units. The molecular weight of designed PAA polymer
was 1600 with the chemical formula C68H96O43. Figure 1a represents the general 2D structure
of the polymer [17].
For polyvinylpyrrolidone/vinyl acetate (PVP VA): Vinyl acetate and vinyl pyrrolidone were
linked by linear random free-radical polymerization at 1:1 weight ratio (8:12 molar ratio,
respectively) as shown in Figure 1b [18]. The molecular weight and formula for 20 monomers
containing PVP VA were 1698 and C80H112N10O30, respectively.
For HPMC AS: Block copolymerization of 5 units of substituted cellulose was carried out in
ABCAB sequence, wherein A represents the cellulose unit with two acetyl, three methoxy and
one hydroxypropoxy substitutions in place of R. B represents the cellulose unit with two acetyl,
two succinoyl and two methoxy substitutions. C represents the cellulose unit with two acetyl,
two methoxy, one succinoyl and one hydroxypropoxy substitutions [19]. The general structure
of HPMCAS is shown in Figure 1c. The designed HPMC AS had the molecular weight and
chemical formula 2428 and C102H164O65, respectively.
b’ c’ g’ h’ i’
a. b. O c.
RO R = -COCH3 (acetyl), -OCH3 (methoxy) Formatted: Centered
a’ c’ OR
CO2H CO2H RO b’ j’ k’ l’
O N d’ O i’ a’ O O
j’ O
e’ O RO
-CH2CH(CH3)OH (hydroxypropoxy)
a’ d’ m’ n’ o’ p’
b’ n f’ g’ RO f’ OR
e’ x h’ y n -COCH2CH2CO2H (succinoyl)
b’ c’
a. b. O c.
a’ OR RO
c’
CO2H CO2H RO b’
N d’
O a’ O
O i’ O
j’ O
e’ O RO
a’ d’
b’ n f’ g’ RO f’ OR
e’ x h’ y n
g’ h’ i’
d. R = -COCH3 (acetyl), -OCH3 (methoxy)
O j’ k’ l’
F3C S
NH2 m’ n’ o’ p’
N -COCH2CH2CO2H (succinoyl)
Figure 1. General structural representation of the monomeric units of a) PAA, b) PVP VA, and c)
HPMC AS, and d) RLZ.
6
ForPreparation of 3D structure of RLZ: The chemical structure of RLZ is shown in Figure 1d. Formatted: Font: Italic
The initial crystal structure of RLZ was adopted from Cambridge Crystallographic Data Centre
(CSD Deposition number: 1820861). The coordinates of the 3D structure of RLZ were obtained Formatted: Font: Not Italic
from the crystal structure using Mercury software. The same coordinates were used to generate
3D model of RLZ for further computational calculations.
The choice of the 20 monomeric units in PAA and PVV VA, and 5 monomeric units in HPMC
AS is based on an assumption that the optimized structures of the three polymers should possess
equivalent surface area and flexibility to interact with RLZ molecule. During molecular
dynamics simulation and quantum chemical energy minimization of drug-polymer complexes,
the uniformity in the surface area and volume of the complexes can minimize the possibility of
inconsistencies.
2.1.2 Molecular docking:
Molecular docking analysis of RLZ on designed polymers was donecarried out using Glide
module of Schrodinger software package [20]. RLZ was docked at multiple sites of the
polymers to obtain the best possible sites of interactions and binding affinity of the drug-
polymer complexes.
2.1.3 Molecular dynamics simulations (MDS):
MDS were carried out using Desmond module of Schrodinger package. Systems were built in
vacuum using orthorhombic box with buffer size of 10 Å buffer from all the sides using OPLS3
force field. Simulations were run using NPT ensemble at 300 K temperature. Polymers were
simulated for 20 ns at 2 ps record interval and 8ps trajectory. The drug-polymer complexes
were simulated for 40 ns at 2 ps record interval and 8ps trajectory.
7
2.1.4 Density functional theory (DFT) calculations
DFT calculations were performed using Gaussian09 program suite [21] to obtain the drug-
polymer interaction energies in the complexes. The 3D geometries of the three polymers were
optimized at semi empirical PM6 level of theory [22] and the 3D structure of RLZ molecule
was optimized at B3LYP/6-31+G(d, p) level of theory [23]. RLZ-polymer complex werewas
optimized using ONIOM (Our own n-layered Integrated molecular Orbital and Molecular
mechanics) method [24], wherein the polymer was optimized as low layer at semi-empirical
PM6 level and RLZ was optimized as high layer at B3LYP/6-31+G(d, p) level. All the
geometries were optimized in both gas phase and solvent phase using IEFPCM solvent model
[25] and methanol MeOH (=32.7) as the implicit solvent. Since, methanol was used as the
polar protic solvent to dissolve the RLZ and polymers in experimental conditions, it was the
choice of solvent to study the effect of solvent on drug-polymer interaction energies. The
interaction energies of drug-polymer complexes were calculated using the following formula:
Einteraction = Ecomplex – [Epolymer + Edrug]
Where Ecomplex, Epolymer and Edrug are the Gibbs free energies of drug-polymer complex, polymer
and drug respectively.
2.2 Preparation of amorphous solid dispersion and physical mixtures of RLZ and polymer
ASDs with various different ratios of drug-polymer ratio and polymers (Table 1) were
prepared by rapid solvent evaporation method by taking RLZ and the polymers in the ratio
mentioned in the Ttable 1. The solvent system methanol:dichloromethane (1:1) was used as
solvent to solubilize RLZ and the polymers to prepare uniform solutions. The resultant solution
was subjected to vacuum evaporation on water bath with temperature set to 55 oC and gradually
applying vacuum from 500 mBar to 2 mBar. The residual solvent was removed by drying the
8
ASDs in vacuum oven at 25 oC for 24 hours. The samples were passed through 150 µm sieve
and stored over P2O5 until further analysis.
Different weight ratios of RLZ and each of the polymer were taken and geometrically mixed
to avoid any kind of change in the drug’s solid form of the drug. RLZ was mixed in the
composition of 95% w/w, 90% w/w, 85% w/w and 80% w/w with each of the polymer. For T g
determination, the physical mixtures (PM) were prepared with 90% w/w, 70% w/w and 50%
w/w of RLZ with each of the polymer. The PM were stored over P2O5 until use.
Table 1: Composition details for the preparation of ASD with different polymers Formatted: Line spacing: single
Formulation %w/w of
S.No Drug-/Polymer Formatted: Font: 11 pt
Code Drug
1 F1A RLZ-PAA Formatted: Font: 11 pt
2 F2A 50 RLZ-PVP VA Formatted: Font: 11 pt
3 F3A RLZ-HPMC AS
4 F1B RLZ-PAA Formatted: Font: 11 pt
5 F2B 70 RLZ-PVP VA Formatted: Font: 11 pt
6 F3B RLZ-HPMC AS Formatted: Font: 11 pt
7 F1C RLZ-PAA
8 F2C 90 RLZ-PVP VA Formatted: Font: 11 pt
9 F3C RLZ-HPMC AS Formatted: Font: 11 pt
Formatted: Font: 11 pt
2.2.1 High Performance Liquid Chromatography
To check the chemical stability of RLZ in the prepared ASDs, high-performance liquid Formatted: Font: 11 pt
chromatography (HPLC) system from Waters Milliford, USA, was used. The system was
equipped with a photodiode array (PDA) detector, model 2998 from Waters, USA. The LC
system comprised of a binary pump (model 1525; Waters, USA), a manual injector valve with
a 20 µL loop, and a C18 column measuring 150 mm × 4 mm with a particle size of 5 µm
particle size.
For the analysis, a sample volume of 20 µL was injected, andnd the detection of the sample
detections was conducted at a wavelength of 264 nm. Acetonitrile and water in the ratio of
65:35 at with a flow rate of 1 ml/min .was used as the mobile phase.The data acquisition process
was performed using the Breeze program. The mobile phase consisted of a mixture of
9
acetonitrile and water in a ratio of 65:35, with a flow rate of 1 ml/min. The entire analysis
process was completed within a run time of 10 minutes. The data acquisition process was
performed using the Breeze program. The plots confirming the chemical stability are provided
in supporting information (Figure S1).
2.3 Powder X-ray diffraction study
Powder XRD study was done on benchtop Rigaku Miniflex 600 X-ray diffractometer to
confirm the formation of ASDs prepared with different polymers and their different weight
ratios. Analysis was carried out at 40 kV voltage, 15 mA current and Ni-filtered Cu-K radiation
with 20 mm monochromator slit was used to take the pattern of samples. The radiation scattered
on the samples was measured with a vertical goniometer. X-ray diffraction patterns were
obtained at the 2θ values ranging from of 5° to 50° at the scan speed of 5°/min with a step size
of 0.02° [26].
2.4 Determination of melting point and glass transition temperature
The thermal behavior of physical mixtures of RLZ with PAA, PVP VA, HPMC AS and ASDs
of RLZ prepared with these polymers were analyzed through DSC. The melting point of the
pure crystalline RLZ was determined by observing the presence of endothermic peak when the
ramp rate was set at 10oC/min until 140 oC.
The physical mixtures in different composition (90% w/w, 70% w/w, 50% w/w of drug) were
subjected to heat-cool-heat cycle in the DSC. Samples were heated till 140 oC, at a rate of 10
oC/min followed by cooling at a rate of 20 oC/min to 0 oC. From 0 oC onwards, the sample was
reheated until 140 oC at the rate of 20 oC/min. The Tg value from the second heating run was
recorded. For the To determination ofe the Tg of RLZ, it was heated till its melting point and
kept for 1 min at the same temperature to ensure complete melting. To the melted RLZ, liquid
10
nitrogen was poured over it to convert it into amorphous form. Its Tg was determined at a
heating rate of 20 oC/min till 140 oC in DSC.
2.5 FT-IR characterization
FT-IR spectra was obtained in the region of 400-4000 cm-1 in attenuated reflectance mode
(Bruker, Massachusetts, U.S.A) assisted with ZnSe crystal sample holder. Each sample (pure
RLZ, physical mixture of drug and polymer and ASDs of RLZ with each of the polymer in
50:50 weight ratio) was placed on the sample holderATR chamber and scanned between the
given range with an average of sixty-four scans and a resolution of 2 cm-1 [27], [28].
2.6 Solid-state NMR
To study the molecular and structural insight to the drug-polymer interactions, ssNMR
experiments were recorded at 298 K using Bruker 750 MHz Avance III spectrometer connected
with a variable temperature unit. Topspin 3.6.2 pl6 and Topspin 4.1.3 Bruker NMR software
were used for data acquisition and processing. ssNMR experiments were carried out on a 3.2
mm HXY triple resonance probe, and for better sensitivity, it was configured to double
resonance mode. 1D 13C cross polarization magic angle spinning (CP-MAS) experiment was
recorded with an optimized cross-polarization (CP) contact time of 3 ms at a MAS frequency
of 12 kHz. Linear gradient pulse of ramp 100-70 was used during cross-polarization. 1H 90o
pulse of 3.3 mS corresponding to a RF of power 76 kHz and 13C power was kept ~ 56 kHz.
Proton RF power of 76 kHz was used during acquisition of the decoupling. A total of 1024
scans were acquired for each spectrum with a spectral width of 56818 Hz, corresponding to an
acquisition time of 18 ms. Data processing was done using the window function (exponential
multiplication) with a line broadening of 50 Hz. The interscan delay was 5 s for all experiments.
11
2.7 Crystallization tendency of amorphous solid dispersion
Thermal analysis of RLZ, polymers and RLZ ASDs prepared with different drug polymer
compositions (90:10, 70:30, 50:50, 30:70 and 10:90) were subjected to heat-cool-heat cycle.
The temperature range for the analysis wasere -20-140 oC. The starting point of temperature
was at 25 oC. The heating rate was 10 oC/min [29].
2.8 Microscopic analysis
Scanning electron microscopy (SEM) and transmission electron microscopy (TEM) studies
were conducted for determining the morphology of drug of the drug, ASDs and the state of
ASDs after its exposure to 90 oC for 12 hours. The microscopic analysis was done to detect the
possible phase separation on exposure to a higher temperature. The study of phase separation
due to moisture exposure was avoided pertaining to its unknown effect on solubilities [30]. The
study was done to access the thermodynamic stability of ASDs of RLZ with each of the three
polymers. For this analysis, thin ASD films of RLZ with the respective polymers in a 1:1
mixture of DCM and methanol was prepared. 100 uµl of solution was spin coated (EZ spinSD,
Apex Instruments, India) on a 1cm2 glass substrate and 200 mesh copper grids to be analyzed
under SEM ((MA15/18, CARL ZEISS MICROSCOPY LTD) and Tecnai G2T20 LaB6 TEM
operating at 200 kV respectively..
2.9 In-vitro dissolution study
The rate of release of RLZ from ASD (50:50 RLZ:-polymer ratio) versus pure crystalline RLZ
was assessed employing USP type-II apparatus (Electrolab India Pvt Ltd.). 1000 mL of
phosphate buffer (pH 6.8) was used for the dissolution experiments, which were carried out at
37.0± 0.5 °C with a paddle rotating at a speed of 50 rpm [31]. 20 mg of RLZ and ASD
equivalent to 20 mg of RLZ were filled inside hard gelatin capsules and added in each vessel.
12
5 mL of aliquot was taken at each time interval (5, 15, 30, 45, 60, 90, and 120 min). The volume
was kept constant by adding equal volume of fresh media. HPLC method as described above
was used to determine the concentration of RLZ. The study was done in triplicate and mean
values are reported.
2.10 Drug polymer miscibility study
2.10.1 Melting Point Depression Study
For melting point depression study, 3-5 mg of samples (100% w/w 95% w/w, 90% w/w, 85%
w/w and 80% w/w of drug with each polymer) were taken in hermitically sealed aluminum
pans and heated under DSC from 20 oC to 140 oC at a rate of 2 oC/min. During the DSC run50
ml/min nitrogen flow was maintained at 50 ml/minthroughout the DSC run. The readings have
been taken in triplicate and mean of the onset of melting point have been reported.
2.10.2 True density
The density of the drug and the polymers were determined using a helium pyrometer. Prior to
the measurements, the samples were kept over P2O5 for 24 hours. The readings were taken in
triplicate [321].
3 Results and Discussion
3.1 In silico studies
3.1.1 Preparation of 3D structures of polymers
The selection of the most appropriate polymer for preparing an ASD depends upon the
interactions between drug and polymer. In order to assess the drug-polymer interactions,
accurate or close to accurate 3-dimensional structures of both drug and the polymer are
required. Therefore, 3D structures of the three polymers PAA, PVP VA and HPMC AS were
13
modelled initially using computational tools. Modelled polymer chains were subjected to
molecular dynamic simulations to obtain the stable conformation of the chains. PAA polymer
chain consisting of 20 monomers of acrylic acid acquires a linear geometry (Figure S2a, in
supporting information). Similarly, PVP VA also exists in linear form containing vinyl acetate
and vinyl pyrrolidone linked alternatively (Figure S2b). On the other hand, HPMC AS block
copolymer acquires a stable bowl shape conformation by forming various intramolecular
hydrogen bonds as shown in Figure S2c. Formatted: Font color: Auto
Formatted: Font color: Auto
The choice of the 20 monomeric units in PAA and PVP VA, and 5 monomeric units in HPMC
AS is based on assumption that the optimized structures of the threThe choice of the 20 Formatted: Not Highlight
monomeric units in PAA and PVP VA, and 5 monomeric units in HPMC AS is based on an
assumption that the optimized structures of the three polymers should possess equivalent
surface area and flexibility to interact with RLZ molecule. During molecular dynamics Formatted: Not Highlight
simulation and quantum chemical energy minimization of drug-polymer complexes, the
uniformity in the surface polymers should possess equivalent surface area and flexibility to
interact with RLZ molecule. During molecular dynamics simulation and quantum chemical
energy minimization of drug-polymer complexes, the uniformity in the surface area and Formatted: Not Highlight
volume of the complexes can minimize the possibility of inconsistencies. possibility of Formatted: Not Highlight
inconsistencies. DFT calculation are best performed at the small model system of large Formatted: Font color: Auto
polymers to save the calculation time and computational cost, which allow more detailed and
precise structural investigations [33], [34].. The molar ratio of polymer to drug in the three3 Formatted: Font color: Auto
ASDs in the experimental condition vary for 50%w/w drug load at w/w% because of the
variable molecular weights of the three polymers. For PAA-RLZ ASD, the polymer to drug Formatted: Font color: Auto
molar ration is 1923:1, for PVP VA-RLZ ASD, it is 218:1 and for HPMC AS-RLZ ASD, it is
77:1. The DFT calculations of the large polymers collapse because of high degree of freedom Formatted: Font color: Auto
14
and many local minima energy points. When the model system is extended to the actual length,
the similar interactions are expected with the other RLZ molecules with the long polymer
chains in ASDs. Formatted: Font color: Green
3.1.2 Drug-polymer interaction
To determine possible drug-polymer interactions, drug-polymer complexes are obtained in
three steps. Firstly, To determine possible drug-polymer interactions, RLZ was docked on the
simulated geometries of PAA, PVP VA and HPMC AS at multiple sites around the polymeric
chains so as to obtain the initial RLZ-polymer complexes for further quantification of
interaction energies. At the second step, the docked complexes were subjected to molecular
dynamics simulation to obtain the best fit of drug within the flexible polymer cavity. In the last
step, DFT calculations were performed on the simulated complexes to calculate the precise
interaction energies between drug and polymer. In the docking studies, The the most stable
docked poses resulted the complexes with docking score -4.15, -3.83 and -4.14 kcal/mol for
PAA, PVP VA and HPMC AS, respectively. Based on the initial assessment through molecular
docking the stability of RLZ-polymer complexes was found to be in the order RLZ-PAA >
RLZ-HPMC AS > RLZ-PVP VA. In the three complexes, the free NH2 group of RLZ shows
hydrogens bonding interactions with the oxygen of carboxylic acid or ester groups of the
polymers. RLZ resides at the surface of the polymers, however the additional hydrogen
bonding interaction is visible between ring nitrogen of RLZ and carboxylic acid group of
HPMC AS. The C–H···F and van der Waals interactions in RLZ-PAA lead to increase the
docking score higher than that of RLZ-HPMC AS (Figure S3). However, the docking scores
usually do not take the electron density based interaction into account. Thus, To to further
validate the order of the stability and quantify the strength of interaction in terms of energy
values more robust computational algorithm that is DFT calculations were performed. RLZ,
polymers and RLZ-polymer complexes were optimized using ONIOM method, wherein RLZ
15
was optimized using density functionals and polymer segment was optimized using semi-
empirical method. The drug-polymer complex formation is an endergonic process for which
the interaction energies were calculated in gas phase and in MeOH solvent. The interactions
energies for RLZ-PAA, RLZ-PVP VA and RLZ-HPMC AS in gas phase were calculated to be
2.29, 2.64 and 15.24 kcal/mol, respectively. The interaction energies in MeOH solvent were
found to be 10.75, 11.37 and 25.06 kcal/mol for RLZ-PAA, RLZ-PVP VA and RLZ-HPMC
AS, respectively. Based on the interaction (complexation) energy values, the order of the
stability was RLZ-PAA > RLZ-PVP VA > RLZ-HPMC AS under both gas phase and solvent
conditions. The normalized interactions energies were calculated with respect to RLZ-PAA as
shown in Figure 2. Since, ASDs do not contain the solvent molecule, the gas phase energy
value can be correlated with the experimental results. The optimized geometries of the RLZ-
PAA complex indicates that RLZ is fitting well inside the polymer cage. The free PAA was
transformed from linear conformation to a folded conformation in the RLZ-PAA complex.
There are four types of prominent molecular level interaction visible in the RLZ-PAA, (i) a
strong hydrogen bond between ring nitrogen of RLZ and COOH group of PAA, (ii) a strong
hydrogen bond between NH2 group of RLZ and COO– group of PAA, (iii) C–H···O interaction
between RLZ and COOH, and (iv) C–H··· interaction between CH2 linker of polymer and 
cloud of thiazole ring of RLZ (Figure 2a). In case of RLZ-PVP VA, the most prominent
interactions are hydrogen bonding interactions of the two hydrogens of NH 2 of RLZ with
COO– and pyrrolidone carbonyl oxygen of PVP VA (Figure 2b). The strong hydrogen bonding
interaction of NH2, and ring nitrogen of RLZ with carboxylic groups of HPMC AS are visible
in RLZ-HPMC AS complex. An additional C–H···O interaction is also visible; (Figure 2c). It
was observed that major conformational change occurred in PAA as a result of drug-polymer
complexation whereas PVP VA does not allow much conformational change due to its rigid
structure. The free HPMC AS polymer was already stabilized by forming multiple
16
intramolecular hydrogen bonds and thus does not undergo further conformational changes after
complexation with RLZ. This molecular level observation indicates that PAA may be the
suitable polymer for forming a stable drug-polymer non-covalently bonded complex, when a
slight energy is provided. PVP VA may be the next better polymer of choice. In order to
compare the propensity of RLZ to undergo crystallization, an attempt was made to calculate
the non-covalent bonding interaction energies between two RLZ molecules through DFT
calculation. The interaction energy for RLZ-RLZ was calculated to be -10.18 kcal/mol, which
reflect very strong interactions due to bidirectional dual hydrogen-bonding as shown in figure
S4. However, the crystallization behaviour of drug in 1:1 ratio of drug-polymer and drug-drug
cannot be compared through DFT calculation, because it provides the quantification of
intermolecular interactions at smaller model systems. The interaction energy of -10.18 kcal/mol
justifies the requirement of heating (enthalpy) to break these drug-drug intermolecular
interactions within the crystal lattice of drug.
17
Figure 2. RLZ-polymer complexes along with their respective interaction energies. Bond distances are
in Å, interaction energies are in kcal/mol. Energy values are the complexation energy values
referenced to PAA-RLZ in italics and in brackets are obtained under implicit solvent condition using
MeOH as the solvent.
Docking and MD simulation studies depend upon molecular mechanics wavefunction which is Formatted: Font: 12 pt, Font color: Auto
Formatted: Justified, Space After: 0 pt, Line spacing:
suitable for computational studies of large systems like polymer, lipids and nanocarriers etc. Double
These studies can provide the qualitative determination of molecular behavioral patterns in the
simulated conditions. On the other hand, DFT studies depend upon ab initio or semi empirical
wavefunctions which are more robust and provide more accurate estimation of energy values.
However, DFT studies can be performed on truncated model systems within the agreeable time
frame and available computational facilities. DFT studies are more suitable when quantitative
comparison and more precise atomic details of molecular interactions are required such as bond
distance and bond angle for the interactions. On one hand, simulation of larger system
corresponds to the more realistic systems with diverse chemical behavior, while on the other
hand, the smaller systems lead to quick, more accurate calculations and atomic details but limits
the study due to a gap between simulated and realistic systems. The selection of computational
methods depends upon the formulation scientist. For example, MD simulation techniques are
more suitable when one wants to calculate the physicochemical properties like miscibility,
solubility or glass transition temperature of ASDs. DFT calculations are preferred choice when
estimation of molecular level interactions, energy calculations or polymer profiling is desired. Formatted: Font: 12 pt, Font color: Auto
3.2 Powder X-ray diffraction study
The PXRD pattern of pure crystalline RLZ shows distinct sharp peaks at 9.14 o, 13.64o, 18.18o,
22.74o, and 25.22o, which indicates the crystalline nature of the drug (Figure 3). The PXRD
patterns of ASDs with different drug loading (50, 70, and 90 %w/w) with PAA, PVP VA, and
HPMC AS show the transformation of the sharp peaks to the absolute halo patterns at 50%
w/w drug loading, which indicate the presence of single-phase binary amorphous system. At
18
70% w/w, a few indistinct sharp peaks can be observed at 19.68o, and 25.3o in the case of ASD
RLZ-HPMC AS. In ASD RLZ-PVP VA (70:30) also minor sharpness at 19.68o was observed.
After increasing the drug loading to 90% w/w, crystallinity was observed with all three
polymers in the order PAA<PVP VA<HPMC AS. Shift in the two theta values RLZ from
18.18o to ~19.4o and 25.22o to 26.6o could be attributed to the occurrence of polycrystallinity
in RLZ after association with the polymer [352]. This is an indication of the presence of two
phases in the analyzed ASD. This concludes that the ASD formation capacity of polymers with
RLZ is in the order of PAA>PVP VA>HPMC AS.
a) b) c) Formatted: Font: 12 pt
Figure 3. Overlay spectra of PXRD patterns of ASDs having (a) 50:50 of RLZ with polymers, b)
70:30 of RLZ with polymer, and c) 90:10 of RLZ with polymer.
3.3 Determination of melting point and glass transition temperature
Tghe glass transition temperature is the indicator of molecular mobility, where lower molecular Formatted: Subscript
mobility refers to high physical stability of the amorphous systems. DSC is well known for its
evaluation of transition process occurring in amorphous solid at a higher temperature at which
the surrounding molecules influences the orientation and reorientation of the molecules [363].
Tm and Tg of RLZ, physical mixtures of RLZ with each polymer, and ASDs are shown in Figure
4. A melting endotherm was shown by RLZ at 120 oC demonstrating its pure crystalline nature,
while amorphous RLZ showed a Tg at 28 oC (Figure S54). When 90:10 ratio of drug-polymer
were subjected to heat-cool-heat cycle in the DSC for in situ ASD preparation, and the
19
determination of their Tg, sharp endotherm was observed with all the three polymers. This
confirmed the absence of in situ ASD formation at 10% w/w ratio. At 70:30 ratio of the drug-
polymer, the Tg of RLZ-PAA and RLZ-PVP VA were observed at 91 oC and 39 oC,
respectively. However, RLZ-HPMC AS did not show any change in the specific heat capacity
and, a melting endotherm was observed confirming that HPMC AS was not efficient to form
ASD with RLZ at this weight ratio as well. At 50:50 ratio of drug-polymer, with all three
polymers glass transition temperatures were observed. A Tg of 106 oC, 53 oC, and 40 oC were
observed for RLZ-PAA, RLZ-PVP VA, and RLZ-HPMC AS, respectively. In the
aforementioned cases, a single Tg was observed whenever in situ ASD formation occurred
inside the DSC. This clearly indicates the formation of a miscible system since the obtained Tg
of in situ formed ASDs were between the Tg of pure amorphous drug and polymer (Table 2)
[374]. This study implies that PAA is imparting most promising stability to the amorphous
RLZ followed by PVP VA and HPMC AS.
Table 2. Tg and Tm of different solid forms of RLZ and ASD with different drug polymer compositions. Formatted: Space After: 0 pt
Drug to polymer
S.No Composition Tg (oC) Formatted Table
weight ratio
1 RLZ-PAA 106
2 50:50 RLZ-PVP VA 53
3 RLZ-HPMC AS 40
4 RLZ-PAA 91
5 70:30 RLZ-PVP VA 39
6 RLZ-HPMC AS -
7 RLZ-PAA -
8 90:10 RLZ-PVP VA -
9 RLZ-HPMC AS -
10a NA PAA 102
11b NA PVP VA 101
12c NA HPMC AS 117
13 NA RLZ (Amorphous) 28
a
Ref [385], bRef [30], cRef [396]
20
Figure 4. Overlay of DSC plots of in situ ASDs of RLZ with each of the polymer in different drug to
polymer weight ratios formed inside the DSC with heat-cool-heat cycle
3.4 FT-IR characterization
Figure 5 represents the comparative FTIR spectra of RLZ, PMs, and ASDs of RLZ-polymer at
a 50:50 weight ratio. In the spectra of RLZ the prominent peaks at 3364 cm -1, 1543 cm-1, 1458
cm-1,1293 cm-1 and 1143 cm-1 correspond to N-H stretching, C=C stretching, C-F stretching,
C-O stretching, and C-N stretching, respectively. Comparing the spectra of RLZ with the
physical mixture of each polymer, minor differences like the shift from 3364 cm-1 to 3368 cm-
1, 1538 cm-1 to 1543 cm-1, 1293 cm-1 to 1300 cm-1 and 1143 cm-1 to 1203 cm-1 were observed
in PM RLZ-PAA. In PM RLZ-PVP VA, peaks were observed at 3366 cm-1 and 1541 cm-1. In
PM RLZ-HPMC AS, the characteristic peaks were observed at 3323 cm-1, 1248 cm-1, and 1213
cm-1. However, in the case of ASDs, much more changes in the wavenumber were observed.
The N-H stretching peaks were broadened significantly in the ASDs. The C=C stretching peaks
were shifted to 1528 cm-1 (in ASD RLZ-PAA) and 1538 cm-1 (in ASD RLZ-PVP VA and ASD
RLZ-HPMC AS). The C-O stretching, C-N stretching and C-F stretching peaks were also
21
shifted in the ASD samples. These shifts in wavenumber confirmed the formation of ASD and
presence of interactions between drug and polymer. The evidence of strong drug-polymer
interactions assures the formation of a highly stable ASD by increasing the onset temperature
of crystallization and by attenuating the crystallization of the drug [4037].
Figure 5. FTIR spectra of RLZ, physical mixture of RLZ with polymers and ASD of RLZ with each
of the respective polymer
3.5 Solid-state NMR
ssNMR is a very powerful tool to identify the presence of molecular interactions between the
drug and the polymer, responsible for the stability of ASD. The 13C CPMAS spectra of
crystalline RLZ, pure polymers, and RLZ ASD with each of the polymers in the ratio of 70:30
were recorded to study the interactions occurring at the molecular level. The typical sharp
resonances in the spectra of pure RLZ are reflecting the crystalline nature of RLZ. In the spectra
of ASDs, the peaks of both drug and polymer are merged hence confirming the formation of
molecular level dispersion. The broad Gaussian resonances are indicative of disorder in the
crystal arrangement as seen in amorphous solid [4138]. ASD RLZ-PAA is showing more
22
nonhomogeneous line broadening with respect to ASDs RLZ-PVP VA and RLZ-HPMC AS,
which could be possible due to more random distribution of the chemical environment
subjected to more loss of crystallinity [4239] in comparison to the former and the presence of
a small amount of crystalline drug in the later two spectra. This is also in close agreement with
the results obtained in PXRD studies and in silico studies.
The overlay of pure RLZ and ASD RLZ-PAA 13C CP/MAS spectra shows the downfield shift
in the quaternary carbon (Ca) of the thiazole ring from 169 to 170 ppm, as shown in Figure 6a.
Another quaternary carbons (Ce and Cc) of pure RLZ occurring at 150 and 143 ppm,
respectively, are getting merged with the C b in the case of ASD RLZ-PAA. A similar
observation is found for another aromatic carbon Ch and Cg, which are occurring at 132 and
129 ppm, respectively. The trifluoromethyl carbon (Cf) shown at 120 ppm is also merged with
the aromatic Cd carbon of RLZ. The downfield shift from 178 ppm to 180 ppm in the carbonyl
carbon (Ca’) and from 37 ppm to 43 ppm in the methylene carbon (Cb’) of PAA [430] indicate
that these carbons are experiencing the electron-withdrawing effect through non-covalent
interactions. The merging of peaks and changes in the 13C chemical shifts explain the rigidity
in the structure due to the prominent interaction between RLZ and PAA functional groups.
In the case of PVP VA (Figure 6b), the resonance at 175 and 168 ppm are assigned to the
carbonyl carbons of the pyrrolidone (Ca’) and acetate (Ci’) functionality, respectively.
Resonances of 65 and 34, 18 and 16 ppm are assigned to Cg’, Cf’, Cc’ and Cj’ carbon
respectively. However, the carbon present in the pyrrolidone ring (Cb’) and the vinyl chain (Ch’)
have been assigned resonance values of 40 and 29, respectively [441]. Despite the ASD RLZ-
PVP VA showing evidence of the formation of the molecular dispersion, there is clearly the
presence of crystalline RLZ which is evident by the sharp resonances. The interaction could be
seen in the shortening and the slight deshielding of the carbons present in the range of 40-15
ppm of PVP VA. Hence, it can be inferred that despite of interaction shown by the polymer the
23
crystalline structure of RLZ persists making it a two-phased dispersion. In Figure 6c, the
anomeric carbons labeled as a', d', b', c', and e', j' are responsible for resonances lying between
101-58 ppm for polymer HPMC AS. At 67 and 56 ppm there are shoulders due to methylene
Cf', and methoxy carbons (Ci' and Ch'), respectively. The most shielded carbons lying in the
range of 26-14 ppm are those in the acetyl (Ch’), succinoyl chain (Cn’ and Co’) and methyl
carbon (Cl’) in the hydroxypropyl chain. Interactions between RLZ and polymer in ASD RLZ-
HPMC AS can be seen through the deshielding of carbons lying between 100-58 ppm of HPMC
AS, but the resonances of RLZ seem unaffected in terms of chemical shifts; however, there is
a change in the peak intensities from that of the crystalline RLZ.
Hence, it can be concluded that the interaction and the capacity to make a stable amorphous
solid dispersion system by the polymer is in the order of PAA > PVP VA > HPMC AS. These
results corroborate the findings from the in silico work.
24
a)
Cb’
Ca’
Cb
Cf
Ca Cd
Ce Ch Cg
Cc
b)
Cd’,e’ Cb’,h’
Cc’
Cj’
Cf’
Ca’
Ci’ Cg’
c)
Cb’,c’
Ce’,k’
Cd’ Cf’,j’ Ch’

Cg’ Ca’ Ci’
Cm’,p’ Cn’,o’
Cl’
25
Figure 6. The 13C CP-/MAS spectra of RLZ in comparison with a) PAA and ASD RLZ-PAA, b) PVP
VA and ASD RLZ-PVP VA, and c) HPMC AS and RLZ-HPMC AS
3.6 Crystallization tendency of amorphous solid dispersion
Heat-cool-heat cycle in DSC was used to assess the crystallization tendency of drug in the
prepared ASDs. The thermograms of RLZ ASDs prepared with different composition of drug
and polymer are shown in Figure S56 (a to f). Fusion and crystallization enthalpies of RLZ and
RLZ ASDs are mentioned in table 3. In the thermograms it can be seen that the samples with
higher weight percentage of drug (90% w/w) the cooling crystallization event is prominent
inwhile ASDs prepared with all the polymers. The enthalpy of fusion values are in close
agreement with the enthalpy of crystallization values showing the extent of crystallization
occurring in 90% w/w drug loaded ASDs. Exothermic effect due to the crystallization of RLZ
were not observed in the ASDs prepared with polymer wight percentage of 30% to 90%. A
slight event of crystallization was observed in RLZ ASDs prepared HPMC AS at 70% w/w
drug loading. The crystallinity estimated from the fusion of enthalpy were around 15%, 16%
and 29% for 90:10 RLZ-PAA, RLZ-PVP VA, and RLZ-HPMC AS ASDs respectively. This
gives an indication that solid dispersions of RLZ when prepared with PAA and PVP VA will
offer highest resistance against drug recrystallization followed by ASDsolid dispersions
prepared with HPMC AS. The higher value of crystallinity describes the higher mobility of
HPMC AS [29]. This confirms the higher kinetic stability of ASDs prepared with PAA and
PVP VA with respect to HPMC AS.
Table 3: Fusion enthalpies (ΔHm) and crystallization enthalpies (ΔHcr), with corresponding onset Formatted: Justified, Space After: 0 pt
temperatures of RLZ and its ASDs with polymers in different compositions.
Enthalpy ΔHcr (Jg-1) ΔHm (Jg-1)
RLZ -332.94 (87.25 oC) 341.82 (119.36 oC)
ASDs 90:10 70:30 50:50 30:70 10:90 90:10 70:30 50:50 30:70 10:90
RLZ: PAA -49.94 - - - - 52.87 - - - -
RLZ: PVP
-51.65 - - - - 55.57 - - - -
VA
RLZ:
-95.18 -23.88 - - - 101.73 - - - -
HPMC AS
26
The TEM and SEM images of ASDs have been taken in pre- and post-annealed conditions.
TEM with the selected area electron diffraction (SAED) patterns was used to identify the
regions where crystallinity might have developed due to annealing. Figure 7 shows the bright
field images of the ASDs for the pre- and post-annealed conditions. A large sample size (50
areas on each grid) was chosen to study the morphology of RLZ ASDs. The bright field TEM
images (Figure 7 (a, e and i)) show the spherical morphology of ASDs prepared with the
polymers. The diffused rings produced by the particles in the SAED pattern confirm their
amorphous nature [452]. The TEM images of all three ASDs were observed for the post-
annealed condition and found that no significant change has occurred in RLZ-PAA ASD
(Figure 7c). However, in the RLZ-PVP VA and RLZ-HPMC AS ASD, the bright field images
along with SAED showed a high degree of crystallization as there was the presence of a regular
array of diffraction spots (Figure 7 (g,h) and (k,l)) [463], [474]. The SAED pattern consists of
systematic spots which means the grains or crystallites formed are in the ranges of nm length
scale. Along with the array of diffraction, diffused circular rings can also be seen in the pattern
explaining the presence of both amorphous and crystalline regions. To confirm the presence of
RLZ only, the crystallographic information was collected from the CIF file [485] and it showed
that the compound belongs to the triclinic system with a space group of P1̅ suggesting that the
crystal has got center of inversion (centrosymmetric). As per the crystallography data available
on the Cambridge Crystallographic Data Centre (CCDC), the lattice parameters of RLZ are:
a=8.08 Å, b= 11.78 Å, c=19.74 Å and α = 78.44 o, β= 84.378o, γ= 89.318o. The information
obtained from the zone axis table of Figure 7 h and l indicated that the crystals present have
the zone axis of [311] and [-112] which corresponds to the RLZ crystals.
The presence of RLZ crystals due to demixing after annealing from the supersaturated ASD is
in agreement with the Tg results. It appears that due to high Tg, ASD RLZ-PAA exhibited
27
superior stability in comparison to RLZ-PVP VA and RLZ-HPMC AS ASD and retained its
amorphous form at higher temperatures. This confirms the suitability of PAA as polymer over
PVP VA and HPMC AS for the preparation of a thermodynamically stable ASD with RLZ.
The chemical stability of all the three ASDs of RLZ with PAA, PVP VA and HPMC AS, which
were estimated through HPLC, was in line with the thermodynamic and kinetic stability. The
retention time (Rt ~ 3.9 min) of RLZ in the ASDs was same as that in the pure RLZ sample,
which confirmed the unchanged that there was no change in the chemical environment around
the pure RLZ molecules (Figure S1). This may also confirm that after demixing the
recrystallized component was RLZ only.
SEM images of ASDs in pre- and post-annealed conditions have been shown in Figure S67a-
f.
a) b) c) d)
e) f) g) h)
i) j) k) l)
Figure 7. Bright-field images of (a) freshly prepared RLZ-PAA ASD, (c) annealed at 90oC for 12 hrs
RLZ-PAA ASD, (e) freshly prepared RLZ-PVP VA ASD, (g) annealed at 90oC for 12 hrs RLZ-PVP
VA ASD, (i) freshly prepared RLZ-HPMC AS ASD, (k) annealed at 90oC for 12 hrs RLZ-HPMC AS
ASD, and its corresponding diffraction patterns (b), (d), (f), (h), (j) and (l) respectively.
28
3.9 In this study 50:50 RLZ:polymer ASD were taken considering the fact that at this ratio Formatted: Indent: Left: 0.1", No bullets or
all the ASDs showed complete amorphization as shown in PXRD study (section 3.2). RLZ is
a weakly basic drug with an aqueous solubility of 0.3 mg/ml at neutral pH and 12 mg/ml
(approx.) at acidic pH of 1.2 [49]. It has a pKa of 3.47, and it shows higher solubility at lower
pH and lower solubility at neutral to high pH [50]. Out of the three polymers, HPMC AS
shows solubility at pH above 5 [51], hence selection of acidic medium for dissolution would
not have provided the comparative results, hence pH 6.8 medium was selected for dissolution
study. The dissolution profile of all three ASDs versus crystalline RLZ is shown is Figure
S78. RLZ:PAA ASD showed maximum release of drug (100%) in the initial 45 min of study
followed by RLZ:PVP VA (91%) and RLZ:HPMC AS (87%). Pure crystalline RLZ is having
a low dissolution profile as compared to the ASD formulation, however it was also showing
around 83% release in initial 45 min. Almost every solid state of drug showed 100% drug
release in 120 mins. The difference between the dissolution profile of each ASD formulation
is smallless, which canould be attributed to the higher sink condition that is maintained during
the dissolution study. SinceAs ASDs are supersaturated drug delivery systems, they tend to
increase the dissolution of any poorly aqueous soluble drug. ButHowever,, this may lead to
drug precipitation due to supersaturation;, hence, higher sink condition was maintained [52].
Formatted: Heading 2
3.103.9 Drug polymer miscibility and phase diagram
The determination of miscibility of drug in polymer is widely estimated by Flory-Huggins Formatted: Indent: Left: 0.1"
theory. This theory explains the thermodynamics of the mixing of polymers with solvent
molecules by using the Flory-Huggins interaction parameter χ [5346]. The degree and type
29
of interaction play a major role in free energy of mixing (ΔGmix), which can be determined by
equation 1.
Δ𝐺𝑚𝑖𝑥 Φ𝑝𝑜𝑙𝑦
𝑅𝑇
= Φ𝑑𝑟𝑢𝑔 𝑙𝑛Φ𝑑𝑟𝑢𝑔 + 𝑚
𝑙𝑛Φ𝑝𝑜𝑙𝑦 + χΦ𝑑𝑟𝑢𝑔 Φ𝑝𝑜𝑙𝑦 …Eq 1
Where Φ is the volume fraction of the drug and the polymer; R is the universal gas constant; T
is the absolute temperature of the system; m is the ratio of the volume of a polymer chain to
drug molecular volume [5447], [5548]. Equation 1 shows the contribution of the entropy and
enthalpy of mixing of drug and polymer, which are represented by the first and second terms,
respectively. Entropy is the decisive factor in the determination of spontaneous mixing since
enthalpy always favors the mixing. It is so because the enthalpy of mixing cannot surpass a
small critical value unless any phase separation occurs.
To determine the entropy of mixing the two components, interaction parameter (χ) (shown in
equation 2) was calculated by the melting point depression method, which is an indicator of
the decreased chemical potential of drug in the mixture with respect to the pure crystalline drug
[5649].
1 1 𝑅 1 2
𝑇𝑚
− 𝑜
𝑇𝑚
= − Δ𝐻 [ln Φ𝑑𝑟𝑢𝑔 + (1 − 𝑚) Φ𝑝𝑜𝑙𝑦 + χΦ𝑝𝑜𝑙𝑦 ] …Eq 2.
𝑓
Where Tm is the drug’s melting temperature of the drug in the mixture of drug and polymer,
𝑇𝑚𝑜 is the melting point of the pure crystalline drug, Φ is the volume fraction of the drug and
polymer and ΔHf is the heat of fusion of the pure crystalline drug. The DSC plots shown in
Figure S89 indicate the depression in the onset of the melting point of RLZ with varying weight
fractions (Φpoly) of each polymer. Figure S98 shows that increase in weight fraction of each
polymer is proportionately related to the depression in the onset of melting point of the
crystalline RLZ. This verifies that there is a certain degree of mixing between the drug and
polymer near the melting temperature of the drug in the respective compositions of physical
30
mixtures. Using equation 2 (values of the required parameters are mentioned in Table 4), χ of
drug with each of the polymers were calculated. The calculated χ were plotted against
temperature empirically described by equation 3 to show its dependence on temperature.
χ=A+ B/T …Eq 3.
Where A and B/T are the relative contributions of entropy (temperature-independent) and the
enthalpy, respectively towards the free energy of mixing. T is the melting temperature of drug
in a drug-polymer mixture. Equation 3 is a representation of straight-line equation, where the
value of R2 were 0.9477, 0.9817, and 0.9410, respectively, for physical mixtures of RLZ-PAA,
RLZ-PVP VA, and RLZ-HPMC AS showing the goodness of fit in χ vs 1/T plot (shown in
Figure S109). Hence, by using the value of A and B, the value of χ was extrapolated for even
lower temperatures.
Table 4: Physical properties of RLZ, PAA, PVP VA and HPMC AS used for quantitative drug-polymer Formatted: Justified
miscibility.
True Molar
Molecular ΔHf Tm
S.No. Components Density volume Formatted: Font: 11 pt
weight (g/mol) (kJ/mol) (K)
(g/cm3) (cm3/mol)
Formatted: Line spacing: single
1 RLZ 234 1.66 140.96 18043.74 391.98 Formatted Table
2 PAA 450000 1.33 338345.86 - - Formatted: Font: 11 pt
3 PVP VA 51000 1.13 45132.74 - - Formatted: Space After: 0 pt, Line spacing: single
4 HPMC AS 18000 1.23 14634.15 - - Formatted: Font: 11 pt
Formatted: Space After: 0 pt, Line spacing: single
In the case of drug-polymer miscibility, since the miscible diluent is a polymer that, which is
amorphous in nature, the entropy of mixing becomes negligible [570]. Hence, the second Formatted: Font: 11 pt
derivative of the free energy (equation 1) is equated to zero and as stated in equation 4, the
maximum drug-polymer miscibility boundary may be calculated as:

2𝐵
𝑇𝑆 = 1 1
…Eq 4.
( )+ ( −2𝐴)
Φ𝑑𝑟𝑢𝑔 𝑚(1−Φ𝑑𝑟𝑢𝑔 )
31
Drug solubility and miscibility can be plotted as a function of drug weight fraction and
temperature after the understanding of how ΔGmix varies as a function of T for each system.
ΔGmix is an indicator of a stable drug-polymer system and was calculated using χ through
equation 1 and then ΔGmix/RT was plotted against the drug volume fraction (Φ drug). The
negative value of ΔGmix in all the three systems was seen across all drug weight fraction near
the melting point of the drug. This shows the thermodynamic stability of all three drug-polymer
systems near drug’s melting point. ΔGmix/RT vs Φdrug were also plotted at Tg shown by each
drug-polymer system prepared by in-situ ASD formation inside DSC instrument (discussed in
section 3.3). In this curve shown in Figure 8 a, b, and c it was seen that all the drug-polymer
system at 106oC (Tg of 50:50 RLZ-PAA ASD), ΔGmix is negative and convex across all Φdrug.
However, at 53 oC (Tg of RLZ-PVP VA ASD) and at 40 oC (Tg of 50:50 RLZ-HPMC AS ASD),
all the systems showed instability due to positive ΔGmix at their respective Tg, which may result
into heterogeneous drug-polymer mixture. It is noteworthy that RLZ-PAA showed negative
ΔGmix at its Tg while RLZ-PVP VA and RLZ-HPMC AS showed positive ΔGmix. Contrary to
this result, in experimental settings, we observed single Tg of freshly prepared ASDs, which
correspond to the single phased drug-polymer system. Usually, immiscible-systems, two
different glass transition temperatures are observed, one for drug and another for polymer
[518]. Miscibility study is mainly concerned with the thermodynamic stability of ASD but
alone cannot govern the absolute stability of the ASD because it does not account the kinetic
stability [474]. At room temperature as well, the positive ΔGmix was showing the instability of
all the compositions of drug-polymer. From equation 1 and 2, it can be seen that smaller the
value of χ more negative will be the ΔGmix, designating the drug-polymer complex to be more
stable and a greater number of interactions between drug and polymer [581]. Therefore,
positive values of χ obtained at 25 oC showed the formation of immiscible drug-polymer system
(Table 5). The negative value of χ is obtained at <94 oC, <80 oC and <66 oC for RLZ-PAA,
32
RLZ-PVP VA and RLZ-HPMC AS, respectively (Figure 8 d, e, and f). This highlights the
temperature at which the drug-polymer system may exist as a homogenous phase. The value
of interaction parameter was observed to be influenced by each factor encompassing Flory-
Huggins model, thus could not be used for comparison between different drug-polymer systems
[592]. So, the χ values for each drug-polymer system at 25 oC were not sufficient to conclude
the grading of polymer in terms of its efficacy in complete amorphousization to give stable
ASDs. Hence, the system was further evaluated by solubility and miscibility curves.
Table 5: Values of Constant A, Constant B and interaction parameter (Ӽ) for Three RLZ-Polymer Formatted: Justified
Systems.
Drug-Polymer Interaction Parameter (χ)
S.No.
System
Constant A Constant B Formatted: Font: 11 pt
at 25oC At melting point
1 RLZ-PAA -210.73 77574 49.45 -12.83
Formatted Table
2 RLZ-PVP VA -45.08 16009 8.61 -4.24
Formatted: Space After: 0 pt
3 RLZ-HPMC AS -155.86 53012 21.94 -20.62 Formatted: Font: 11 pt
Henceforth, further insight to understand the potential of polymer towards formation of stable
ASD was taken by preparing the phase diagram of miscibility temperature and solubility Formatted: Font: 11 pt
temperature (calculated using equation 4) vs drug volume fraction. This information was paired
with predicted Tg of the system with different weight ratios of drug and polymer, which were Formatted: Space After: 0 pt, Line spacing: single
calculated using Gordon-Taylor equation (equation 5).
𝑊1 𝑇𝑔1 + 𝐾𝐺 𝑊2 𝑇𝑔2 Formatted: Line spacing: single
𝑇𝑔 = 𝑊1 + 𝐾𝐺 𝑊2
…Eq 5
Where Tg, Tg1, and Tg2 are the glass transition temperatures of the drug-polymer mixture, drug
in amorphous form and the polymer, respectively. The constant value KG can be calculated
using equation 6.
33
𝜌1 𝑇𝑔1
𝐾𝐺 = 𝜌2 𝑇𝑔2
…Eq 6
Where ρ1 and ρ2 are the true density value of amorphous drug and polymer, respectively. This
phase diagram concerns the range of temperature and composition, below and above which the
system exists in single-phase and two-phase thermodynamic region. Through phase diagram
(Figure 8 d, e and f), information about the maximum solubility and miscibility of amorphous
drug within the polymeric matrix and its dependence on temperature can be obtained. The
region lying at the right-hand side of the spinodal Tg curve is considered as unstable where the
drug-polymer exists in two phase heterogenous system. To the left-hand side exists the region
on drug-polymer single-phase homogenous system. At the right-hand side of the Tg curve
recrystallization of the drug can occur without any significant energy barrier. The solubility
and the miscibility curve intersecting the predicted glass transition curve can be seen.
Quantitative determination of miscibility of RLZ in PAA, PVP VA and HPMC AS was
observed to be approximately 10% w/w at <85 oC, 25% w/w at <71 oC and 35% w/w at <61
oC, respectively. This revealed the extent of drug loading possible until these temperatures
achieved which could result into a stable ASD of RLZ without rendering any sort of
recrystallization. Additionally, in the case of PAA, and PVP VA at 50%w/w drug loading there
is positive deviation in the Tg value i.e., the observed value is higher than predicted. In case of
HPMC AS at the same drug loading a negative deviation in the Tg value is observed. A positive
deviation is the evidence of heteronuclear interaction which leads to enhanced stability. On the
other hand, the negative deviation is the evidence of homonuclear interaction which has
destabilizing effect [5548]. From this, it can be concluded that PAA can offer high stability to
RLZ over higher temperatures; however, drug loading will be low. PVP VA observed to offer
higher drug loading among the three polymers with stability over moderate temperature. While
HPMC AS can produce a stable ASD with the highest drug loading but observed to have
temperature limited lower thermodynamic stability compared to PAA and PVP VA. However,
34
low drug loading observed in case of PAA can be easily addressed by adjusting the ASDs
amount as dose of RLZ was not very high (50mg taken orally twice daily). In addition, it may
not result in any pill burden at the cost of its stability. This experimental methodology is based
on the measurement of the dissolution of crystalline drug into polymeric matrix at high drug
loading and temperature. In this method, data generated is extrapolated to lower temperatures,
which is often argued to give overestimation or underestimation of miscibility of drug in
polymer. Therefore, this study is backed up with other studies to give a multi-faceted
understanding of the interaction of drug with the selected polymers. For instance,
intermolecular interactions like the formation of H-bonds or ionic bonds (as discussed in earlier
study) plays the key role in determining the capacity of polymer to prevent the drug present in
the ASD to recrystallize. The evaluation of the results obtained from all the studies should be
considered to generate evidences in grading of polymers according to their efficacy to produce
a stable ASD. Present study revealed PAA as effective polymer based on all above studies
conducted for grading of polymers for development of RLZ ASDs.
35
14
a) 25 C 40 C 53 C 106 C 120 C 120 d)
12 Solubility Miscibility Predicted Tg
10 100
8
Temperature (oC) 80
6
ΔGmix/RT
4 60
2
40
0
0 0.2 0.4 0.6 0.8 1
-2 20
-4
Φdrug 0
-6 0.0 0.2 0.4 0.6 0.8 1.0
Φdrug
120
2 25 C 40 C 53 C 106 C 120 C Solubility Miscibility Predicted Tg
b)
e)
1.5 100
Temperature (oC)
1
80
0.5
ΔGmix/RT
60
0
0 0.2 0.4 0.6 0.8 1
-0.5 40
-1
20
-1.5
Φdrug 0
-2 0.0 0.2 0.4 0.6 0.8 1.0
Φdrug
6 25 C 40 C 53 C 106 C 120 C 125 f)

c)
Solubility Miscibility Predicted Tg
4
105
2
Temperature (oC)
ΔGmix/RT
0 85
0 0.2 0.4 0.6 0.8 1
-2 65
-4
45
-6
Φdrug
-8 25
0.0 0.2 0.4 0.6 0.8 1.0
ΦDrug
Figure 8. a) Plot of ΔGmix/RT vs drug weight fraction for RLZ-PAA b) Plot of ΔGmix/RT vs drug
weight fraction (Φdrug) for RLZ-PVP VA c) Plot of ΔGmix/RT vs drug weight fraction (Φdrug) for RLZ-
HPMC AS d) Binary phase diagram for RLZ-PAA e) Binary phase diagram for RLZ-PVP VA f)
Binary phase diagram for RLZ-HPMC AS showing the solid liquid equilibrium (blue), miscibility
(orange) and glass transition temperature (grey) curve.
4 Conclusion
The current study provides a multi-pronged approach in deciding the use of most appropriate
polymer according to its ability to provide stability to the ASD system. The initial screening of
36
polymer done by molecular modeling and DFT studies, deciphered the mode of interactions
between the drug and polymer along with the quantification of interactions in terms of energies.
The XRD studies of the developed ASD with different drug and polymer ratios, revealed the
capacity of polymer to stabilize a complete amorphous system. DSC thermogram showed the
homogenous phase by exhibiting single Tg for each ASD. In addition to this, in situ ASD
preparation helped in determining the drug-polymer system giving the Tg value of 106 oC, 53
oC and 40 oC in 50:50 drug/-polymer weight ratio, for ASD RLZ-PAA, RLZ-PVP VA and
RLZ-HPMC AS respectively. This defines the capacity of polymers to increase the low Tg (28
oC) of pure amorphous RLZ, hence enhancing its stability. Chemical interaction studies with
the help of FT-IR and ssNMR showed the formation of a molecular dispersion and the possible
interactions occurring between the drug and polymer. Chemical stability was also confirmed
using HPLC analysis. Assessment of crystallization tendency of the prepared ASDs indicated
that PAA and PVP VA showed lower phase separation as the crystallization and fusion
enthalpies were lower with respect to ASDs prepared with HPMC AS. This study evaluated
the kinetic stability of ASDs prepared with different drug-polymer compositions. TEM studies
were capable in detecting the presence of crystalline drug even in nanometer scale after it was
exposed to high temperature, to ensure its thermodynamic stability. Additionally, in vitro
dissolution study also showed superior dissolutionrug release capacity of RLZ:PAA ASD over
other ASDs. The drug-polymer miscibility studies elaborated the stability space lying between
the percent drug loading and the temperature. ΔGmix of mixing at room, glass transition and
melting temperatures of the drug-polymer systems were also taken into account. Hence, it can
be inferred that PAA within its drug loading capacity would provide maximum stability to the
RLZ ASD followed by PVP VA and HPMC AS. This multi-faceted approach will aid a faster
and empirical selection of the most suitable polymer for the development of ASD which will
be stable throughout its stipulated shelf-life. Even though these studies provide sufficient
37
insight into the stability of ASDs, a real time stability study at accelerated conditions would
provide exact results, which will be the future scope of this study. By exploiting this approach
industries can equip themselves to produce more such products; as till date, commercially
available ASD products are very limited.
Acknowledgment
Authors would like to express their appreciation to Dr. Jagadish Sharma of the National
Institute of Pharmaceutical Education and Research, S.A.S Nagar for his invaluable guidance
throughout the project. Authors also extends their sincere thanks to Dr. Aman Kumar Lal Das
of the Department of Metallurgical Engineering, IT (BHU) for his assistance in TEM studies.
Additionally, authors would like to acknowledge the support of the Central Instrument Facility
at IIT (BHU) for providing the necessary resources to conduct the analytical studies.
Data Availability:
The raw/processed data required to reproduce these findings cannot be shared at this time as
the data also forms part of an ongoing study. However, original data included in this study will
be shared prior to publication, if required.
References
1. Baird, J.A., Taylor, L.S., 2012. Evaluation of amorphous solid dispersion properties using
thermal analysis techniques. Adv. Drug Deliv. Rev., The role of thermal analysis and
calorimetry in pharmaceutical design and development 64, 396–421.
https://doi.org/10.1016/j.addr.2011.07.009
2. Bhujbal, S.V., Mitra, B., Jain, U., Gong, Y., Agrawal, A., Karki, S., Taylor, L.S., Kumar,
S., (Tony) Zhou, Q., 2021. Pharmaceutical amorphous solid dispersion: A review of
manufacturing strategies. Acta Pharm. Sin. B, Hot Topic Reviews in Drug Delivery 11,
2505–2536. https://doi.org/10.1016/j.apsb.2021.05.014
3. Azad, M., Moreno, J., Davé, R., 2018. Stable and Fast-Dissolving Amorphous Drug
Composites Preparation via Impregnation of Neusilin® UFL2. J. Pharm. Sci. 107, 170–182.
https://doi.org/10.1016/j.xphs.2017.10.007
38
4. Thakore, S.D., Akhtar, J., Jain, R., Paudel, A., Bansal, A.K., 2021. Analytical and
Computational Methods for the Determination of Drug-Polymer Solubility and Miscibility.
Mol. Pharm. 18, 2835–2866. https://doi.org/10.1021/acs.molpharmaceut.1c00141
5. Wilson, V.R., Lou, X., Osterling, D.J., Stolarik, D.F., Jenkins, G.J., Nichols, B.L.B., Dong,
Y., Edgar, K.J., Zhang, G.G.Z., Taylor, L.S., 2020. Amorphous solid dispersions of
enzalutamide and novel polysaccharide derivatives: investigation of relationships between
polymer structure and performance. Sci. Rep. 10, 18535. https://doi.org/10.1038/s41598-
020-75077-7
6. Hiew, T.N., Zemlyanov, D.Y., Taylor, L.S., 2022. Balancing Solid-State Stability and
Dissolution Performance of Lumefantrine Amorphous Solid Dispersions: The Role of
Polymer Choice and Drug–Polymer Interactions. Mol. Pharm. 19, 392–413.
https://doi.org/10.1021/acs.molpharmaceut.1c00481
7. Wang, Y., Grohganz, H. and Rades, T., 2022. Effects of polymer addition on the non-
strongly interacting binary co-amorphous system carvedilol-tryptophan. International
Journal of Pharmaceutics, 617, p.121625. https://doi.org/10.1016/j.ijpharm.2022.121625
8. Sharma, J., Singh, B., Agrawal, A.K., Bansal, A.K., 2021. Correlationship of Drug-Polymer
Miscibility, Molecular Relaxation and Phase Behavior of Dipyridamole Amorphous Solid
Dispersions. J. Pharm. Sci. 110, 1470–1479. https://doi.org/10.1016/j.xphs.2020.12.007
9. He, Y., Ho, C., 2015. Amorphous Solid Dispersions: Utilization and Challenges in Drug
Discovery and Development. J. Pharm. Sci. 104, 3237–3258.
https://doi.org/10.1002/jps.24541
10. Yani, Y., Kanaujia, P., Chow, P.S., Tan, R.B.H., 2017. Effect of API-Polymer Miscibility
and Interaction on the Stabilization of Amorphous Solid Dispersion: A Molecular
Simulation Study. Ind. Eng. Chem. Res. 56, 12698–12707.
https://doi.org/10.1021/acs.iecr.7b03187
11. Wang, B., Wang, D., Zhao, S., Huang, X., Zhang, J., Lv, Y., Liu, X., Lv, G., Ma, X., 2017.
Evaluate the ability of PVP to inhibit crystallization of amorphous solid dispersions by
density functional theory and experimental verify. Eur. J. Pharm. Sci. 96, 45–52.
https://doi.org/10.1016/j.ejps.2016.08.046
12. Rezaeisadat, M., Bordbar, A.-K., Omidyan, R., 2021. Molecular dynamics simulation study
of curcumin interaction with nano-micelle of PNIPAAm-b-PEG co-polymer as a smart
efficient drug delivery system. J. Mol. Liq. 332, 115862.
https://doi.org/10.1016/j.molliq.2021.115862
13. Stipa, P., Marano, S., Galeazzi, R., Minnelli, C., Mobbili, G., Laudadio, E., 2021. Prediction
of drug-carrier interactions of PLA and PLGA drug-loaded nanoparticles by molecular
dynamics simulations. Eur. Polym. J. 147, 110292.
https://doi.org/10.1016/j.eurpolymj.2021.110292
14. Bazyari-Delavar, S., Badalkhani-Khamseh, F., Ebrahim-Habibi, A., Hadipour, N.L., 2020.
Investigation of host-guest interactions between polyester dendrimers and ibuprofen using
density functional theory (DFT). Comput. Theor. Chem. 1189, 112983.
https://doi.org/10.1016/j.comptc.2020.112983
15. Schrödinger Release 2019-4: MacroModel, Schrödinger; LLC: New York, NY, USA, 2019
16. Schrödinger Release 2019-4: Desmond Molecular Dynamics System, D. E. Shaw Research,
New York, NY, 2019. Maestro-Desmond Interoperability Tools, Schrödinger, New York,
NY, 2019
17. Terao, K. (2014). Poly(acrylic acid) (PAA). In: Kobayashi, S., Müllen, K. (eds)
Encyclopedia of Polymeric Nanomaterials. Springer, Berlin, Heidelberg
18. https://coatings.specialchem.com/product/r-ashland-plasdone-s-630-copovidone, accessed
on 16th January 2023
39
19. https://www.ashland.com/file_source/Ashland/Industries/Pharmaceutical/Links/PC-
12624.6_AquaSolve_HPMCAS_Physical_Chemical_Properties, accessed on 16th January
2023.
20. Schrödinger Release 2019-4: Glide, Schrödinger, LLC, New York, NY, 2019
21. Gaussian 09, Revision B.01, M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M.
A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G. A. Petersson, H.
Nakatsuji, M. Caricato, X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L.
Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T.
Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, J. A. Montgomery, Jr., J. E. Peralta, F.
Ogliaro, M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, T. Keith, R.
Kobayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi,
M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox, J. B. Cross, V. Bakken, C. Adamo,
J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli,
J. W. Ochterski, R. L. Martin, K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J.
J. Dannenberg, S. Dapprich, A. D. Daniels, O. Farkas, J. B. Foresman, J. V. Ortiz, J.
Cioslowski, and D. J. Fox, Gaussian, Inc., Wallingford CT, 2010
22. Stewart, J.J.P., 2008. Application of the PM6 method to modeling the solid state. J. Mol.
Model. 14, 499–535. https://doi.org/10.1007/s00894-008-0299-7
23. Hill, J.G., 2013. Gaussian basis sets for molecular applications. Int. J. Quantum Chem. 113,
21–34. https://doi.org/10.1002/qua.24355
24. Vreven, T., Byun, K.S., Komáromi, I., Dapprich, S., Montgomery, J.A.Jr., Morokuma, K.,
Frisch, M.J., 2006. Combining Quantum Mechanics Methods with Molecular Mechanics
Methods in ONIOM. J. Chem. Theory Comput. 2, 815–826.
https://doi.org/10.1021/ct050289g
25. Tomasi, J., Mennucci, B., Cancès, E., 1999. The IEF version of the PCM solvation method:
an overview of a new method addressed to study molecular solutes at the QM ab initio level.
J. Mol. Struct. THEOCHEM 464, 211–226. https://doi.org/10.1016/S0166-1280(98)00553-
3
26. Bharti, K., Mittal, P., Mishra, B., 2019. Formulation and characterization of fast dissolving
oral films containing buspirone hydrochloride nanoparticles using design of experiment. J.
Drug Deliv. Sci. Technol. 49, 420–432. https://doi.org/10.1016/j.jddst.2018.12.013
27. Saxena, D., Maiti, P., 2021. Utilization of ABS from plastic waste through single-step
reactive extrusion of LDPE/ABS blends of improved properties. Polymer 221, 123626.
https://doi.org/10.1016/j.polymer.2021.123626
28. Vikas, Viswanadh, M.K., Mehata, A.K., Sharma, V., Priya, V., Varshney, N., Mahto, S.K.,
Muthu, M.S., 2021. Bioadhesive chitosan nanoparticles: Dual targeting and
pharmacokinetic aspects for advanced lung cancer treatment. Carbohydr. Polym. 274,
118617. https://doi.org/10.1016/j.carbpol.2021.118617
29. Lapuk, S.E., Zubaidullina, L.S., Ziganshin, M.A., Mukhametzyanov, T.A., Schick, C. and
Gerasimov, A.V., 2019. Kinetic stability of amorphous solid dispersions with high content
of the drug: A fast scanning calorimetry investigation. International Journal of
Pharmaceutics, 562, pp.113-123. https://doi.org/10.1016/j.ijpharm.2019.03.039
30. Sun, Y., Tao, J., Zhang, G.G.Z., Yu, L., 2010. Solubilities of Crystalline Drugs in Polymers:
An Improved Analytical Method and Comparison of Solubilities of Indomethacin and
Nifedipine in PVP, PVP/VA, and PVAc. J. Pharm. Sci. 99, 4023–4031.
https://doi.org/10.1002/jps.22251
31. Yadav, B., Balasubramanian, S., Chavan, R. B., Thipparaboina, R., Naidu, V. G. M., &
Shastri, N. R. 2018. Hepatoprotective cocrystals and salts of riluzole: prediction, synthesis,
solid state characterization, and evaluation. Crystal Growth & Design, 18(2), 1047-1061.
https://doi.org/10.1021/acs.cgd.7b01514
40
31.32. Chieng, N., Aaltonen, J., Saville, D. and Rades, T., 2009. Physical characterization and
stability of amorphous indomethacin and ranitidine hydrochloride binary systems prepared
by mechanical activation. European Journal of Pharmaceutics and Biopharmaceutics, 71(1),
pp.47-54. https://doi.org/10.1016/j.ejpb.2008.06.022
33. Walden, D. M., Bundey, Y., Jagarapu, A., Antontsev, V., Chakravarty, K., & Varshney, J.
2021. Molecular simulation and statistical learning methods toward predicting drug–
polymer amorphous solid dispersion miscibility, stability, and formulation design.
Molecules, 26(1), 182. https://doi.org/10.3390/molecules26010182
34. Wang, B., Wang, D., Zhao, S., Huang, X., Zhang, J., Lv, Y., ... & Ma, X. 2017. Evaluate
the ability of PVP to inhibit crystallization of amorphous solid dispersions by density
functional theory and experimental verify. European Journal of Pharmaceutical Sciences,
96, 45-52. https://doi.org/10.1016/j.ejps.2016.08.046
32.35. Yang, G., Park, S.-J., 2019. Deformation of Single Crystals, Polycrystalline Materials,
and Thin Films: A Review. Materials 12, 2003. https://doi.org/10.3390/ma12122003
33.36. Kissi, E.O., Kasten, G., Löbmann, K., Rades, T., Grohganz, H., 2018. The Role of Glass
Transition Temperatures in Coamorphous Drug–Amino Acid Formulations. Mol. Pharm.
15, 4247–4256. https://doi.org/10.1021/acs.molpharmaceut.8b00650
34.37. Forster, A., Hempenstall, J., Tucker, I., Rades, T., 2001. Selection of excipients for melt
extrusion with two poorly water-soluble drugs by solubility parameter calculation and
thermal analysis. Int. J. Pharm. 226, 147–161. https://doi.org/10.1016/S0378-
5173(01)00801-8.
35.38. Chan, C.K. and Chu, I.M., 2001. Effect of hydrogen bonding on the glass transition
behavior of poly (acrylic acid)/silica hybrid materials prepared by sol–gel process. Polymer,
42(14), pp.6089-6093. https://doi.org/10.1016/S0032-3861(01)00091-X
36.39. Al-Obaidi, H. and Buckton, G., 2009. Evaluation of griseofulvin binary and ternary
solid dispersions with HPMCAS. AAPS PharmSciTech, 10, pp.1172-1177.
https://doi.org/10.1208/s12249-009-9319-x
37.40. Mistry, P., Mohapatra, S., Gopinath, T., Vogt, F.G., Suryanarayanan, R., 2015. Role of
the Strength of Drug–Polymer Interactions on the Molecular Mobility and Crystallization
Inhibition in Ketoconazole Solid Dispersions. Mol. Pharm. 12, 3339–3350.
https://doi.org/10.1021/acs.molpharmaceut.5b00333
38.41. Song, Y., Yang, X., Chen, X., Nie, H., Byrn, S., Lubach, J.W., 2015. Investigation of
Drug–Excipient Interactions in Lapatinib Amorphous Solid Dispersions Using Solid-State
NMR Spectroscopy. Mol. Pharm. 12, 857–866. https://doi.org/10.1021/mp500692a
39.42. Pugliese, A., Toresco, M., McNamara, D., Iuga, D., Abraham, A., Tobyn, M.,
Hawarden, L.E., Blanc, F., 2021. Drug–Polymer Interactions in
Acetaminophen/Hydroxypropylmethylcellulose Acetyl Succinate Amorphous Solid
Dispersions Revealed by Multidimensional Multinuclear Solid-State NMR Spectroscopy.
Mol. Pharm. 18, 3519–3531. https://doi.org/10.1021/acs.molpharmaceut.1c00427
40.43. Miyoshi, T., Takegoshi, K., Hikichi, K., 1997. High-resolution solid state 13C n.m.r.
study of the interpolymer interaction, morphology and chain dynamics of the poly(acrylic
acid)/poly(ethylene oxide) complex. Polymer 38, 2315–2320.
https://doi.org/10.1016/S0032-3861(96)00799-9
41.44. Sarpal, K., Delaney, S., Zhang, G.G.Z., Munson, E.J., 2019. Phase Behavior of
Amorphous Solid Dispersions of Felodipine: Homogeneity and Drug–Polymer Interactions.
Mol. Pharm. 16, 4836–4851. https://doi.org/10.1021/acs.molpharmaceut.9b00731
42.45. Mahamid, J., Sharir, A., Addadi, L., Weiner, S., 2008. Amorphous calcium phosphate
is a major component of the forming fin bones of zebrafish: Indications for an amorphous
precursor phase. Proc. Natl. Acad. Sci. 105, 12748–12753.
https://doi.org/10.1073/pnas.0803354105
41
43.46. Li, N., Gilpin, C.J., Taylor, L.S., 2017. Understanding the Impact of Water on the
Miscibility and Microstructure of Amorphous Solid Dispersions: An AFM–LCR and TEM–
EDX Study. Mol. Pharm. 14, 1691–1705.
44.47. S’ari, M., Blade, H., Cosgrove, S., Drummond-Brydson, R., Hondow, N., Hughes, L.P.,
Brown, A., 2021. Characterization of Amorphous Solid Dispersions and Identification of
Low Levels of Crystallinity by Transmission Electron Microscopy. Mol. Pharm. 18, 1905–
1919. https://doi.org/10.1021/acs.molpharmaceut.0c00918
48. Mondal, P.K., T, A., Rao, V., Chopra, D., 2019. Crystal structure analysis of the biologically
active drug mol-ecule riluzole and riluzolium chloride. Acta Crystallogr. Sect. E
Crystallogr. Commun. 75, 1084–1089. https://doi.org/10.1107/S2056989019009022
45.49. Dyer, A. M., & Smith, A. 2016. Riluzole 5 mg/mL oral suspension: for optimized drug
delivery in amyotrophic lateral sclerosis. Drug design, development and therapy, 59-64.
https://doi.org/10.2147/DDDT.S123776
50. https://drugcentral.org/drugcard/2382
51. Brady, J., Dürig, T., Lee, P. I., & Li, J. X. 2017. Polymer properties and characterization. In
Developing solid oral dosage forms (pp. 181-223). Academic Press.
https://doi.org/10.1016/B978-0-12-802447-8.00007-8
52. Thakore, S. D., Thakur, P. S., Shete, G., Gangwal, R., Narang, A. S., Sangamwar, A. T., &
Bansal, A. K. (2019). Assessment of biopharmaceutical performance of supersaturating
formulations of carbamazepine in rats using physiologically based pharmacokinetic
modeling. AAPS PharmSciTech, 20, 1-12. https://doi.org/10.1208/s12249-019-1386-z
46.53. Potter, C.B., Davis, M.T., Albadarin, A.B., Walker, G.M., 2018. Investigation of the
Dependence of the Flory–Huggins Interaction Parameter on Temperature and Composition
in a Drug–Polymer System. Mol. Pharm. 15, 5327–5335.
47.54. Meng, F., Dave, V., Chauhan, H., 2015. Qualitative and quantitative methods to
determine miscibility in amorphous drug–polymer systems. Eur. J. Pharm. Sci. 77, 106–
111. https://doi.org/10.1016/j.ejps.2015.05.018
48.55. Baghel, S., Cathcart, H., O’Reilly, N.J., 2016. Theoretical and experimental
investigation of drug-polymer interaction and miscibility and its impact on drug
supersaturation in aqueous medium. Eur. J. Pharm. Biopharm. 107, 16–31.
https://doi.org/10.1016/j.ejpb.2016.06.024
49.56. Marsac, P.J., Li, T., Taylor, L.S., 2009. Estimation of Drug–Polymer Miscibility and
Solubility in Amorphous Solid Dispersions Using Experimentally Determined Interaction
Parameters. Pharm. Res. 26, 139–151. https://doi.org/10.1007/s11095-008-9721-1
50.57. Rim, P.B., Runt, J.P., 2002. Melting point depression in crystalline/compatible polymer
blends [WWW Document]. ACS Publ. https://doi.org/10.1021/ma00138a017
58. Newman, A., Zografi, G., 2019. Commentary: Considerations in the Measurement of Glass
Transition Temperatures of Pharmaceutical Amorphous Solids. AAPS PharmSciTech 21,
26. https://doi.org/10.1208/s12249-019-1562-1
59. Knopp, M.M., Tajber, L., Tian, Y., Olesen, N.E., Jones, D.S., Kozyra, A., Löbmann, K., Formatted: Indent: Left: -0.1", Hanging: 0.3"
Paluch, K., Brennan, C.M., Holm, R., Healy, A.M., Andrews, G.P., Rades, T., 2015.
Comparative Study of Different Methods for the Prediction of Drug–Polymer Solubility.
51. Formatted: Normal, No bullets or numbering
52. Knopp, M.M., Tajber, L., Tian, Y., Olesen, N.E., Jones, D.S., Kozyra, A., Löbmann, K.,
Paluch, K., Brennan, C.M., Holm, R., Healy, A.M., Andrews, G.P., Rades, T., 2015.
42
Comparative Study of Different Methods for the Prediction of Drug–Polymer Solubility.
Formatted: Font color: Red
Formatted: List Paragraph, Indent: Left: -0.1",
Hanging: 0.3", Numbered + Level: 1 + Numbering
Style: 1, 2, 3, … + Start at: 1 + Alignment: Left +
Aligned at: 0.25" + Indent at: 0.5"
43
Graphical Abstract Click here to access/download;Graphical Abstract;2 Graph Abs 18_8_23.tif
Declaration of Interest Statement
A Multifaceted Approach for Grading of Polymers for the Development of Stable
Amorphous Solid Dispersion of Riluzole
Declaration of Interest
The authors declare no conflicting financial interests or personal connections that may have influenced the
research reported in this paper.
1
Author Statement
A Multifaceted Approach for Grading of Polymers for the Development of Stable Amorphous Solid
Dispersion of Riluzole
Author statement
The work being described in the manuscript has never been published before and is not currently being
considered for publication elsewhere. It has the approval of all authors and the responsible authorities where
the work was carried out. If accepted, it will not be published elsewhere in the same form, in English or in
any other language, including electronically, without the written consent of the copyright-honoring party.
1
Supplementary Material
Click here to access/download

Supplementary Material
8 JDDST-D-23-02773 supporting Rev.docx
Revised Manuscript (Clean Version) Click here to view linked References
1
A Multifaceted Approach for Grading of Polymers for the
2
3 Development of Stable Amorphous Solid Dispersion of Riluzole
4
5
6 Kanchan Bharti1, Gurudutt Dubey2, Manish Kumar1, Abhishek Jha1, Manjit1, Mansi
7 Upadhyay3, Pramod S Mali3, Ashutosh Kumar3, Prasad V. Bharatam2, Brahmeshwar
8 Mishra1*
9
10 1
Department of Pharmaceutical Engineering and Technology, Indian Institute of Technology
11
12
(Banaras Hindu University), Varanasi, U.P. 221005, India
13 2
14 Department of Medicinal Chemistry National Institute of Pharmaceutical Education and
15 Research, S.A.S Nagar, Punjab 160062, India
16
3
17 Department of Biosciences and Bioengineering, Indian Institute of Technology Bombay,
18 Powai, Mumbai 400076, India
19
20 *Corresponding author at Department of Pharmaceutical Engineering & Technology, Indian
21
Institute of Technology (Banaras Hindu University), Varanasi- 221005, India
22
23
24
25
26 Email addresses:
27
28 kanchan.bharti.rs.phe18@itbhu.ac.in (Kanchan Bharti)
29
30
31
dubeygurudutt93@gmail.com (Dr. Gurudutt Dubey)
32
33 manishkumar.rs.phe19@itbhu.ac.in (Manish Kumar)
34
35 abhishekjha.phe17@itbhu.ac.in (Abhishek Jha)
36
37
38 manjit.rs.phe20@itbhu.ac.in (Manjit)
39
40 mansiupadhyay14@gmail.com (Dr. Mansi Upadhyay)
41
42
43 malips01@gmail.com (Pramod S Mali)
44
45 ashutoshk@iitb.ac.in (Prof. Ashutosh Kumar)
46
47
48 pvbharatam@niper.ac.in (Prof. Prasad V. Bharatam)
49
50 bmishrabhu@rediffmail.com (Prof. Brahmeshwar Mishra)
51
52
53
54
55
56
57
58
59
60
61
62 1
63
64
65
Abstract
1
2
3 The current work involves grading three widely used polymers for preparing a kinetically and
4
5
6 thermodynamically stable amorphous solid dispersion (ASD) of a neuroprotective drug
7
8 Riluzole (RLZ) by evaluating the drug-polymer interactions. Polymers were screened based on
9
10
11 their chemical interaction with RLZ and their ability to inhibit the crystallization of the drug.
12
13 Detailed computational studies were performed to quantify the non-covalent interactions
14
15
16
between RLZ and the modeled structures of polymers; polyacrylic acid (PAA),
17
18 polyvinylpyrrolidone vinyl acetate (PVP VA), and hydroxypropyl methyl cellulose acetate
19
20 succinate (HPMC AS) for calculating the interaction energies of drug-polymer complexes.
21
22
23 Experimental characterization of drug-polymer complexes through analytical techniques
24
25 further validated the formation of the drug-polymer complexes. The results indicated that RLZ
26
27
28 interacts with the polymers in the order PAA > PVP VA > HPMC AS, giving an insight into
29
30 the stability of drug-polymer complexes. In vitro dissolution study also showed higher
31
32
33
dissolution profile of RLZ with PAA. Further, the drug-polymer miscibility was determined
34
35 by employing Flory-Huggins theory and stability in accordance with Gibbs free energy of
36
37 mixing and phase diagram. This work presented a multi-technique approach for the grading of
38
39
40 polymers in search of a stable ASD of the poorly water-soluble drug RLZ.
41
42
43 Keywords: Amorphous Solid Dispersion, Drug-Polymer Interactions, Density functional
44
45
46 theory, Computational study, Miscibility, Flory-Huggins theory.
47
48
49
50
51 1 Introduction
52
53
54 A large number of active pharmaceutical ingredients (API) and marketed drug products give
55
56 poor therapeutic results due to their poor solubility in water. Poorly soluble drugs have low
57
58
59
bioavailability, compromised therapeutic efficacy, and dose escalation. This problem is a major
60
61
62 2
63
64
65
bottleneck in the development of an effective drug delivery system especially for orally
1
2 administered formulations. ASDs and co-amorphous systems have higher solubility owing to
3
4
5 their high thermodynamic properties and high internal energy; however, these very properties
6
7 of amorphous solids render them highly unstable. Therefore, they need to be transformed into
8
9
10 a form that is stable enough. ASDs with drugs in an amorphous form in a highly stable state
11
12 can be a potential tool for both solubility as well as stability enhancement [1]. Development of
13
14
15 such ASDs is a widely employed technique for drug bioavailability enhancement owing to its
16
17 enhanced solubility profile with respect to their crystalline counterpart and high stability [2].
18
19
ASD consists of a polymeric matrix for drug dispersion which acts as a crystallization inhibitor
20
21
22 (CI). In the presence of CI, the amorphous drug can form a supersaturated solution that has a
23
24 substantially higher solubility than its counterpart crystalline drug’s saturation solubility. The
25
26
27 higher degree of supersaturation favors drug concentration gradient guided drug transport
28
29 across the gastrointestinal lumen, responsible for higher drug bioavailability and thus improved
30
31
32 therapeutic activity [3].
33
34
35 Polymer selection is a vital step in designing a stable ASD. Physical stability is imperative in
36
37 the case of ASDs and is governed by the phase behavior during storage. The type of polymer
38
39
40 highly influences the formation of ASD and its stability, which is an interplay of the anti-
41
42 plasticizing effect of polymer, increase in the glass transition temperature (Tg) of the system,
43
44
45 and reduction in the chemical potential of the drug and its molecular mobility [4]. These cohort
46
47 effects demand an improved approach of mechanistic understanding of drug and polymers
48
49
50 interplay in crystallization inhibition [5]. Single-phase amorphous binary systems such as
51
52 ASDs and co-amorphous solids exhibit interactions like van der Waal forces, ionic interactions,
53
54
and hydrogen bonding that are responsible for the inhibition of crystallization [6], [7].
55
56
57 Therefore, drug-polymer interactions need to be studied well to determine the homogenous
58
59 single phase of drug and polymer for a stable ASD. The solubility and miscibility study of the
60
61
62 3
63
64
65
drug in the polymer matrix can be used as an indicator for the interaction of drug and polymer
1
2 and ASD formation, as phase separation can occur in an amorphous-amorphous or amorphous-
3
4
5 crystalline phase that depends upon the miscibility of drug into polymer [4], [8].
6
7
8 Screening of polymers can be done by predicting miscibility through in silico and experimental
9
10
11 methods [9]. Analytical techniques like FT-IR and solid-state NMR (ssNMR) are capable of
12
13 identifying the non-covalent interactions between the drug and polymers. However, the
14
15
16
strength of such interactions is also very important. This could be possible by employing in
17
18 silico tools. In silico tool has emerged as a powerful technique for the virtual screening of
19
20 polymers based on the degree of interactions between the drug and the polymer in systems like
21
22
23 ASD [10], [11], nano-micelle [12], nanoparticles [13], dendrimers, [14] etc.
24
25
26 The current study involves a detailed study to comprehend the drug-polymer interactions in the
27
28
29
formulation of a kinetically and thermodynamically stable ASD. RLZ is a glutamate antagonist
30
31 and the only drug approved by US FDA for amyotrophic lateral sclerosis by oral route. It
32
33 belongs to BCS Class II and thus has low aqueous solubility which limits its bioavailability.
34
35
36 The polymers having Tg lower or equivalent to the melting point (Tm) of RLZ were selected
37
38 and screened for this study. The initial screening of the suitable polymers for ASD preparation
39
40
41 and understanding the molecular level interactions were carried out by extensive in silico
42
43 studies. The experimental validation of the results was done by powder X-ray diffraction
44
45
46 (PXRD), differential scanning calorimetry (DSC), fourier transform infrared spectroscopy (FT-
47
48 IR), solid-state NMR (ssNMR), microscopic analysis by using scanning electron microscopy
49
50
(SEM) and transmission electron microscopy (TEM). Mathematical modeling using Flory-
51
52
53 Huggins theory was also used to determine the extent of drug loading and the temperature, up
54
55 to which an ASD system will uphold its homogenous phase. The studies aided in understanding
56
57
58 the mechanistic insight of drug’s interaction with polymers in terms of Tg, drug crystallization,
59
60 demixing, drug-polymer miscibility, phase solubility of drugs in a polymer matrix, physical
61
62 4
63
64
65
stability of the drug, and preparation process, that may significantly affect the physical stability
1
2 of ASDs.
3
4
5
6
7
8
2 Materials and Method
9
10
Crystalline RLZ was received as a gift sample from Alkem Laboratories, Maharashtra, India.
11
12
13 The three polymers PAA, PVP VA and HPMC AS was received as gift samples from Lubrizol
14
15 Life Science, Mumbai, Maharashtra India. Phosphorus pentoxide was purchased from Central
16
17
18 Drug House (P) Ltd. Analytical grade anhydrous methanol and dichloromethane were used for
19
20 the preparation of ASD.
21
22
23
24 2.1 Methodology for in silico studies
25
26
27 2.1.1 Preparation of 3D structures
28
29
30 The 3D structures of polymers were designed using polymer builder tool in Macromodel
31
32 module of Schrodinger software package [15]. Chain growth with backbone dihedral option
33
34
35 was set to variable between 120-210 degrees to avoid clashes. The energy minimization of the
36
37 models was performed on the single polymer chains using OPLS3 force field in vacuum using
38
39
40
Powell-Reeves conjugate gradient (PRCG) method. The resultant optimized geometries of
41
42 polymers were subjected to 20 ns of molecular dynamics simulations using Desmond module
43
44 [16] of Schrodinger software package. The most stable forms of polymer chains were taken for
45
46
47 further study. To design the polymers: PAA, PVP VA and HPMC AS, following monomeric
48
49 compositions were considered based on the literature reports:
50
51
52
53 For PAA: 19 units of acrylic acid and one unit of acrylate allyl pentaerythreitol were used in a
54
55 linear combination (AAAAAAAAAABAAAAAAAAA) to obtain a polymer of length equal
56
57
58
to 20 units. The molecular weight of designed PAA polymer was 1600 with the chemical
59
60 formula C68H96O43. Figure 1a represents the general 2D structure of the polymer [17].
61
62 5
63
64
65
For PVP VA: Vinyl acetate and vinyl pyrrolidone were linked by linear random free-radical
1
2 polymerization at 1:1 weight ratio (8:12 molar ratio, respectively) as shown in Figure 1b [18].
3
4
5 The molecular weight and formula for 20 monomers containing PVP VA were 1698 and
6
7 C80H112N10O30, respectively.
8
9
10
11 For HPMC AS: Block copolymerization of 5 units of substituted cellulose was carried out in
12
13 ABCAB sequence, wherein A represents the cellulose unit with two acetyl, three methoxy and
14
15
16
one hydroxypropoxy substitutions in place of R. B represents the cellulose unit with two acetyl,
17
18 two succinoyl and two methoxy substitutions. C represents the cellulose unit with two acetyl,
19
20 two methoxy, one succinoyl and one hydroxypropoxy substitutions [19]. The general structure
21
22
23 of HPMCAS is shown in Figure 1c. The designed HPMC AS had the molecular weight and
24
25 chemical formula 2428 and C102H164O65, respectively.
26
27
28 b’ c’
a. b. O c.
29 a’ OR RO
c’
30 CO2H CO2H RO b’
N d’
O a’ O
O i’ O
31 j’ O
e’ O RO
a’ d’
32 b’ n f’ g’ RO f’ OR
e’ x h’ y n
33
34 g’ h’ i’
35 d. R = -COCH3 (acetyl), -OCH3 (methoxy)
36 O j’ k’ l’
F3C S
37 NH2 m’ n’ o’ p’
38 N -COCH2CH2CO2H (succinoyl)
39
40
41 Figure 1. General structural representation of the monomeric units of a) PAA, b) PVP VA, c) HPMC
42 AS, and d) RLZ.
43
44 For RLZ: The chemical structure of RLZ is shown in Figure 1d. The initial crystal structure of
45
46 RLZ was adopted from Cambridge Crystallographic Data Centre (CSD Deposition number:
47
48
49 1820861). The coordinates of the 3D structure of RLZ were obtained from the crystal structure
50
51 using Mercury software. The same coordinates were used to generate 3D model of RLZ for
52
53
54 further computational calculations.
55
56
57
58
59
60
61
62 6
63
64
65
2.1.2 Molecular docking:
1
2 Molecular docking analysis of RLZ on designed polymers was done using Glide module of
3
4
5 Schrodinger software package [20]. RLZ was docked at multiple sites of the polymers to obtain
6
7 the best possible sites of interactions and binding affinity of the drug-polymer complexes.
8
9
10
11 2.1.3 Molecular dynamics simulations (MDS):
12
13 MDS were carried out using Desmond module of Schrodinger package. Systems were built in
14
15
16
vacuum using orthorhombic box with buffer size of 10 Å buffer from all the sides using OPLS3
17
18 force field. Simulations were run using NPT ensemble at 300 K temperature. Polymers were
19
20 simulated for 20 ns at 2 ps record interval and 8ps trajectory. The drug-polymer complexes
21
22
23 were simulated for 40 ns at 2 ps record interval and 8ps trajectory.
24
25
26 2.1.4 Density functional theory (DFT) calculations
27
28
29
DFT calculations were performed using Gaussian09 program suite [21] to obtain the drug-
30
31 polymer interaction energies in the complexes. The 3D geometries of the three polymers were
32
33 optimized at semi empirical PM6 level of theory [22] and the 3D structure of RLZ molecule
34
35
36 was optimized at B3LYP/6-31+G(d,p) level of theory [23]. RLZ-polymer complex was
37
38 optimized using ONIOM (Our own n-layered Integrated molecular Orbital and Molecular
39
40
41 mechanics) method [24], wherein the polymer was optimized as low layer at semi-empirical
42
43 PM6 level and RLZ was optimized as high layer at B3LYP/6-31+G(d,p) level. All the
44
45
46 geometries were optimized in both gas phase and solvent phase using IEFPCM solvent model
47
48 [25] and methanol (=32.7) as the implicit solvent. Since, methanol was used as the polar protic
49
50
51 solvent to dissolve the RLZ and polymers in experimental conditions, it was the choice of
52
53 solvent to study the effect of solvent on drug-polymer interaction energies. The interaction
54
55
56
energies of drug-polymer complexes were calculated using the following formula:
57
58
59 Einteraction = Ecomplex – [Epolymer + Edrug]
60
61
62 7
63
64
65
Where Ecomplex, Epolymer and Edrug are the Gibbs free energies of drug-polymer complex, polymer
1
2 and drug respectively.
3
4
5
6 2.2 Preparation of amorphous solid dispersion and physical mixtures of RLZ and polymer
7
8
9 ASDs with various drug-polymer ratio were prepared by rapid solvent evaporation method by
10
11
12 taking RLZ and the polymers in the ratio mentioned in the Table 1. The solvent system
13
14 methanol:dichloromethane (1:1) was used as solvent to solubilize RLZ and the polymers to
15
16
17
prepare uniform solutions. The resultant solution was subjected to vacuum evaporation on
18
19 water bath with temperature set to 55 oC and gradually applying vacuum from 500 mBar to 2
20
21 mBar. The residual solvent was removed by drying the ASDs in vacuum oven at 25 oC for 24
22
23
24 hours. The samples were passed through 150 µm sieve and stored over P2O5 until further
25
26 analysis.
27
28
29
30 Different weight ratios of RLZ and each of the polymer were taken and geometrically mixed
31
32 to avoid any kind of change in the drug’s solid form. RLZ was mixed in the composition of
33
34
35
95% w/w, 90% w/w, 85% w/w and 80% w/w with each of the polymer. For Tg determination,
36
37 the physical mixtures (PM) were prepared with 90% w/w, 70% w/w and 50% w/w of RLZ with
38
39 each of the polymer. The PM were stored over P2O5 until use.
40
41
42
43 Table 1: Composition details for the preparation of ASD with different polymers
44 Formulation %w/w of
S.No Drug-Polymer
45 Code Drug
46 1 F1A RLZ-PAA
47 2 F2A 50 RLZ-PVP VA
48 3 F3A RLZ-HPMC AS
49
50
4 F1B RLZ-PAA
51 5 F2B 70 RLZ-PVP VA
52 6 F3B RLZ-HPMC AS
53 7 F1C RLZ-PAA
54 8 F2C 90 RLZ-PVP VA
55 9 F3C RLZ-HPMC AS
56
57
58
59
60
61
62 8
63
64
65
2.2.1 High Performance Liquid Chromatography
1
2 To check the chemical stability of RLZ in the prepared ASDs, high-performance liquid
3
4
5 chromatography (HPLC) system from Waters Milliford, USA, was used. The system was
6
7 equipped with a photodiode array (PDA) detector, model 2998 from Waters, USA. The LC
8
9
10 system comprised of a binary pump (model 1525; Waters, USA), a manual injector valve with
11
12 a 20 µL loop, and a C18 column measuring 150 mm × 4 mm with 5 µm particle size.
13
14
15
16
For the analysis, a sample volume of 20 µL was injected, and sample detection was conducted
17
18 at a wavelength of 264 nm. Acetonitrile and water in the ratio of 65:35 at a flow rate of 1
19
20 ml/min was used as the mobile phase. The entire analysis process was completed within a run
21
22
23 time of 10 minutes. The data acquisition process was performed using the Breeze program. The
24
25 plots confirming the chemical stability are provided in supporting information (Figure S1).
26
27
28
29 2.3 Powder X-ray diffraction study
30
31
32 Powder XRD study was done on benchtop Rigaku Miniflex 600 X-ray diffractometer to
33
34
35
confirm the formation of ASDs prepared with different polymers and their different weight
36
37 ratios. Analysis was carried out at 40 kV voltage, 15 mA current and Ni-filtered Cu-K radiation
38
39 with 20 mm monochromator slit . The radiation scattered on the samples was measured with a
40
41
42 vertical goniometer. X-ray diffraction patterns were obtained at the 2θ values ranging from of
43
44 5° to 50° at the scan speed of 5°/min with a step size of 0.02° [26].
45
46
47
48 2.4 Determination of melting point and glass transition temperature
49
50
51 The thermal behavior of physical mixtures of RLZ with PAA, PVP VA, HPMC AS and ASDs
52
53
54
of RLZ prepared with these polymers were analyzed through DSC. The melting point of the
55
56 pure crystalline RLZ was determined by observing the presence of endothermic peak when the
57
58 ramp rate was set at 10oC/min until 140 oC.
59
60
61
62 9
63
64
65
The physical mixtures in different composition (90% w/w, 70% w/w, 50% w/w of drug) were
1
2 subjected to heat-cool-heat cycle in the DSC. Samples were heated till 140 oC, at a rate of 10
3
4
o
5 C/min followed by cooling at a rate of 20 oC/min to 0 oC. From 0 oC onwards, the sample was
6
7 reheated until 140 oC at the rate of 20 oC/min. The Tg value from the second heating run was
8
9
10 recorded. For the determination of the Tg of RLZ, it was heated till its melting point and kept
11
12 for 1 min at the same temperature to ensure complete melting. To the melted RLZ, liquid
13
14
15 nitrogen was poured over it to convert it into amorphous form. Its Tg was determined at a
16
17 heating rate of 20 oC/min till 140 oC in DSC.
18
19
20
21 2.5 FT-IR characterization
22
23
24 FT-IR spectra was obtained in the region of 400-4000 cm-1 in attenuated reflectance mode
25
26 (Bruker, Massachusetts, U.S.A) assisted with ZnSe crystal sample holder. Each sample (pure
27
28
29 RLZ, physical mixture of drug and polymer and ASDs of RLZ with each of the polymer in
30
31 50:50 weight ratio) was placed on the sample holder and scanned between the given range with
32
33
34 an average of sixty-four scans and a resolution of 2 cm-1 [27], [28].
35
36
37
2.6 Solid-state NMR
38
39
40 To study the molecular and structural insight to the drug-polymer interactions, ssNMR
41
42
43 experiments were recorded at 298 K using Bruker 750 MHz Avance III spectrometer connected
44
45 with a variable temperature unit. Topspin 3.6.2 pl6 and Topspin 4.1.3 Bruker NMR software
46
47
48 were used for data acquisition and processing. ssNMR experiments were carried out on a 3.2
49
50 mm HXY triple resonance probe, and for better sensitivity, it was configured to double
51
52
53 resonance mode. 1D 13C cross polarization magic angle spinning (CP-MAS) experiment was
54
55 recorded with an optimized cross-polarization (CP) contact time of 3 ms at a MAS frequency
56
57
of 12 kHz. Linear gradient pulse of ramp 100-70 was used during cross-polarization. 1H 90o
58
59
60 pulse of 3.3 mS corresponding to a RF of power 76 kHz and 13C power was kept ~ 56 kHz.
61
62 10
63
64
65
Proton RF power of 76 kHz was used during acquisition of the decoupling. A total of 1024
1
2 scans were acquired for each spectrum with a spectral width of 56818 Hz, corresponding to an
3
4
5 acquisition time of 18 ms. Data processing was done using the window function (exponential
6
7 multiplication) with a line broadening of 50 Hz. The interscan delay was 5 s for all experiments.
8
9
10
11 2.7 Crystallization tendency of amorphous solid dispersion
12
13
14 Thermal analysis of RLZ, polymers and RLZ ASDs prepared with different drug polymer
15
16
17
compositions (90:10, 70:30, 50:50, 30:70 and 10:90) were subjected to heat-cool-heat cycle.
18
19 The temperature range for the analysis was -20-140 oC. The starting point of temperature was
20
21 at 25 oC. The heating rate was 10 oC/min [29].
22
23
24
25 2.8 Microscopic analysis
26
27
28 Scanning electron microscopy (SEM) and transmission electron microscopy (TEM) studies
29
30
31 were conducted for determining the morphology of drug, ASDs and the state of ASDs after its
32
33 exposure to 90 oC for 12 hours. The microscopic analysis was done to detect the possible phase
34
35
36
separation on exposure to a higher temperature. The study of phase separation due to moisture
37
38 exposure was avoided pertaining to its unknown effect on solubilities [30]. The study was done
39
40 to access the thermodynamic stability of ASDs of RLZ with each of the three polymers. For
41
42
43 this analysis, thin ASD films of RLZ with the respective polymers in a 1:1 mixture of DCM
44
45 and methanol was prepared. 100 µl of solution was spin coated (EZ spinSD, Apex Instruments,
46
47
48 India) on a 1cm2 glass substrate and 200 mesh copper grids to be analyzed under SEM
49
50 ((MA15/18, CARL ZEISS MICROSCOPY LTD) and Tecnai G2T20 LaB6 TEM operating at
51
52
53 200 kV respectively.
54
55
56
57
58
59
60
61
62 11
63
64
65
1
2
3 The rate of release of RLZ from ASD (50:50 RLZ-polymer ratio) versus pure crystalline RLZ
4
5 was assessed employing USP type-II apparatus (Electrolab India Pvt Ltd.). 1000 mL of
6
7
8 phosphate buffer (pH 6.8) was used for the dissolution experiments, which were carried out at
9
10 37.0± 0.5 °C with a paddle rotating at a speed of 50 rpm [31]. 20 mg of RLZ and ASD
11
12
13 equivalent to 20 mg of RLZ were filled inside hard gelatin capsules and added in each vessel.
14
15 5 mL of aliquot was taken at each time interval (5, 15, 30, 45, 60, 90, and 120 min). The volume
16
17
18
was kept constant by adding equal volume of fresh media. HPLC method as described above
19
20 was used to determine the concentration of RLZ. The study was done in triplicate and mean
21
22 values are reported.
23
24
25
26 2.10 Drug polymer miscibility study
27
28
29
30 2.10.1 Melting Point Depression Study
31
32 For melting point depression study, 3-5 mg of samples (100% w/w 95% w/w, 90% w/w, 85%
33
34
35
w/w and 80% w/w of drug with each polymer) were taken in hermitically sealed aluminum
36
37 pans and heated under DSC from 20 oC to 140 oC at a rate of 2 oC/min. 50 ml/min nitrogen
38
39 flow was maintained throughout the DSC run. The readings have been taken in triplicate and
40
41
42 mean of the onset of melting point have been reported.
43
44
45 2.10.2 True density
46
47
48
The density of the drug and the polymers were determined using a helium pyrometer. Prior to
49
50 the measurements, the samples were kept over P2O5 for 24 hours. The readings were taken in
51
52 triplicate [32].
53
54
55
56
57
58
59
60
61
62 12
63
64
65
3 Results and Discussion
1
2
3
4 3.1 In silico studies
5
6
7 3.1.1 Preparation of 3D structures of polymers
8
9
10 The selection of the most appropriate polymer for preparing an ASD depends upon the
11
12 interactions between drug and polymer. In order to assess the drug-polymer interactions,
13
14
accurate or close to accurate 3-dimensional structures of both drug and the polymer are
15
16
17 required. Therefore, 3D structures of the three polymers PAA, PVP VA and HPMC AS were
18
19 modelled initially using computational tools. Modelled polymer chains were subjected to
20
21
22 molecular dynamic simulations to obtain the stable conformation of the chains. PAA polymer
23
24 chain consisting of 20 monomers of acrylic acid acquires a linear geometry (Figure S2a, in
25
26
27 supporting information). Similarly, PVP VA also exists in linear form containing vinyl acetate
28
29 and vinyl pyrrolidone linked alternatively (Figure S2b). On the other hand, HPMC AS block
30
31
32
copolymer acquires a stable bowl shape conformation by forming various intramolecular
33
34 hydrogen bonds as shown in Figure S2c.
35
36 The choice of the 20 monomeric units in PAA and PVP VA, and 5 monomeric units in HPMC
37
38
39 AS is based on assumption that the optimized structures of the three polymers should possess
40
41 equivalent surface area and flexibility to interact with RLZ molecule. During molecular
42
43
44 dynamics simulation and quantum chemical energy minimization of drug-polymer complexes,
45
46 the uniformity in the surface area and volume of the complexes can minimize the possibility of
47
48
49
inconsistencies. DFT calculation are best performed at the small model system of large
50
51 polymers to save the calculation time and computational cost, which allow more detailed and
52
53 precise structural investigations [33], [34]. The molar ratio of polymer to drug in the three
54
55
56 ASDs in the experimental condition vary for 50%w/w drug load because of the variable
57
58 molecular weights of the three polymers. For PAA-RLZ ASD, the polymer to drug molar ration
59
60
61
62 13
63
64
65
is 1923:1, for PVP VA-RLZ ASD, it is 218:1 and for HPMC AS-RLZ ASD, it is 77:1. The
1
2 DFT calculations of the large polymers collapse because of high degree of freedom and many
3
4
5 local minima energy points. When the model system is extended to the actual length, the similar
6
7 interactions are expected with the other RLZ molecules with the long polymer chains in ASDs.
8
9
10 3.1.2 Drug-polymer interaction
11
12
13 To determine possible drug-polymer interactions, drug-polymer complexes are obtained in
14
15
16 three steps. Firstly, RLZ was docked on the simulated geometries of PAA, PVP VA and HPMC
17
18 AS at multiple sites around the polymeric chains so as to obtain the initial RLZ-polymer
19
20
21
complexes for further quantification of interaction energies. At the second step, the docked
22
23 complexes were subjected to molecular dynamics simulation to obtain the best fit of drug
24
25 within the flexible polymer cavity. In the last step, DFT calculations were performed on the
26
27
28 simulated complexes to calculate the precise interaction energies between drug and polymer.
29
30 In the docking studies, the most stable docked poses resulted the complexes with docking score
31
32
33 -4.15, -3.83 and -4.14 kcal/mol for PAA, PVP VA and HPMC AS, respectively. Based on the
34
35 initial assessment through molecular docking the stability of RLZ-polymer complexes was
36
37
38 found to be in the order RLZ-PAA > RLZ-HPMC AS > RLZ-PVP VA. In the three complexes,
39
40 the free NH2 group of RLZ shows hydrogens bonding interactions with the oxygen of
41
42
carboxylic acid or ester groups of the polymers. RLZ resides at the surface of the polymers,
43
44
45 however the additional hydrogen bonding interaction is visible between ring nitrogen of RLZ
46
47 and carboxylic acid group of HPMC AS. The C–H···F and van der Waals interactions in RLZ-
48
49
50 PAA lead to increase the docking score higher than that of RLZ-HPMC AS (Figure S3).
51
52 However, the docking scores usually do not take the electron density based interaction into
53
54
55 account. Thus, to further validate the order of the stability and quantify the strength of
56
57 interaction in terms of energy values more robust computational algorithm that is DFT
58
59
60
calculations were performed. RLZ, polymers and RLZ-polymer complexes were optimized
61
62 14
63
64
65
using ONIOM method, wherein RLZ was optimized using density functionals and polymer
1
2 segment was optimized using semi-empirical method. The drug-polymer complex formation is
3
4
5 an endergonic process for which the interaction energies were calculated in gas phase and in
6
7 MeOH solvent. The interactions energies for RLZ-PAA, RLZ-PVP VA and RLZ-HPMC AS
8
9
10 in gas phase were calculated to be 2.29, 2.64 and 15.24 kcal/mol, respectively. The interaction
11
12 energies in MeOH solvent were found to be 10.75, 11.37 and 25.06 kcal/mol for RLZ-PAA,
13
14
15 RLZ-PVP VA and RLZ-HPMC AS, respectively. Based on the interaction (complexation)
16
17 energy values, the order of the stability was RLZ-PAA > RLZ-PVP VA > RLZ-HPMC AS
18
19
under both gas phase and solvent conditions. The normalized interactions energies were
20
21
22 calculated with respect to RLZ-PAA as shown in Figure 2. Since, ASDs do not contain the
23
24 solvent molecule, the gas phase energy value can be correlated with the experimental results.
25
26
27 The optimized geometries of the RLZ-PAA complex indicates that RLZ is fitting well inside
28
29 the polymer cage. The free PAA was transformed from linear conformation to a folded
30
31
32 conformation in the RLZ-PAA complex. There are four types of prominent molecular level
33
34 interaction visible in the RLZ-PAA, (i) a strong hydrogen bond between ring nitrogen of RLZ
35
36
37
and COOH group of PAA, (ii) a strong hydrogen bond between NH2 group of RLZ and COO–
38
39 group of PAA, (iii) C–H···O interaction between RLZ and COOH, and (iv) C–H··· interaction
40
41
42 between CH2 linker of polymer and  cloud of thiazole ring of RLZ (Figure 2a). In case of
43
44 RLZ-PVP VA, the most prominent interactions are hydrogen bonding interactions of the two
45
46
47 hydrogens of NH2 of RLZ with COO– and pyrrolidone carbonyl oxygen of PVP VA (Figure
48
49 2b). The strong hydrogen bonding interaction of NH2, and ring nitrogen of RLZ with carboxylic
50
51
groups of HPMC AS are visible in RLZ-HPMC AS complex. An additional C–H···O
52
53
54 interaction is also visible; (Figure 2c). It was observed that major conformational change
55
56 occurred in PAA as a result of drug-polymer complexation whereas PVP VA does not allow
57
58
59 much conformational change due to its rigid structure. The free HPMC AS polymer was
60
61
62 15
63
64
65
already stabilized by forming multiple intramolecular hydrogen bonds and thus does not
1
2 undergo further conformational changes after complexation with RLZ. This molecular level
3
4
5 observation indicates that PAA may be the suitable polymer for forming a stable drug-polymer
6
7 non-covalently bonded complex, when a slight energy is provided. PVP VA may be the next
8
9
10 better polymer of choice. In order to compare the propensity of RLZ to undergo crystallization,
11
12 an attempt was made to calculate the non-covalent bonding interaction energies between two
13
14
15 RLZ molecules through DFT calculation. The interaction energy for RLZ-RLZ was calculated
16
17 to be -10.18 kcal/mol, which reflect very strong interactions due to bidirectional dual hydrogen-
18
19
bonding as shown in figure S4. However, the crystallization behaviour of drug in 1:1 ratio of
20
21
22 drug-polymer and drug-drug cannot be compared through DFT calculation, because it provides
23
24 the quantification of intermolecular interactions at smaller model systems. The interaction
25
26
27 energy of -10.18 kcal/mol justifies the requirement of heating (enthalpy) to break these drug-
28
29 drug intermolecular interactions within the crystal lattice of drug.
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62 16
63
64
65
Figure 2. RLZ-polymer complexes along with their respective interaction energies. Bond distances are
1 in Å, interaction energies are in kcal/mol. Energy values are the complexation energy values
2 referenced to PAA-RLZ in italics and in brackets are obtained under implicit solvent condition using
3 MeOH as the solvent.
4
5 Docking and MD simulation studies depend upon molecular mechanics wavefunction which is
6
7
8 suitable for computational studies of large systems like polymer, lipids and nanocarriers etc.
9
10 These studies can provide the qualitative determination of molecular behavioral patterns in the
11
12
13 simulated conditions. On the other hand, DFT studies depend upon ab initio or semi empirical
14
15 wavefunctions which are more robust and provide more accurate estimation of energy values.
16
17
18
However, DFT studies can be performed on truncated model systems within the agreeable time
19
20 frame and available computational facilities. DFT studies are more suitable when quantitative
21
22 comparison and more precise atomic details of molecular interactions are required such as bond
23
24
25 distance and bond angle for the interactions. On one hand, simulation of larger system
26
27 corresponds to the more realistic systems with diverse chemical behavior, while on the other
28
29
30 hand, the smaller systems lead to quick, more accurate calculations and atomic details but limits
31
32 the study due to a gap between simulated and realistic systems. The selection of computational
33
34
35 methods depends upon the formulation scientist. For example, MD simulation techniques are
36
37 more suitable when one wants to calculate the physicochemical properties like miscibility,
38
39 solubility or glass transition temperature of ASDs. DFT calculations are preferred choice when
40
41
42 estimation of molecular level interactions, energy calculations or polymer profiling is desired.
43
44
45
46 3.2 Powder X-ray diffraction study
47
48
49 The PXRD pattern of pure crystalline RLZ shows distinct sharp peaks at 9.14o, 13.64o, 18.18o,
50
51 22.74o, and 25.22o, which indicates the crystalline nature of the drug (Figure 3). The PXRD
52
53
54 patterns of ASDs with different drug loading (50, 70, and 90 %w/w) with PAA, PVP VA, and
55
56 HPMC AS show the transformation of the sharp peaks to the absolute halo patterns at 50%
57
58 w/w drug loading, which indicate the presence of single-phase binary amorphous system. At
59
60
61
62 17
63
64
65
70% w/w, a few indistinct sharp peaks can be observed at 19.68o, and 25.3o in the case of ASD
1
2 RLZ-HPMC AS. In ASD RLZ-PVP VA (70:30) also minor sharpness at 19.68o was observed.
3
4
5 After increasing the drug loading to 90% w/w, crystallinity was observed with all three
6
7 polymers in the order PAA<PVP VA<HPMC AS. Shift in the two theta values RLZ from
8
9
10 18.18o to ~19.4o and 25.22o to 26.6o could be attributed to the occurrence of polycrystallinity
11
12 in RLZ after association with the polymer [35]. This is an indication of the presence of two
13
14
15 phases in the analyzed ASD. This concludes that the ASD formation capacity of polymers with
16
17 RLZ is in the order of PAA>PVP VA>HPMC AS.
18
19
20
21
22 a) b) c)
23
24
25
26
27
28
29
30
31
32
33
34 Figure 3. Overlay spectra of PXRD patterns of ASDs having (a) 50:50 of RLZ with polymers, b)
35 70:30 of RLZ with polymer, and c) 90:10 of RLZ with polymer.
36
37
38 3.3 Determination of melting point and glass transition temperature
39
40
41 Tg is the indicator of molecular mobility, where lower molecular mobility refers to high
42
43 physical stability of the amorphous systems. DSC is well known for its evaluation of transition
44
45
46 process occurring in amorphous solid at a higher temperature at which the surrounding
47
48 molecules influences the orientation and reorientation of the molecules [36].
49
50
51 Tm and Tg of RLZ, physical mixtures of RLZ with each polymer, and ASDs are shown in Figure
52
53
54 4. A melting endotherm was shown by RLZ at 120 oC demonstrating its pure crystalline nature,
55
56 while amorphous RLZ showed a Tg at 28 oC (Figure S5). When 90:10 ratio of drug-polymer
57
58
59 were subjected to heat-cool-heat cycle in the DSC for in situ ASD preparation, and the
60
61
62 18
63
64
65
determination of their Tg, sharp endotherm was observed with all the three polymers. This
1
2 confirmed the absence of in situ ASD formation at 10% w/w ratio. At 70:30 ratio of the drug-
3
4
5 polymer, the Tg of RLZ-PAA and RLZ-PVP VA were observed at 91 oC and 39 oC,
6
7 respectively. However, RLZ-HPMC AS did not show any change in the specific heat capacity
8
9
10 and, a melting endotherm was observed confirming that HPMC AS was not efficient to form
11
12 ASD with RLZ at this weight ratio as well. At 50:50 ratio of drug-polymer, with all three
13
14
15 polymers glass transition temperatures were observed. A Tg of 106 oC, 53 oC, and 40 oC were
16
17 observed for RLZ-PAA, RLZ-PVP VA, and RLZ-HPMC AS, respectively. In the
18
19
aforementioned cases, a single Tg was observed whenever in situ ASD formation occurred
20
21
22 inside the DSC. This clearly indicates the formation of a miscible system since the obtained Tg
23
24 of in situ formed ASDs were between the Tg of pure amorphous drug and polymer (Table 2)
25
26
27 [37]. This study implies that PAA is imparting most promising stability to the amorphous RLZ
28
29 followed by PVP VA and HPMC AS.
30
31
32
Table 2. Tg of different solid forms of RLZ and ASD with different drug polymer compositions.
33
Drug to polymer
34 S.No Composition Tg (oC)
35 weight ratio
36 1 RLZ-PAA 106
37 2 50:50 RLZ-PVP VA 53
38 3 RLZ-HPMC AS 40
39 4 RLZ-PAA 91
40
41
5 70:30 RLZ-PVP VA 39
42 6 RLZ-HPMC AS -
43 7 RLZ-PAA -
44 8 90:10 RLZ-PVP VA -
45 9 RLZ-HPMC AS -
46 10 a
NA PAA 102
47
11b NA PVP VA 101
48
49 12c NA HPMC AS 117
50 13 NA RLZ (Amorphous) 28
a b c
51 Ref [38], Ref [30], Ref [39]
52
53
54
55
56
57
58
59
60
61
62 19
63
64
65
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26 Figure 4. Overlay of DSC plots of in situ ASDs of RLZ with each of the polymer in different drug to
27 polymer weight ratios formed inside the DSC with heat-cool-heat cycle
28
29
30 3.4 FT-IR characterization
31
32
33 Figure 5 represents the comparative FTIR spectra of RLZ, PMs, and ASDs of RLZ-polymer at
34
35 a 50:50 weight ratio. In the spectra of RLZ the prominent peaks at 3364 cm-1, 1543 cm-1, 1458
36
37
38 cm-1,1293 cm-1 and 1143 cm-1 correspond to N-H stretching, C=C stretching, C-F stretching,
39
40 C-O stretching, and C-N stretching, respectively. Comparing the spectra of RLZ with the
41
42
physical mixture of each polymer, minor differences like the shift from 3364 cm-1 to 3368 cm-
43
44
1
45 , 1538 cm-1 to 1543 cm-1, 1293 cm-1 to 1300 cm-1 and 1143 cm-1 to 1203 cm-1 were observed
46
47 in PM RLZ-PAA. In PM RLZ-PVP VA, peaks were observed at 3366 cm-1 and 1541 cm-1. In
48
49
50 PM RLZ-HPMC AS, the characteristic peaks were observed at 3323 cm-1, 1248 cm-1, and 1213
51
52 cm-1. However, in the case of ASDs, much more changes in the wavenumber were observed.
53
54
55 The N-H stretching peaks were broadened significantly in the ASDs. The C=C stretching peaks
56
57 were shifted to 1528 cm-1 (in ASD RLZ-PAA) and 1538 cm-1 (in ASD RLZ-PVP VA and ASD
58
59
60
RLZ-HPMC AS). The C-O stretching, C-N stretching and C-F stretching peaks were also
61
62 20
63
64
65
shifted in the ASD samples. These shifts in wavenumber confirmed the formation of ASD and
1
2 presence of interactions between drug and polymer. The evidence of strong drug-polymer
3
4
5 interactions assures the formation of a highly stable ASD by increasing the onset temperature
6
7 of crystallization and by attenuating the crystallization of the drug [40].
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36 Figure 5. FTIR spectra of RLZ, physical mixture of RLZ with polymers and ASD of RLZ with each
37 of the respective polymer
38
39
40 3.5 Solid-state NMR
41
42
43 ssNMR is a very powerful tool to identify the presence of molecular interactions between the
44
13
45 drug and the polymer, responsible for the stability of ASD. The C CPMAS spectra of
46
47
crystalline RLZ, pure polymers, and RLZ ASD with each of the polymers in the ratio of 70:30
48
49
50 were recorded to study the interactions occurring at the molecular level. The typical sharp
51
52 resonances in the spectra of pure RLZ are reflecting the crystalline nature of RLZ. In the spectra
53
54
55 of ASDs, the peaks of both drug and polymer are merged hence confirming the formation of
56
57 molecular level dispersion. The broad Gaussian resonances are indicative of disorder in the
58
59
60 crystal arrangement as seen in amorphous solid [41]. ASD RLZ-PAA is showing more
61
62 21
63
64
65
nonhomogeneous line broadening with respect to ASDs RLZ-PVP VA and RLZ-HPMC AS,
1
2 which could be possible due to more random distribution of the chemical environment
3
4
5 subjected to more loss of crystallinity [42] in comparison to the former and the presence of a
6
7 small amount of crystalline drug in the later two spectra. This is also in close agreement with
8
9
10 the results obtained in PXRD studies and in silico studies.
11
12
13 The overlay of pure RLZ and ASD RLZ-PAA 13C CP/MAS spectra shows the downfield shift
14
15
16
in the quaternary carbon (Ca) of the thiazole ring from 169 to 170 ppm, as shown in Figure 6a.
17
18 Another quaternary carbons (Ce and Cc) of pure RLZ occurring at 150 and 143 ppm,
19
20 respectively, are getting merged with the Cb in the case of ASD RLZ-PAA. A similar
21
22
23 observation is found for another aromatic carbon Ch and Cg, which are occurring at 132 and
24
25 129 ppm, respectively. The trifluoromethyl carbon (Cf) shown at 120 ppm is also merged with
26
27
28 the aromatic Cd carbon of RLZ. The downfield shift from 178 ppm to 180 ppm in the carbonyl
29
30 carbon (Ca’) and from 37 ppm to 43 ppm in the methylene carbon (Cb’) of PAA [43] indicate
31
32
33
that these carbons are experiencing the electron-withdrawing effect through non-covalent
34
35 interactions. The merging of peaks and changes in the 13C chemical shifts explain the rigidity
36
37 in the structure due to the prominent interaction between RLZ and PAA functional groups.
38
39
40
41 In the case of PVP VA (Figure 6b), the resonance at 175 and 168 ppm are assigned to the
42
43 carbonyl carbons of the pyrrolidone (Ca’) and acetate (Ci’) functionality, respectively.
44
45
46 Resonances of 65 and 34, 18 and 16 ppm are assigned to Cg’, Cf’, Cc’ and Cj’ carbon
47
48 respectively. However, the carbon present in the pyrrolidone ring (Cb’) and the vinyl chain (Ch’)
49
50
have been assigned resonance values of 40 and 29, respectively [44]. Despite the ASD RLZ-
51
52
53 PVP VA showing evidence of the formation of the molecular dispersion, there is clearly the
54
55 presence of crystalline RLZ which is evident by the sharp resonances. The interaction could be
56
57
58 seen in the shortening and the slight deshielding of the carbons present in the range of 40-15
59
60 ppm of PVP VA. Hence, it can be inferred that despite of interaction shown by the polymer the
61
62 22
63
64
65
crystalline structure of RLZ persists making it a two-phased dispersion. In Figure 6c, the
1
2 anomeric carbons labeled as a', d', b', c', and e', j' are responsible for resonances lying between
3
4
5 101-58 ppm for polymer HPMC AS. At 67 and 56 ppm there are shoulders due to methylene
6
7 Cf', and methoxy carbons (Ci' and Ch'), respectively. The most shielded carbons lying in the
8
9
10 range of 26-14 ppm are those in the acetyl (Ch’), succinoyl chain (Cn’ and Co’) and methyl
11
12 carbon (Cl’) in the hydroxypropyl chain. Interactions between RLZ and polymer in ASD RLZ-
13
14
15 HPMC AS can be seen through the deshielding of carbons lying between 100-58 ppm of HPMC
16
17 AS, but the resonances of RLZ seem unaffected in terms of chemical shifts; however, there is
18
19
20
a change in the peak intensities from that of the crystalline RLZ.
21
22
23 Hence, it can be concluded that the interaction and the capacity to make a stable amorphous
24
25 solid dispersion system by the polymer is in the order of PAA > PVP VA > HPMC AS. These
26
27
28 results corroborate the findings from the in silico work.
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62 23
63
64
65
a)
1
2
3
4
5
6
7 Cb’
8
9 Ca’
10
11
12
Cb
13 Cf
14 Ca Cd
Ce Ch Cg
Cc
15
16
17
18
19
20 b)
21
22
23
24
25
26
27 Cb’,h’
Cd’,e’
28 Cc’
Cj’
29 Ca’
Cf’
Ci’ Cg’
30
31
32
33
34
35
36
37
38
39
40
c)
41
42
43
44
45
46 Cb’,c’
Ce’,k’
47
48 Cd’ Cf’,j’ Ch’
49 Cg’ Ca’ Ci’
Cm’,p’ Cn’,o’
50 Cl’
51
52
53
54
55
56
57
58
59
60
61
62 24
63
64
65
Figure 6. The 13C CP-MAS spectra of RLZ in comparison with a) PAA and ASD RLZ-PAA, b) PVP
1 VA and ASD RLZ-PVP VA, and c) HPMC AS and RLZ-HPMC AS
2
3
4 3.6 Crystallization tendency of amorphous solid dispersion
5
6
7 Heat-cool-heat cycle in DSC was used to assess the crystallization tendency of drug in the
8
9 prepared ASDs. The thermograms of RLZ ASDs prepared with different composition of drug
10
11
and polymer are shown in Figure S6 (a to f). Fusion and crystallization enthalpies of RLZ and
12
13
14 RLZ ASDs are mentioned in table 3. In the thermograms it can be seen that the samples with
15
16 higher weight percentage of drug (90% w/w) the cooling crystallization event is prominent in
17
18
19 ASDs prepared with all the polymers. The enthalpy of fusion values are in close agreement
20
21 with the enthalpy of crystallization values showing the extent of crystallization occurring in
22
23
24 90% w/w drug loaded ASDs. Exothermic effect due to the crystallization of RLZ were not
25
26 observed in the ASDs prepared with polymer wight percentage of 30% to 90%. A slight event
27
28
29
of crystallization was observed in RLZ ASDs prepared HPMC AS at 70% w/w drug loading.
30
31 The crystallinity estimated from the fusion of enthalpy were around 15%, 16% and 29% for
32
33 90:10 RLZ-PAA, RLZ-PVP VA, and RLZ-HPMC AS ASDs respectively. This gives an
34
35
36 indication that solid dispersions of RLZ when prepared with PAA and PVP VA will offer
37
38 highest resistance against drug recrystallization followed by ASD prepared with HPMC AS.
39
40
41 The higher value of crystallinity describes the higher mobility of HPMC AS [29]. This confirms
42
43 the higher kinetic stability of ASDs prepared with PAA and PVP VA with respect to HPMC
44
45
46 AS.
47
48
49 Table 3: Fusion enthalpies (ΔHm) and crystallization enthalpies (ΔHcr), with corresponding onset
50 temperatures of RLZ and its ASDs with polymers in different compositions.
51 Enthalpy ΔHcr (Jg-1) ΔHm (Jg-1)
52 RLZ -332.94 (87.25 oC) 341.82 (119.36 oC)
53
54 ASDs 90:10 70:30 50:50 30:70 10:90 90:10 70:30 50:50 30:70 10:90
55 RLZ: PAA -49.94 - - - - 52.87 - - - -
56 RLZ: PVP
57 -51.65 - - - - 55.57 - - - -
VA
58 RLZ:
59 -95.18 -23.88 - - - 101.73 - - - -
HPMC AS
60
61
62 25
63
64
65
1
2
3 The TEM and SEM images of ASDs have been taken in pre- and post-annealed conditions.
4
5 TEM with the selected area electron diffraction (SAED) patterns was used to identify the
6
7
8 regions where crystallinity might have developed due to annealing. Figure 7 shows the bright
9
10 field images of the ASDs for the pre- and post-annealed conditions. A large sample size (50
11
12
13 areas on each grid) was chosen to study the morphology of RLZ ASDs. The bright field TEM
14
15 images (Figure 7 (a, e and i)) show the spherical morphology of ASDs prepared with the
16
17
18
polymers. The diffused rings produced by the particles in the SAED pattern confirm their
19
20 amorphous nature [45]. The TEM images of all three ASDs were observed for the post-
21
22 annealed condition and found that no significant change has occurred in RLZ-PAA ASD
23
24
25 (Figure 7c). However, in the RLZ-PVP VA and RLZ-HPMC AS ASD, the bright field images
26
27 along with SAED showed a high degree of crystallization as there was the presence of a regular
28
29
30 array of diffraction spots (Figure 7 (g,h) and (k,l)) [46], [47]. The SAED pattern consists of
31
32 systematic spots which means the grains or crystallites formed are in the ranges of nm length
33
34
35 scale. Along with the array of diffraction, diffused circular rings can also be seen in the pattern
36
37 explaining the presence of both amorphous and crystalline regions. To confirm the presence of
38
39
40
RLZ only, the crystallographic information was collected from the CIF file [48] and it showed
41
42 that the compound belongs to the triclinic system with a space group of P1̅ suggesting that the
43
44 crystal has got center of inversion (centrosymmetric). As per the crystallography data available
45
46
47 on the Cambridge Crystallographic Data Centre (CCDC), the lattice parameters of RLZ are:
48
49 a=8.08 Å, b= 11.78 Å, c=19.74 Å and α = 78.44o, β= 84.378o, γ= 89.318o. The information
50
51
52 obtained from the zone axis table of Figure 7 h and l indicated that the crystals present have
53
54 the zone axis of [311] and [-112] which corresponds to the RLZ crystals.
55
56
57 The presence of RLZ crystals due to demixing after annealing from the supersaturated ASD is
58
59 in agreement with the Tg results. It appears that due to high Tg, ASD RLZ-PAA exhibited
60
61
62 26
63
64
65
superior stability in comparison to RLZ-PVP VA and RLZ-HPMC AS ASD and retained its
1
2 amorphous form at higher temperatures. This confirms the suitability of PAA as polymer over
3
4
5 PVP VA and HPMC AS for the preparation of a thermodynamically stable ASD with RLZ.
6
7 The chemical stability of all the three ASDs of RLZ with PAA, PVP VA and HPMC AS, which
8
9
10 were estimated through HPLC, was in line with the thermodynamic and kinetic stability. The
11
12 retention time (Rt ~ 3.9 min) of RLZ in the ASDs was same as that in the pure RLZ sample,
13
14
15 which confirmed the unchanged chemical environment around pure RLZ molecules (Figure
16
17 S1). This may also confirm that after demixing the recrystallized component was RLZ only.
18
19
SEM images of ASDs in pre- and post-annealed conditions have been shown in Figure S7a-f.
20
21
22
23
24
a) b) c) d)
25
26
27
28
29
30
31
32 e) f) g) h)
33
34
35
36
37
38
39 i) j) k) l)
40
41
42
43
44
45
46
47 Figure 7. Bright-field images of (a) freshly prepared RLZ-PAA ASD, (c) annealed at 90oC for 12 hrs
48 RLZ-PAA ASD, (e) freshly prepared RLZ-PVP VA ASD, (g) annealed at 90oC for 12 hrs RLZ-PVP
49 VA ASD, (i) freshly prepared RLZ-HPMC AS ASD, (k) annealed at 90oC for 12 hrs RLZ-HPMC AS
50
51
ASD, and its corresponding diffraction patterns (b), (d), (f), (h), (j) and (l) respectively.
52
53
54
55
56
57
58
59
60
61
62 27
63
64
65
1
2
3
4 In this study 50:50 RLZ:polymer ASD were taken considering the fact that at this ratio all the
5
6 ASDs showed complete amorphization as shown in PXRD study (section 3.2). RLZ is a weakly
7
8
9 basic drug with an aqueous solubility of 0.3 mg/ml at neutral pH and 12 mg/ml (approx.) at
10
11 acidic pH of 1.2 [49]. It has a pKa of 3.47, and it shows higher solubility at lower pH and lower
12
13
14 solubility at neutral to high pH [50]. Out of the three polymers, HPMC AS shows solubility at
15
16 pH above 5 [51], hence selection of acidic medium for dissolution would not have provided
17
18 the comparative results, hence pH 6.8 medium was selected for dissolution study. The
19
20
21 dissolution profile of all three ASDs versus crystalline RLZ is shown is Figure S8. RLZ:PAA
22
23 ASD showed maximum release of drug (100%) in the initial 45 min of study followed by
24
25
26 RLZ:PVP VA (91%) and RLZ:HPMC AS (87%). Pure crystalline RLZ is having a low
27
28 dissolution profile as compared to the ASD formulation, however it was also showing around
29
30
31 83% release in initial 45 min. Almost every solid state of drug showed 100% drug release in
32
33 120 mins. The difference between the dissolution profile of each ASD formulation is small,
34
35
36
which can attribute to the higher sink condition that is maintained during the dissolution study.
37
38 Since ASDs are supersaturated drug delivery systems, they tend to increase the dissolution of
39
40 any poorly aqueous soluble drug. However, this may lead to drug precipitation due to
41
42
43 supersaturation; hence, higher sink condition was maintained [52].
44
45
46
47 3.9 Drug polymer miscibility and phase diagram
48
49
50 The determination of miscibility of drug in polymer is widely estimated by Flory-Huggins
51
52 theory. This theory explains the thermodynamics of the mixing of polymers with solvent
53
54
55
molecules by using the Flory-Huggins interaction parameter χ [53]. The degree and type of
56
57 interaction play a major role in free energy of mixing (ΔGmix), which can be determined by
58
59 equation 1.
60
61
62 28
63
64
65
Δ𝐺𝑚𝑖𝑥 Φ𝑝𝑜𝑙𝑦
= Φ𝑑𝑟𝑢𝑔 𝑙𝑛Φ𝑑𝑟𝑢𝑔 + 𝑙𝑛Φ𝑝𝑜𝑙𝑦 + χΦ𝑑𝑟𝑢𝑔 Φ𝑝𝑜𝑙𝑦 …Eq 1
1 𝑅𝑇 𝑚
2
3
4 Where Φ is the volume fraction of the drug and the polymer; R is the universal gas constant; T
5
6 is the absolute temperature of the system; m is the ratio of the volume of a polymer chain to
7
8
9 drug molecular volume [54], [55]. Equation 1 shows the contribution of the entropy and
10
11 enthalpy of mixing of drug and polymer, which are represented by the first and second terms,
12
13
14 respectively. Entropy is the decisive factor in the determination of spontaneous mixing since
15
16 enthalpy always favors the mixing. It is so because the enthalpy of mixing cannot surpass a
17
18
19 small critical value unless any phase separation occurs.
20
21
22 To determine the entropy of mixing the two components, interaction parameter (χ) (shown in
23
24 equation 2) was calculated by the melting point depression method, which is an indicator of
25
26
27 the decreased chemical potential of drug in the mixture with respect to the pure crystalline drug
28
29 [56].
30
31
32 1 1 𝑅 1 2
33 − 𝑜 = − Δ𝐻 [ln Φ𝑑𝑟𝑢𝑔 + (1 − 𝑚) Φ𝑝𝑜𝑙𝑦 + χΦ𝑝𝑜𝑙𝑦 ] …Eq 2.
𝑇𝑚 𝑇𝑚 𝑓
34
35
36
37 Where Tm is the drug’s melting temperature in the mixture of drug and polymer, 𝑇𝑚𝑜 is the
38
39 melting point of the pure crystalline drug, Φ is the volume fraction of the drug and polymer
40
41
42 and ΔHf is the heat of fusion of the pure crystalline drug. The DSC plots shown in Figure S9
43
44 indicate the depression in the onset of the melting point of RLZ with varying weight fractions
45
46
47 (Φpoly) of each polymer. Figure S9 shows that increase in weight fraction of each polymer is
48
49 proportionately related to the depression in the onset of melting point of the crystalline RLZ.
50
51
52
This verifies that there is a certain degree of mixing between the drug and polymer near the
53
54 melting temperature of the drug in the respective compositions of physical mixtures. Using
55
56 equation 2 (values of the required parameters are mentioned in Table 4), χ of drug with each
57
58
59
60
61
62 29
63
64
65
of the polymers were calculated. The calculated χ were plotted against temperature empirically
1
2 described by equation 3 to show its dependence on temperature.
3
4
5
6 χ=A+ B/T …Eq 3.
7
8
9 Where A and B/T are the relative contributions of entropy (temperature-independent) and the
10
11
12 enthalpy, respectively towards the free energy of mixing. T is the melting temperature of drug
13
14 in a drug-polymer mixture. Equation 3 is a representation of straight-line equation, where the
15
16
17 value of R2 were 0.9477, 0.9817, and 0.9410, respectively, for physical mixtures of RLZ-PAA,
18
19 RLZ-PVP VA, and RLZ-HPMC AS showing the goodness of fit in χ vs 1/T plot (shown in
20
21
22
Figure S10). Hence, by using the value of A and B, the value of χ was extrapolated for even
23
24 lower temperatures.
25
26
27 Table 4: Physical properties of RLZ, PAA, PVP VA and HPMC AS used for quantitative drug-polymer
28 miscibility.
29
30 True Molar
Molecular ΔHf Tm
31 S.No. Components Density volume
weight (g/mol) (kJ/mol) (K)
32 (g/cm3) (cm3/mol)
33
34 1 RLZ 234 1.66 140.96 18043.74 391.98
35 2 PAA 450000 1.33 338345.86 - -
36
37 3 PVP VA 51000 1.13 45132.74 - -
38
39 4 HPMC AS 18000 1.23 14634.15 - -
40
41
42
43 In the case of drug-polymer miscibility, since the miscible diluent is a polymer that, which is
44
45 amorphous in nature, the entropy of mixing becomes negligible [57]. Hence, the second
46
47
derivative of the free energy (equation 1) is equated to zero and as stated in equation 4, the
48
49
50 maximum drug-polymer miscibility boundary may be calculated as:
51
52 2𝐵
53 𝑇𝑆 = …Eq 4.
1 1
( )+ ( −2𝐴)
54 Φ𝑑𝑟𝑢𝑔 𝑚(1−Φ𝑑𝑟𝑢𝑔 )
55
56
57
58 Drug solubility and miscibility can be plotted as a function of drug weight fraction and
59
60 temperature after the understanding of how ΔGmix varies as a function of T for each system.
61
62 30
63
64
65
ΔGmix is an indicator of a stable drug-polymer system and was calculated using χ through
1
2 equation 1 and then ΔGmix/RT was plotted against the drug volume fraction (Φdrug). The
3
4
5 negative value of ΔGmix in all the three systems was seen across all drug weight fraction near
6
7 the melting point of the drug. This shows the thermodynamic stability of all three drug-polymer
8
9
10 systems near drug’s melting point. ΔGmix/RT vs Φdrug were also plotted at Tg shown by each
11
12 drug-polymer system prepared by in-situ ASD formation inside DSC instrument (discussed in
13
14
15 section 3.3). In this curve shown in Figure 8 a, b, and c it was seen that all the drug-polymer
16
17 system at 106oC (Tg of 50:50 RLZ-PAA ASD), ΔGmix is negative and convex across all Φdrug.
18
19
However, at 53 oC (Tg of RLZ-PVP VA ASD) and at 40 oC (Tg of 50:50 RLZ-HPMC AS ASD),
20
21
22 all the systems showed instability due to positive ΔGmix at their respective Tg, which may result
23
24 into heterogeneous drug-polymer mixture. It is noteworthy that RLZ-PAA showed negative
25
26
27 ΔGmix at its Tg while RLZ-PVP VA and RLZ-HPMC AS showed positive ΔGmix. Contrary to
28
29 this result, in experimental settings, we observed single Tg of freshly prepared ASDs, which
30
31
32 correspond to the single phased drug-polymer system. Usually, immiscible-systems, two
33
34 different glass transition temperatures are observed, one for drug and another for polymer [58].
35
36
37
Miscibility study is mainly concerned with the thermodynamic stability of ASD but alone
38
39 cannot govern the absolute stability of the ASD because it does not account the kinetic stability
40
41 [47]. At room temperature as well, the positive ΔGmix was showing the instability of all the
42
43
44 compositions of drug-polymer. From equation 1 and 2, it can be seen that smaller the value of
45
46 χ more negative will be the ΔGmix, designating the drug-polymer complex to be more stable
47
48
49 and a greater number of interactions between drug and polymer [58]. Therefore, positive values
50
51 of χ obtained at 25 oC showed the formation of immiscible drug-polymer system (Table 5). The
52
53
54 negative value of χ is obtained at <94 oC, <80 oC and <66 oC for RLZ-PAA, RLZ-PVP VA and
55
56 RLZ-HPMC AS, respectively (Figure 8 d, e, and f). This highlights the temperature at which
57
58
the drug-polymer system may exist as a homogenous phase. The value of interaction parameter
59
60
61
62 31
63
64
65
was observed to be influenced by each factor encompassing Flory-Huggins model, thus could
1
2 not be used for comparison between different drug-polymer systems [59]. So, the χ values for
3
4
5 each drug-polymer system at 25 oC were not sufficient to conclude the grading of polymer in
6
7 terms of its efficacy in complete amorphousization to give stable ASDs. Hence, the system was
8
9
10 further evaluated by solubility and miscibility curves.
11
12
13 Table 5: Values of Constant A, Constant B and interaction parameter (Ӽ) for Three RLZ-Polymer
14 Systems.
15
16 Drug-Polymer Interaction Parameter (χ)
S.No. Constant A Constant B
17 System at 25oC At melting point
18
19 1 RLZ-PAA -210.73 77574 49.45 -12.83
20 2 RLZ-PVP VA -45.08 16009 8.61 -4.24
21
22 3 RLZ-HPMC AS -155.86 53012 21.94 -20.62
23
24
25
26 Henceforth, further insight to understand the potential of polymer towards formation of stable
27
28
29
ASD was taken by preparing the phase diagram of miscibility temperature and solubility
30
31 temperature (calculated using equation 4) vs drug volume fraction. This information was paired
32
33 with predicted Tg of the system with different weight ratios of drug and polymer, which were
34
35
36 calculated using Gordon-Taylor equation (equation 5).
37
38 𝑊1 𝑇𝑔1 + 𝐾𝐺 𝑊2 𝑇𝑔2
𝑇𝑔 = …Eq 5
39 𝑊1 + 𝐾𝐺 𝑊2
40
41
42 Where Tg, Tg1, and Tg2 are the glass transition temperatures of the drug-polymer mixture, drug
43
44
45 in amorphous form and the polymer, respectively. The constant value KG can be calculated
46
47 using equation 6.
48
49
50 𝜌1 𝑇𝑔1
51 𝐾𝐺 = …Eq 6
𝜌2 𝑇𝑔2
52
53
54
55 Where ρ1 and ρ2 are the true density value of amorphous drug and polymer, respectively. This
56
57 phase diagram concerns the range of temperature and composition, below and above which the
58
59
60 system exists in single-phase and two-phase thermodynamic region. Through phase diagram
61
62 32
63
64
65
(Figure 8 d, e and f), information about the maximum solubility and miscibility of amorphous
1
2 drug within the polymeric matrix and its dependence on temperature can be obtained. The
3
4
5 region lying at the right-hand side of the spinodal Tg curve is considered as unstable where the
6
7 drug-polymer exists in two phase heterogenous system. To the left-hand side exists the region
8
9
10 on drug-polymer single-phase homogenous system. At the right-hand side of the Tg curve
11
12 recrystallization of the drug can occur without any significant energy barrier. The solubility
13
14
15 and the miscibility curve intersecting the predicted glass transition curve can be seen.
16
17 Quantitative determination of miscibility of RLZ in PAA, PVP VA and HPMC AS was
18
19
observed to be approximately 10% w/w at <85 oC, 25% w/w at <71 oC and 35% w/w at <61
20
21
o
22 C, respectively. This revealed the extent of drug loading possible until these temperatures
23
24 achieved which could result into a stable ASD of RLZ without rendering any sort of
25
26
27 recrystallization. Additionally, in the case of PAA, and PVP VA at 50%w/w drug loading there
28
29 is positive deviation in the Tg value i.e., the observed value is higher than predicted. In case of
30
31
32 HPMC AS at the same drug loading a negative deviation in the Tg value is observed. A positive
33
34 deviation is the evidence of heteronuclear interaction which leads to enhanced stability. On the
35
36
37
other hand, the negative deviation is the evidence of homonuclear interaction which has
38
39 destabilizing effect [55]. From this, it can be concluded that PAA can offer high stability to
40
41 RLZ over higher temperatures; however, drug loading will be low. PVP VA observed to offer
42
43
44 higher drug loading among the three polymers with stability over moderate temperature. While
45
46 HPMC AS can produce a stable ASD with the highest drug loading but observed to have
47
48
49 temperature limited lower thermodynamic stability compared to PAA and PVP VA. However,
50
51 low drug loading observed in case of PAA can be easily addressed by adjusting the ASDs
52
53
54 amount as dose of RLZ was not very high (50mg taken orally twice daily). In addition, it may
55
56 not result in any pill burden at the cost of its stability. This experimental methodology is based
57
58
on the measurement of the dissolution of crystalline drug into polymeric matrix at high drug
59
60
61
62 33
63
64
65
loading and temperature. In this method, data generated is extrapolated to lower temperatures,
1
2 which is often argued to give overestimation or underestimation of miscibility of drug in
3
4
5 polymer. Therefore, this study is backed up with other studies to give a multi-faceted
6
7 understanding of the interaction of drug with the selected polymers. For instance,
8
9
10 intermolecular interactions like the formation of H-bonds or ionic bonds (as discussed in earlier
11
12 study) plays the key role in determining the capacity of polymer to prevent the drug present in
13
14
15 the ASD to recrystallize. The evaluation of the results obtained from all the studies should be
16
17 considered to generate evidences in grading of polymers according to their efficacy to produce
18
19
a stable ASD. Present study revealed PAA as effective polymer based on all above studies
20
21
22 conducted for grading of polymers for development of RLZ ASDs.
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62 34
63
64
65
1
2
3 14
a) 25 C 40 C 53 C 106 C 120 C 120 d)
4 12 Solubility Miscibility Predicted Tg
5
10 100
6
8
Temperature (oC)
7 80
8 6
ΔGmix/RT
9 4 60
10 2
11 40
0
12 -2
0 0.2 0.4 0.6 0.8 1
20
13
-4
14 Φdrug 0
15 -6 0.0 0.2 0.4 0.6 0.8 1.0
Φdrug
16 120
17 2
b)
25 C 40 C 53 C 106 C 120 C Solubility Miscibility Predicted Tg
18 e)
1.5 100
19 Temperature (oC)
20 1
80
21 0.5
ΔGmix/RT
22 60
0
23 0 0.2 0.4 0.6 0.8 1
-0.5 40
24
25 -1
20
26
-1.5
27 Φdrug 0
28 -2 0.0 0.2 0.4 0.6 0.8 1.0
Φdrug
29
30 6
c) 25 C 40 C 53 C 106 C 120 C 125 f)
31 Solubility Miscibility Predicted Tg
4
32 105
33 2
34
Temperature (oC)
ΔGmix/RT
0 85
35 0 0.2 0.4 0.6 0.8 1
36 -2 65
37
38 -4
45
39 -6
40 Φdrug
41 -8 25
0.0 0.2 0.4 0.6 0.8 1.0
42 ΦDrug
43
44
45
Figure 8. a) Plot of ΔGmix/RT vs drug weight fraction for RLZ-PAA b) Plot of ΔGmix/RT vs drug
46 weight fraction (Φdrug) for RLZ-PVP VA c) Plot of ΔGmix/RT vs drug weight fraction (Φdrug) for RLZ-
47 HPMC AS d) Binary phase diagram for RLZ-PAA e) Binary phase diagram for RLZ-PVP VA f)
48 Binary phase diagram for RLZ-HPMC AS showing the solid liquid equilibrium (blue), miscibility
49 (orange) and glass transition temperature (grey) curve.
50
51
52
53
54 4 Conclusion
55
56
57 The current study provides a multi-pronged approach in deciding the use of most appropriate
58
59 polymer according to its ability to provide stability to the ASD system. The initial screening of
60
61
62 35
63
64
65
polymer done by molecular modeling and DFT studies, deciphered the mode of interactions
1
2 between the drug and polymer along with the quantification of interactions in terms of energies.
3
4
5 The XRD studies of the developed ASD with different drug and polymer ratios, revealed the
6
7 capacity of polymer to stabilize a complete amorphous system. DSC thermogram showed the
8
9
10 homogenous phase by exhibiting single Tg for each ASD. In addition to this, in situ ASD
11
12 preparation helped in determining the drug-polymer system giving the Tg value of 106 oC, 53
13
14 o
15 C and 40 oC in 50:50 drug-polymer weight ratio, for ASD RLZ-PAA, RLZ-PVP VA and RLZ-
16
17 HPMC AS respectively. This defines the capacity of polymers to increase the low Tg (28 oC)
18
19
of pure amorphous RLZ, hence enhancing its stability. Chemical interaction studies with the
20
21
22 help of FT-IR and ssNMR showed the formation of a molecular dispersion and the possible
23
24 interactions occurring between the drug and polymer. Chemical stability was also confirmed
25
26
27 using HPLC analysis. Assessment of crystallization tendency of the prepared ASDs indicated
28
29 that PAA and PVP VA showed lower phase separation as the crystallization and fusion
30
31
32 enthalpies were lower with respect to ASDs prepared with HPMC AS. This study evaluated
33
34 the kinetic stability of ASDs prepared with different drug-polymer compositions. TEM studies
35
36
37
were capable in detecting the presence of crystalline drug even in nanometer scale after it was
38
39 exposed to high temperature, to ensure its thermodynamic stability. Additionally, in vitro
40
41 dissolution study also showed superior dissolution capacity of RLZ:PAA ASD over other
42
43
44 ASDs. The drug-polymer miscibility studies elaborated the stability space lying between the
45
46 percent drug loading and the temperature. ΔGmix of mixing at room, glass transition and melting
47
48
49 temperatures of the drug-polymer systems were also taken into account. Hence, it can be
50
51 inferred that PAA within its drug loading capacity would provide maximum stability to the
52
53
54 RLZ ASD followed by PVP VA and HPMC AS. This multi-faceted approach will aid a faster
55
56 and empirical selection of the most suitable polymer for the development of ASD which will
57
58
be stable throughout its stipulated shelf-life. Even though these studies provide sufficient
59
60
61
62 36
63
64
65
insight into the stability of ASDs, a real time stability study at accelerated conditions would
1
2 provide exact results, which will be the future scope of this study. By exploiting this approach
3
4
5 industries can equip themselves to produce more such products; as till date, commercially
6
7 available ASD products are very limited.
8
9
10
11 Acknowledgment
12
13
14 Authors would like to express their appreciation to Dr. Jagadish Sharma of the National
15
16 Institute of Pharmaceutical Education and Research, S.A.S Nagar for his invaluable guidance
17
18
19 throughout the project. Authors also extends their sincere thanks to Dr. Aman Kumar Lal Das
20
21 of the Department of Metallurgical Engineering, IT (BHU) for his assistance in TEM studies.
22
23
24 Additionally, authors would like to acknowledge the support of the Central Instrument Facility
25
26 at IIT (BHU) for providing the necessary resources to conduct the analytical studies.
27
28
29
Data Availability:
30
31
32
33 The raw/processed data required to reproduce these findings cannot be shared at this time as
34
35 the data also forms part of an ongoing study. However, original data included in this study will
36
37
38 be shared prior to publication, if required.
39
40
41
42
43 References
44
45
46 1. Baird, J.A., Taylor, L.S., 2012. Evaluation of amorphous solid dispersion properties using
47 thermal analysis techniques. Adv. Drug Deliv. Rev., The role of thermal analysis and
48 calorimetry in pharmaceutical design and development 64, 396–421.
49 https://doi.org/10.1016/j.addr.2011.07.009
50
51
2. Bhujbal, S.V., Mitra, B., Jain, U., Gong, Y., Agrawal, A., Karki, S., Taylor, L.S., Kumar,
52 S., (Tony) Zhou, Q., 2021. Pharmaceutical amorphous solid dispersion: A review of
53 manufacturing strategies. Acta Pharm. Sin. B, Hot Topic Reviews in Drug Delivery 11,
54 2505–2536. https://doi.org/10.1016/j.apsb.2021.05.014
55 3. Azad, M., Moreno, J., Davé, R., 2018. Stable and Fast-Dissolving Amorphous Drug
56
57 Composites Preparation via Impregnation of Neusilin® UFL2. J. Pharm. Sci. 107, 170–182.
58 https://doi.org/10.1016/j.xphs.2017.10.007
59
60
61
62 37
63
64
65
4. Thakore, S.D., Akhtar, J., Jain, R., Paudel, A., Bansal, A.K., 2021. Analytical and
1 Computational Methods for the Determination of Drug-Polymer Solubility and Miscibility.
2 Mol. Pharm. 18, 2835–2866. https://doi.org/10.1021/acs.molpharmaceut.1c00141
3
4 5. Wilson, V.R., Lou, X., Osterling, D.J., Stolarik, D.F., Jenkins, G.J., Nichols, B.L.B., Dong,
5 Y., Edgar, K.J., Zhang, G.G.Z., Taylor, L.S., 2020. Amorphous solid dispersions of
6 enzalutamide and novel polysaccharide derivatives: investigation of relationships between
7 polymer structure and performance. Sci. Rep. 10, 18535. https://doi.org/10.1038/s41598-
8
9
020-75077-7
10 6. Hiew, T.N., Zemlyanov, D.Y., Taylor, L.S., 2022. Balancing Solid-State Stability and
11 Dissolution Performance of Lumefantrine Amorphous Solid Dispersions: The Role of
12 Polymer Choice and Drug–Polymer Interactions. Mol. Pharm. 19, 392–413.
13 https://doi.org/10.1021/acs.molpharmaceut.1c00481
14
15 7. Wang, Y., Grohganz, H. and Rades, T., 2022. Effects of polymer addition on the non-
16 strongly interacting binary co-amorphous system carvedilol-tryptophan. International
17 Journal of Pharmaceutics, 617, p.121625. https://doi.org/10.1016/j.ijpharm.2022.121625
18 8. Sharma, J., Singh, B., Agrawal, A.K., Bansal, A.K., 2021. Correlationship of Drug-Polymer
19
Miscibility, Molecular Relaxation and Phase Behavior of Dipyridamole Amorphous Solid
20
21 Dispersions. J. Pharm. Sci. 110, 1470–1479. https://doi.org/10.1016/j.xphs.2020.12.007
22 9. He, Y., Ho, C., 2015. Amorphous Solid Dispersions: Utilization and Challenges in Drug
23 Discovery and Development. J. Pharm. Sci. 104, 3237–3258.
24 https://doi.org/10.1002/jps.24541
25
26 10. Yani, Y., Kanaujia, P., Chow, P.S., Tan, R.B.H., 2017. Effect of API-Polymer Miscibility
27 and Interaction on the Stabilization of Amorphous Solid Dispersion: A Molecular
28 Simulation Study. Ind. Eng. Chem. Res. 56, 12698–12707.
29 https://doi.org/10.1021/acs.iecr.7b03187
30 11. Wang, B., Wang, D., Zhao, S., Huang, X., Zhang, J., Lv, Y., Liu, X., Lv, G., Ma, X., 2017.
31
32 Evaluate the ability of PVP to inhibit crystallization of amorphous solid dispersions by
33 density functional theory and experimental verify. Eur. J. Pharm. Sci. 96, 45–52.
34 https://doi.org/10.1016/j.ejps.2016.08.046
35 12. Rezaeisadat, M., Bordbar, A.-K., Omidyan, R., 2021. Molecular dynamics simulation study
36
37
of curcumin interaction with nano-micelle of PNIPAAm-b-PEG co-polymer as a smart
38 efficient drug delivery system. J. Mol. Liq. 332, 115862.
39 https://doi.org/10.1016/j.molliq.2021.115862
40 13. Stipa, P., Marano, S., Galeazzi, R., Minnelli, C., Mobbili, G., Laudadio, E., 2021. Prediction
41 of drug-carrier interactions of PLA and PLGA drug-loaded nanoparticles by molecular
42
43 dynamics simulations. Eur. Polym. J. 147, 110292.
44 https://doi.org/10.1016/j.eurpolymj.2021.110292
45 14. Bazyari-Delavar, S., Badalkhani-Khamseh, F., Ebrahim-Habibi, A., Hadipour, N.L., 2020.
46 Investigation of host-guest interactions between polyester dendrimers and ibuprofen using
47
48
density functional theory (DFT). Comput. Theor. Chem. 1189, 112983.
49 https://doi.org/10.1016/j.comptc.2020.112983
50 15. Schrödinger Release 2019-4: MacroModel, Schrödinger; LLC: New York, NY, USA, 2019
51 16. Schrödinger Release 2019-4: Desmond Molecular Dynamics System, D. E. Shaw Research,
52 New York, NY, 2019. Maestro-Desmond Interoperability Tools, Schrödinger, New York,
53
54 NY, 2019
55 17. Terao, K. (2014). Poly(acrylic acid) (PAA). In: Kobayashi, S., Müllen, K. (eds)
56 Encyclopedia of Polymeric Nanomaterials. Springer, Berlin, Heidelberg
57 18. https://coatings.specialchem.com/product/r-ashland-plasdone-s-630-copovidone, accessed
58
59 on 16th January 2023
60
61
62 38
63
64
65
19. https://www.ashland.com/file_source/Ashland/Industries/Pharmaceutical/Links/PC-
1 12624.6_AquaSolve_HPMCAS_Physical_Chemical_Properties, accessed on 16th January
2 2023.
3
4 20. Schrödinger Release 2019-4: Glide, Schrödinger, LLC, New York, NY, 2019
5 21. Gaussian 09, Revision B.01, M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M.
6 A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G. A. Petersson, H.
7 Nakatsuji, M. Caricato, X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L.
8
9
Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T.
10 Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, J. A. Montgomery, Jr., J. E. Peralta, F.
11 Ogliaro, M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, T. Keith, R.
12 Kobayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi,
13 M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox, J. B. Cross, V. Bakken, C. Adamo,
14
15 J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli,
16 J. W. Ochterski, R. L. Martin, K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J.
17 J. Dannenberg, S. Dapprich, A. D. Daniels, O. Farkas, J. B. Foresman, J. V. Ortiz, J.
18 Cioslowski, and D. J. Fox, Gaussian, Inc., Wallingford CT, 2010
19
22. Stewart, J.J.P., 2008. Application of the PM6 method to modeling the solid state. J. Mol.
20
21 Model. 14, 499–535. https://doi.org/10.1007/s00894-008-0299-7
22 23. Hill, J.G., 2013. Gaussian basis sets for molecular applications. Int. J. Quantum Chem. 113,
23 21–34. https://doi.org/10.1002/qua.24355
24 24. Vreven, T., Byun, K.S., Komáromi, I., Dapprich, S., Montgomery, J.A.Jr., Morokuma, K.,
25
26 Frisch, M.J., 2006. Combining Quantum Mechanics Methods with Molecular Mechanics
27 Methods in ONIOM. J. Chem. Theory Comput. 2, 815–826.
28 https://doi.org/10.1021/ct050289g
29 25. Tomasi, J., Mennucci, B., Cancès, E., 1999. The IEF version of the PCM solvation method:
30 an overview of a new method addressed to study molecular solutes at the QM ab initio level.
31
32 J. Mol. Struct. THEOCHEM 464, 211–226. https://doi.org/10.1016/S0166-1280(98)00553-
33 3
34 26. Bharti, K., Mittal, P., Mishra, B., 2019. Formulation and characterization of fast dissolving
35 oral films containing buspirone hydrochloride nanoparticles using design of experiment. J.
36
37
Drug Deliv. Sci. Technol. 49, 420–432. https://doi.org/10.1016/j.jddst.2018.12.013
38 27. Saxena, D., Maiti, P., 2021. Utilization of ABS from plastic waste through single-step
39 reactive extrusion of LDPE/ABS blends of improved properties. Polymer 221, 123626.
40 https://doi.org/10.1016/j.polymer.2021.123626
41 28. Vikas, Viswanadh, M.K., Mehata, A.K., Sharma, V., Priya, V., Varshney, N., Mahto, S.K.,
42
43 Muthu, M.S., 2021. Bioadhesive chitosan nanoparticles: Dual targeting and
44 pharmacokinetic aspects for advanced lung cancer treatment. Carbohydr. Polym. 274,
45 118617. https://doi.org/10.1016/j.carbpol.2021.118617
46 29. Lapuk, S.E., Zubaidullina, L.S., Ziganshin, M.A., Mukhametzyanov, T.A., Schick, C. and
47
48
Gerasimov, A.V., 2019. Kinetic stability of amorphous solid dispersions with high content
49 of the drug: A fast scanning calorimetry investigation. International Journal of
50 Pharmaceutics, 562, pp.113-123. https://doi.org/10.1016/j.ijpharm.2019.03.039
51 30. Sun, Y., Tao, J., Zhang, G.G.Z., Yu, L., 2010. Solubilities of Crystalline Drugs in Polymers:
52 An Improved Analytical Method and Comparison of Solubilities of Indomethacin and
53
54 Nifedipine in PVP, PVP/VA, and PVAc. J. Pharm. Sci. 99, 4023–4031.
55 https://doi.org/10.1002/jps.22251
56 31. Yadav, B., Balasubramanian, S., Chavan, R. B., Thipparaboina, R., Naidu, V. G. M., &
57 Shastri, N. R. 2018. Hepatoprotective cocrystals and salts of riluzole: prediction, synthesis,
58
solid state characterization, and evaluation. Crystal Growth & Design, 18(2), 1047-1061.
59
60 https://doi.org/10.1021/acs.cgd.7b01514
61
62 39
63
64
65
32. Chieng, N., Aaltonen, J., Saville, D. and Rades, T., 2009. Physical characterization and
1 stability of amorphous indomethacin and ranitidine hydrochloride binary systems prepared
2 by mechanical activation. European Journal of Pharmaceutics and Biopharmaceutics, 71(1),
3
4 pp.47-54. https://doi.org/10.1016/j.ejpb.2008.06.022
5 33. Walden, D. M., Bundey, Y., Jagarapu, A., Antontsev, V., Chakravarty, K., & Varshney, J.
6 2021. Molecular simulation and statistical learning methods toward predicting drug–
7 polymer amorphous solid dispersion miscibility, stability, and formulation design.
8
9
Molecules, 26(1), 182. https://doi.org/10.3390/molecules26010182
10 34. Wang, B., Wang, D., Zhao, S., Huang, X., Zhang, J., Lv, Y., ... & Ma, X. 2017. Evaluate
11 the ability of PVP to inhibit crystallization of amorphous solid dispersions by density
12 functional theory and experimental verify. European Journal of Pharmaceutical Sciences,
13 96, 45-52. https://doi.org/10.1016/j.ejps.2016.08.046
14
15 35. Yang, G., Park, S.-J., 2019. Deformation of Single Crystals, Polycrystalline Materials, and
16 Thin Films: A Review. Materials 12, 2003. https://doi.org/10.3390/ma12122003
17 36. Kissi, E.O., Kasten, G., Löbmann, K., Rades, T., Grohganz, H., 2018. The Role of Glass
18 Transition Temperatures in Coamorphous Drug–Amino Acid Formulations. Mol. Pharm.
19
15, 4247–4256. https://doi.org/10.1021/acs.molpharmaceut.8b00650
20
21 37. Forster, A., Hempenstall, J., Tucker, I., Rades, T., 2001. Selection of excipients for melt
22 extrusion with two poorly water-soluble drugs by solubility parameter calculation and
23 thermal analysis. Int. J. Pharm. 226, 147–161. https://doi.org/10.1016/S0378-
24 5173(01)00801-8.
25
26 38. Chan, C.K. and Chu, I.M., 2001. Effect of hydrogen bonding on the glass transition behavior
27 of poly (acrylic acid)/silica hybrid materials prepared by sol–gel process. Polymer, 42(14),
28 pp.6089-6093. https://doi.org/10.1016/S0032-3861(01)00091-X
29 39. Al-Obaidi, H. and Buckton, G., 2009. Evaluation of griseofulvin binary and ternary solid
30 dispersions with HPMCAS. AAPS PharmSciTech, 10, pp.1172-1177.
31
32 https://doi.org/10.1208/s12249-009-9319-x
33 40. Mistry, P., Mohapatra, S., Gopinath, T., Vogt, F.G., Suryanarayanan, R., 2015. Role of the
34 Strength of Drug–Polymer Interactions on the Molecular Mobility and Crystallization
35 Inhibition in Ketoconazole Solid Dispersions. Mol. Pharm. 12, 3339–3350.
36
37
38 41. Song, Y., Yang, X., Chen, X., Nie, H., Byrn, S., Lubach, J.W., 2015. Investigation of Drug–
39 Excipient Interactions in Lapatinib Amorphous Solid Dispersions Using Solid-State NMR
40 Spectroscopy. Mol. Pharm. 12, 857–866. https://doi.org/10.1021/mp500692a
41 42. Pugliese, A., Toresco, M., McNamara, D., Iuga, D., Abraham, A., Tobyn, M., Hawarden,
42
43 L.E., Blanc, F., 2021. Drug–Polymer Interactions in
44 Acetaminophen/Hydroxypropylmethylcellulose Acetyl Succinate Amorphous Solid
45 Dispersions Revealed by Multidimensional Multinuclear Solid-State NMR Spectroscopy.
46 Mol. Pharm. 18, 3519–3531. https://doi.org/10.1021/acs.molpharmaceut.1c00427
47
48
43. Miyoshi, T., Takegoshi, K., Hikichi, K., 1997. High-resolution solid state 13C n.m.r. study
49 of the interpolymer interaction, morphology and chain dynamics of the poly(acrylic
50 acid)/poly(ethylene oxide) complex. Polymer 38, 2315–2320.
51 https://doi.org/10.1016/S0032-3861(96)00799-9
52 44. Sarpal, K., Delaney, S., Zhang, G.G.Z., Munson, E.J., 2019. Phase Behavior of Amorphous
53
54 Solid Dispersions of Felodipine: Homogeneity and Drug–Polymer Interactions. Mol.
55 Pharm. 16, 4836–4851. https://doi.org/10.1021/acs.molpharmaceut.9b00731
56 45. Mahamid, J., Sharir, A., Addadi, L., Weiner, S., 2008. Amorphous calcium phosphate is a
57 major component of the forming fin bones of zebrafish: Indications for an amorphous
58
precursor phase. Proc. Natl. Acad. Sci. 105, 12748–12753.
59
60 https://doi.org/10.1073/pnas.0803354105
61
62 40
63
64
65
46. Li, N., Gilpin, C.J., Taylor, L.S., 2017. Understanding the Impact of Water on the
1 Miscibility and Microstructure of Amorphous Solid Dispersions: An AFM–LCR and TEM–
2 EDX Study. Mol. Pharm. 14, 1691–1705.
3
4 https://doi.org/10.1021/acs.molpharmaceut.6b01151
5 47. S’ari, M., Blade, H., Cosgrove, S., Drummond-Brydson, R., Hondow, N., Hughes, L.P.,
6 Brown, A., 2021. Characterization of Amorphous Solid Dispersions and Identification of
7 Low Levels of Crystallinity by Transmission Electron Microscopy. Mol. Pharm. 18, 1905–
8
9
1919. https://doi.org/10.1021/acs.molpharmaceut.0c00918
10 48. Mondal, P.K., T, A., Rao, V., Chopra, D., 2019. Crystal structure analysis of the biologically
11 active drug mol-ecule riluzole and riluzolium chloride. Acta Crystallogr. Sect. E
12 Crystallogr. Commun. 75, 1084–1089. https://doi.org/10.1107/S2056989019009022
13 49. Dyer, A. M., & Smith, A. 2016. Riluzole 5 mg/mL oral suspension: for optimized drug
14
15 delivery in amyotrophic lateral sclerosis. Drug design, development and therapy, 59-
16 64.https://doi.org/10.2147/DDDT.S123776
17 50. https://drugcentral.org/drugcard/2382
18 51. Brady, J., Dürig, T., Lee, P. I., & Li, J. X. 2017. Polymer properties and characterization. In
19
Developing solid oral dosage forms (pp. 181-223). Academic Press.
20
21 https://doi.org/10.1016/B978-0-12-802447-8.00007-8
22 52. Thakore, S. D., Thakur, P. S., Shete, G., Gangwal, R., Narang, A. S., Sangamwar, A. T., &
23 Bansal, A. K. (2019). Assessment of biopharmaceutical performance of supersaturating
24 formulations of carbamazepine in rats using physiologically based pharmacokinetic
25
26 modeling. AAPS PharmSciTech, 20, 1-12. https://doi.org/10.1208/s12249-019-1386-z
27 53. Potter, C.B., Davis, M.T., Albadarin, A.B., Walker, G.M., 2018. Investigation of the
28 Dependence of the Flory–Huggins Interaction Parameter on Temperature and Composition
29 in a Drug–Polymer System. Mol. Pharm. 15, 5327–5335.
30 https://doi.org/10.1021/acs.molpharmaceut.8b00797
31
32 54. Meng, F., Dave, V., Chauhan, H., 2015. Qualitative and quantitative methods to determine
33 miscibility in amorphous drug–polymer systems. Eur. J. Pharm. Sci. 77, 106–111.
34 https://doi.org/10.1016/j.ejps.2015.05.018
35 55. Baghel, S., Cathcart, H., O’Reilly, N.J., 2016. Theoretical and experimental investigation
36
37
of drug-polymer interaction and miscibility and its impact on drug supersaturation in
38 aqueous medium. Eur. J. Pharm. Biopharm. 107, 16–31.
39 https://doi.org/10.1016/j.ejpb.2016.06.024
40 56. Marsac, P.J., Li, T., Taylor, L.S., 2009. Estimation of Drug–Polymer Miscibility and
41 Solubility in Amorphous Solid Dispersions Using Experimentally Determined Interaction
42
43 Parameters. Pharm. Res. 26, 139–151. https://doi.org/10.1007/s11095-008-9721-1
44 57. Rim, P.B., Runt, J.P., 2002. Melting point depression in crystalline/compatible polymer
45 blends [WWW Document]. ACS Publ. https://doi.org/10.1021/ma00138a017
46 58. Newman, A., Zografi, G., 2019. Commentary: Considerations in the Measurement of Glass
47
48
Transition Temperatures of Pharmaceutical Amorphous Solids. AAPS PharmSciTech 21,
49 26. https://doi.org/10.1208/s12249-019-1562-1
50 59. Knopp, M.M., Tajber, L., Tian, Y., Olesen, N.E., Jones, D.S., Kozyra, A., Löbmann, K.,
51 Paluch, K., Brennan, C.M., Holm, R., Healy, A.M., Andrews, G.P., Rades, T., 2015.
52 Comparative Study of Different Methods for the Prediction of Drug–Polymer Solubility.
53
54 Mol. Pharm. 12, 3408–3419. https://doi.org/10.1021/acs.molpharmaceut.5b00423
55
56
57
58
59
60
61
62 41
63
64
65

JDDST D 23 02773 - R1

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

JDDST D 23 02773 - R1

Uploaded by

Copyright:

Available Formats

Journal of Drug Delivery Science and Technology

A Multifaceted Approach for Grading of Polymers for the Development of Stable

Manuscript Number: JDDST-D-23-02773R1

Article Type: Research Paper

Keywords: Amorphous solid dispersion; Drug-Polymer Interactions; Density functional theory;

Corresponding Author: Brahmeshwar Mishra, Ph.D.

First Author: Kanchan Bharti

Order of Authors: Kanchan Bharti

Brahmeshwar Mishra, Ph.D.

Suggested Reviewers: Krishna Kumar Patel, Ph.D.

Rahul Tripathi, Ph.D.

Prof. Dr. Florence Siepmann,

Subject: Submission of Original Research Manuscript for Publication

Dear Prof. Siepmann,

Response sheet on submission JDDST-D-23-02773

Comment 2. Describe the physicochemical properties of RLZ such as solubility in the

A Multifaceted Approach for Grading of Polymers for the

kanchan.bharti.rs.phe18@itbhu.ac.in (Kanchan Bharti)

dubeygurudutt93@gmail.com (Dr. Gurudutt Dubey)

manishkumar.rs.phe19@itbhu.ac.in (Manish Kumar)

abhishekjha.phe17@itbhu.ac.in (Abhishek Jha)

mansiupadhyay14@gmail.com (Dr. Mansi Upadhyay)

malips01@gmail.com (Pramod S Mali)

ashutoshk@iitb.ac.in (Prof. Ashutosh Kumar)

pvbharatam@niper.ac.in (Prof. Prasad V. Bharatam)

bmishrabhu@rediffmail.com (Prof. Brahmeshwar Mishra)

thermodynamically stable amorphous solid dispersion (ASD) of a neuroprotective drug

Detailed computational studies were performed to quantify the non-covalent interactions

Experimental characterization of drug-polymer complexes through analytical techniques

polymers in search of a stable ASD of the poorly water-soluble drug RLZ.

Keywords: Amorphous Solid Dispersion, Drug-Polymer Interactions, Density functional

theory, Computational study, Miscibility, Flory-Huggins theory.

solubility as well as stability enhancement [1]. Development of such ASDs is a widely

therapeutic activity [3].

effects demand an improved approach of mechanistic understanding of drug and polymers

interplay in crystallization inhibition [5]. Single-phase amorphous binary systems such as

Therefore, drug-polymer interactions need to be studied well to determine the homogenous

polymer [4], [8].

The current study involvesed a detailed study to comprehendunderstand the drug-polymer

interactions in the formulation of a kinetically and thermodynamically stable ASD. Riluzole

infrared spectroscopy (FT-IR), solid-state NMR (ssNMR), microscopic analysis by using

scanning electron microscopy (SEM) and transmission electron microscopy (TEM).

polymers in terms of Tg, drug crystallization, demixing, drug-polymer miscibility, phase

that may significantly affect the physical stability of ASDs.

2 Materials and Method

the preparation of ASD.

2.1 Methodology for in silico studies

2.1.1 Preparation of 3D structures of polymers

Powell-Reeves conjugate gradient (PRCG) method. The resultant optimized geometries of

polymers were subjected to 20 ns of molecular dynamics simulations using Desmond module

compositions were considered based on the literature reports:

pentaerythreitol were used in a linear combination (AAAAAAAAAABAAAAAAAAA) to

of the polymer [17].

containing PVP VA were 1698 and C80H112N10O30, respectively.

chemical formula 2428 and C102H164O65, respectively.

3D model of RLZ for further computational calculations.

dynamics simulation and quantum chemical energy minimization of drug-polymer complexes,

2.1.2 Molecular docking:

2.1.3 Molecular dynamics simulations (MDS):

were simulated for 40 ns at 2 ps record interval and 8ps trajectory.

was optimized at B3LYP/6-31+G(d, p) level of theory [23]. RLZ-polymer complex werewas

Einteraction = Ecomplex – [Epolymer + Edrug]

and drug respectively.