NMR Relaxometry to Characterize the Drug 
Structural Phase in a Porous Construct
Authors: Linn W. Thrane, Emily A. Berglund, James 
N. Wilking, David Vodak, and Joseph D. Seymour
This document is the unedited author's version of a Submitted Work that was subsequently 
accepted for publication in Molecular Pharmaceutics, copyright © American Chemical Society 
after peer review. To access the final edited and published work, see https://doi.org/10.1021/
acs.molpharmaceut.8b00144.
Thrane, Linn W., Emily A. Berglund, James N. Wilking, David Vodak, and Joseph D. Seymour, 
“NMR Relaxometry to Characterize the Drug Structural Phase in a Porous Construct,” Molecular 
Pharmaceutics, June 2018; 15(7): 2614-2620. doi: 10.1021/acs.molpharmaceut.8b00144
Made available through Montana State University’s ScholarWorks 
scholarworks.montana.edu 
NMR Relaxometry to Characterize the Drug Structural Phase in a
Porous Construct
Linn W. Thrane,†,‡ Emily A. Berglund,† James N. Wilking,† David Vodak,§ and Joseph D. Seymour*,†
†Department of Chemical and Biological Engineering, Montana State University, Bozeman, Montana 59717-3920, United States
‡Department of Mechanical and Industrial Engineering, Montana State University, Bozeman, Montana 59717, United States
§Bend Research Incorporated, Lonza, Bend, Oregon 97701, United States
ABSTRACT: Nuclear magnetic resonance (NMR) frequency spectra and T2 relaxation time measurements, using a high-power
radio frequency probe, are shown to characterize the presence of an amorphous drug in a porous silica construct. The results
indicate the ability of non-solid-state NMR methods to characterize crystalline and amorphous solid structural phases in drugs.
Two-dimensional T1−T2 magnetic relaxation time correlation experiments are shown to monitor the impact of relative humidity
on the drug in a porous silica tablet.
KEYWORDS: NMR relaxometery, porous construct, amorphous, crystalline, melt solidification
■ INTRODUCTION
Drug solubility depends on whether the drug is in the
crystalline or amorphous solid state.1 Methods such as spray
drying2 and pore entrapment3−6 are well established
approaches to impact molecular dynamics during solidification
and control the drug structural phase. Nuclear magnetic
resonance (NMR) methods have been applied to characterize
the amorphous and crystal structures of pharmaceuticals using
solid-state (ssNMR) methods.7−10 These ssNMR methods
require sample spinning and typically grinding of the sample
into a powder, and the incompatibility of the methods with the
liquid state precludes observing transitions from the liquid melt
to the solid state during solidification. Liquid-state NMR
relaxometry is well established to characterize pore size
distributions in porous media.11,12 NMR relaxation times are
sensitive to phase transitions from the liquid to the solid
state.13,14 Somewhat analogous to the solidification processes
studied here is NMR cryoporometry in porous media, which
relies on the Gibbs−Thomson effect and the presence of
nanoscale unfrozen water layers at the boundary of the solid
porous matrix and the ice crystals in the bulk pores to
characterize porous systems.15,16
In this article, we describe the use of a purpose built high-
power radio frequency (rf) pulse probe to measure the
solidification of the drug fenofibrate in bulk and within a
porous matrix. The probe allows for high-power short duration
rf pulses capable of acquiring signal from the solid-state drug
without sample spinning. NMR spectra indicate amorphous
domains after solidification in the porous matrix that are not
present in the bulk solidification, assisting in the verification
and interpretation of the novel relaxation data presented. The
impact of relative humidity (RH) on the structure of the solid-
state drug in the porous matrix is shown to correlate with NMR
relaxation behavior. The results quantify the impact of porous
matrix restriction on the drug solidification and indicate the
potential for quality control and process monitoring by NMR.
■ THEORY
NMR frequency spectra and magnetic relaxation times are
sensitive to the liquid or solid phase of the sample and the
solid-state structural crystal or amorphous phase. The
frequency spectra of liquids are Lorentzian due to the
exponential form of the time-dependent free induction decay
(FID) voltage signal, while those of solids are Gaussian.17,18
This is due to motional averaging of the dipolar coupling of the
1H protons in liquids by rotational diffusion.17 In solids, the
correlation time of the dipolar coupling is much longer and
results in a Gaussian FID that decays away rapidly in time
(∼1−100 μs), generating a broad Gaussian line width in the
frequency domain.18 The spin−spin T2 relaxation time depends
on the proton dipolar coupling and is short (<1 ms) for solids
and longer (>100 ms) for liquids, and it determines the line
shape and line width. The solid phase ordering in crystalline
phases increases the dipolar coupling with corresponding fast in
time Gaussian FID decays with a T2 shorter than that in
amorphous solids, where disorder and slightly increased
mobility generate longer T2 and potentially shorter time
Gaussian and longer time exponential FIDs.12,19 The spin−
lattice T1 relaxation in liquids is at the order of seconds, and T1
∼ T2, while in solids, T1 is much longer than T2.
12 The solid
phase T1 in more ordered crystalline states is longer than that in
amorphous states.
■ METHODS
The porous matrix used in these experiments was a porous
tablet made from colloidal silica. Silica with a stable-ordered
mesoporous structure has been shown to be a promising
controlled drug delivery system.3 Charge-stabilized, amorphous
colloidal silica (Nissan Chemical, MP-1040-H) suspended in
water with an average particle diameter of d = 100 nm and low
size polydispersity (<5%) was used for all of the experiments.
The surface of the colloidal silica is negatively charged with a
surface charge density between 8 and 10 nm−1 and free of
residual organics, as stated by the manufacturer. Colloidal silica
(300 μL) was added to cylindrical molds and left to solidify for
24 h. The tablet dimensions were set by the dimensions of the
mold used during the slipcasting process. Tablets were
cylindrical in shape, with a diameter and height on the order
of 0.5−1 cm. Silica particles were packed into a random
structure, as revealed by small angle neutron scattering
experiments. The characteristic pore size could be estimated
by calculating the void created by three kissing spheres: D =
0.15d; with d estimated as d ≈ 15 nm. After 24 h, the
fenofibrate was heated to 10 °C above its melting point of 81.1
°C, and the drug was loaded into the tablet by imbibing the
drug in liquid form. The fenofibrate was absorbed into the
tablet by capillary action. Once the imbibition process was
complete, the mass fraction of the drug in the composite was
0.18 ± 0.01, as measured using thermogravimetric analysis.
This corresponds to the silica colloid volume fraction of φ =
0.71, based on the density of amorphous silica, ρsilica = 2.196 g/
cm3, and the density of fenofibrate, ρfenofibrate = 1.18 g/cm
3.
Measurements indicate that the void space was filled
completely by the drug during imbibition. Finally, the tablet
was wiped down to remove any excess drug on the surface
before it was placed in a 5 mm NMR tube. The bulk drug
sample was prepared by adding the powdered drug to a 5 mm
NMR tube and melting the drug in the tube. The sample was
quickly placed in a Bruker 250 MHz superconducting magnet
integrated to an Avance III spectrometer, where frequency
spectra and T2 measurements were performed using a high-
power rf probe with a 5 mm rf coil custom built by Bruker. The
temperature was controlled at 20 °C using the Bruker BTU
system with N2 gas flow. The T2 measurements were executed
using the standard Carr−Purcell−Meiboom−Gill (CPMG)
pulse sequence with an echo time of τE = 12 μs, 500−4000
echoes, depending on the sample, acquisition sampling dwell
time of 1 μs, and 7.5 μs rf pulses at 100 W power. The signal
was acquired using 4 averages of a 16-phase cycle acquisition.
The rapid data acquisition in the short 2·τE period precluded
full acquisition of the echo, so the data were not encoded for
spectral frequency shift. Experiments with τE varying from 12 to
400 μs indicated no T2 dispersion effect due to potential spin
locking of longer T2 components with rotating frame spin−
lattice relaxation times T1ρ of similar magnitude.
10,20 The T2
relaxation data were converted by inverse Laplace trans-
formation to a distribution of relaxation populations.11,19,21
To explore the impact of RH on the solidified drug with
time, tablets were stored in an environment with controlled RH
at T = 20 °C. The 0% RH conditions were set by sealing the
tablet in a container with an excess of dry desiccant (Drierite).
The 100% RH conditions were set by sealing the composites in
a container with excess liquid water. Liquid water was not
allowed to be in contact with the tablet, but it created a
saturated vapor pressure at equilibrium in an enclosed
environment. Measurements of a tablet kept in a 0% RH
environment were compared to the measurements of a tablet
kept in a 100% RH environment for 2 weeks. T1 and T2
relaxation data are acquired by performing 1D T2 experiments
and 2D T1−T2 correlation experiments. The T1-T2 measure-
ments are performed using an inversion recovery combined
with a CPMG pulse sequence.11 The same parameters used in
the T2 measurements were used for the CPMG portion of the
T1−T2 measurement, except that 2 averages of the 16-phase
cycle acquisition were used. Inversion recovery was performed
with inversion times logarithmically spaced from 1 ms to 50 s.
Inverse Laplace transforms of the relaxation data were
performed in 1D for T2 and 2D for T1−T2 experiments to
provide distributions of relaxation times. Data were collected
on day 1, 3, 7, and 14 of the storage in 100% RH. To avoid
exposing the sample to lower RH during measurements, four
samples were prepared so that none of the tablets were
removed from the high RH environment until the measure-
ment was performed.
■ RESULTS AND DISCUSSION
Drug Solidification. Once the tablet or the bulk drug is
placed in the magnet, frequency spectra and T2 measurements
are interleaved and performed continuously until the drug is
fully solidified and no further changes are observed. The
fenofibrate is originally in powder form of ground crystal but
takes on a more ordered crystalline structure when it solidifies
after being melted.22 This is indicated by the T2 measurements
of the drug in powder form and after the melted drug has
solidified.
The T2 distributions in Figure 1 show that the drug has two
T2 populations before and after being melted, and the signal
amplitude distribution between the two peaks remains
approximately the same. A shift toward shorter T2 relaxation
times is observed in the melted and solidified drug. Addition-
ally, the total signal amplitude of the solidified drug is 2 orders
of magnitude lower than that of the powdered drug; however,
this cannot be seen in Figure 1 due to both of the T2
distributions being normalized by the population amplitude
maximum for visualization of the shift in the T2 relaxation time.
The decrease in the total measured signal amplitude in the
recrystallized system is due to the T2 relaxation times of the
drug becoming too short, i.e., less than a τE of 12 μs, for the
NMR pulse sequence to detect, indicating a more ordered
crystalline structure than the powder form. The two
populations thus represent the most rotationally mobile
protons in the system with T2 > ∼50 μs. The populations
cannot be assigned to specific moieties on the molecule.
However, from spectroscopic experiments and quantum
mechanical simulations, the crystal structure indicates stacking
of benzene rings.22 The population at T2 ∼ 10−4 s shifts to
lower T2 values in the melted solidified sample relative to the
powdered form. It is expected that the melt process will allow
reorientation of molecules for a smaller molecular spacing in
the crystal, and protons in this population indicate that effect.
The more rotationally mobile protons at T2 ∼ 1 × 10−3 s are
associated with the methyl groups furthest from the stacked
benzene rings, and the broad distribution of T2 values indicates
that they experience varying rotational mobility depending on
the exact location on the molecule, which is consistent with the
model of Heinz et. al.22
The bulk drug took approximately 30 min to fully solidify
after melting, while it took 3−4 h for the drug in the porous
tablet to solidify. To ensure no further changes occurred in the
tablet past the first 4 h, the sample was stored, and additional
measurements were performed after 21 h. Figure 2a shows the
frequency spectra of the drug in bulk in the liquid melt state at
0 and in the solid state at 27 min. The melt state signal is a
convolution of a Gaussian and Lorentzian decay. This indicates
the high degree of molecular order, likely due to π stacking of
the benzene rings in the molten state. The Lorentzian portion
of the frequency spectrum at 0 min is liquid-like, where the
molecular mobility, i.e., the random modulation of the dipolar
coupling, is fast. In bulk, the drug solidifies with unrestricted
molecular mobility, and the final result is a highly ordered
crystalline structure. The slow random modulation of the
dipolar coupling produces a pure Gaussian line shape in the
frequency spectrum, as can be seen in Figure 2a at 27 min.18
A similar Lorentzian distribution is observed in the frequency
spectra for the drug in a porous tablet in the liquid melt state at
0 min (Figure 2b). However, in addition to the broad Gaussian
distribution seen in the pure drug, a remaining Lorentzian
distribution can be seen at 21 h for this sample. The
combination of Gaussian and Lorentzian line shapes in the
frequency spectrum suggests that when restricted in a porous
media, some of the drug will not reach the highly ordered
crystalline structure of the pure drug. The Gaussian line shape
confirms that the drug has solidified, but the remaining
Lorentzian line shape indicates the presence of a highly
amorphous state of the solidified drug or supercooled
liquid.23,24
When the bulk drug solidifies, a highly ordered bulk crystal
structure forms. In a confined environment of 15 nm pores, as
used here, fenofibrate has been shown to produce nanocrystals
by 13C MAS ssNMR and XRPD.25 In that study, ssNMR data
showed line broadening with the pore size decreasing from 300
to 20 nm, which was attributed to the surface disorder as
nanocrystals became small, thus differentiating the surface drug
versus nanocrystalline core drug.25 At 12 nm pore size, the
ssNMR still indicated crystalline drug, but the 13C spectra
broadening and XRPD data indicated loss of long-range order.
WAXS diffraction on the bulk fenofibrate and fenofibrate in the
porous construct for our samples shows a behavior similar to
the XRPD data of reference,25 with crystalline behavior in bulk
and a large background silica signal in the small pore tablet
(Figure 3). The line width of the Gaussian portion of the
spectra for the bulk and tablet-solidified samples (Figure 2) is
similar, indicating the presence of nanocrystals of drug in the
Figure 1. T2 distributions of the drug in powder form (blue/black)
and the solidified drug after being melted and recrystallized (red/
gray). Two populations are observed in both samples, and the signal
intensity ratio between the two populations remains approximately the
same before and after melting. The population with the shortest T2
time experiences a shift toward shorter relaxation times when
solidifying after being melted.
Figure 2. Frequency spectra of drug solidifying in (a) bulk and (b) tablet. In both samples, the drug starts out (red/gray) as highly amorphous, as
indicated by the Lorentzian shape of the frequency spectra. Once the solidification process is complete, the pure drug has solidified (black) to a
highly ordered crystalline structure, resulting in a broad Gaussian frequency peak. The final frequency spectrum for the drug in the tablet is a
combination of a broad Gaussian distribution and a Lorentzian, indicating that the drug solidifying in the restricted environment of the porous matrix
is more amorphous than that of the bulk sample.
pore. The Lorentzian component indicates an amorphous
component due to the impact of the silica−fenofibrate surface
interactions. This could be due to the supercooled liquid or
amorphous surface solid. Given the rapid crystallization
behavior of supercooled fenofibrate, which occurs on the
time scale of minutes in solution in the presence of a nucleation
site,23,24 we believe it is a surface-impacted solid.25
The T2 data collected during the solidification process
(Figure 4) indicate that a majority of the signal originates from
the population with T2 relaxation times of ∼10−3 s, 9.9 × 10−4 s
in the bulk fenofibrate and 9.3 × 10−4 s in the tablet, when the
drug is in liquid form. Note that the most rotationally mobile
protons in the melt at T2 ∼ 10−2 s are no longer fully resolved
in the solidifying sample. These most mobile protons associated
with the methyl protons become a shoulder on the T2 ∼ 10−3 s
population as the crystal forms. In both systems, the signal
amplitude of the ∼10−3 s T2 time population decreases
significantly, and the signal amplitude of the shorter T2 time
population, ∼10−4 s, increases during the drug solidification.
These results indicate that the relaxation time of a portion of
the fenofibrate shifts from the decreasing long T2 population to
the short T2 population, and the remaining drug exhibits T2
times too short to detect with the 12 μs τE. In the bulk
fenofibrate (Figure 4a), the signal amplitude of the ∼10−3 s T2
population keeps decreasing until a greater portion of the
remaining signal originates from the ∼10−4 s T2 population. For
the fenofibrate in the tablet (Figure 4b), the signal amplitude of
the ∼10−3 s T2 population stops decreasing when the remaining
signal is evenly distributed between the two populations. The
even distribution of signal amplitude in the final state of
fenofibrate in the porous medium compared to that of the bulk
fenofibrate confirms the results from the frequency spectra that
the drug takes on a more amorphous structure when solidifying
within a restricted environment.25
Impact of Relative Humidity. The impact of RH on the
structure of a solid-state drug is relevant to long time storage of
the drug. If stored in a high RH environment over time, water
will begin to inhabit the pore space of the tablet due to capillary
condensation. T1−T2 correlation measurements are used to
quantify the effect of 100% RH on fenofibrate confined in a
porous matrix over a 2 week period. For comparison, T1-T2
data were collected from fenofibrate in a tablet stored at 0%
RH. At 0% RH, the T1−T2 correlation data (Figure 5) reveal
two main populations with identical T1 relaxation times and T2
times at the order of 10−4 and 10−3 s and the broad shoulder
out to 10−2 s on the 10−3 s population. These are similar T2
times to those observed in the T2 distributions in Figures 1 and
4. The tablet stored at 0% RH showed no significant changes
over time.
T1−T2 measurements from day 1 of a tablet stored at 100%
RH reveal a third population in the T1−T2 correlation map
closer to the parity line (gray dashed line, Figure 6a),
representing T1 = T2. Populations located on or close to the
parity line are representative of a liquid phase, and it can
therefore be assumed that this third population is water that has
entered the porous matrix of the tablet as a result of its high RH
environment. The signal amplitude of the population
representing water increases with time and is approximately 1
order of magnitude greater by day 14. The water population is
also shifting toward longer T1 and T2 relaxation times, with T1
increasing from 10−2 s on day 1 to 10−1 s by day 14, and T2
increasing from 10−3 to 10−2 s by day 14. The increase in
relaxation time along with the increase in signal amplitude of
the water population indicate that the water is progressively
occupying more of the pore space as well as larger pores in the
Figure 3. Wide-angle X-ray (WAXS) diffraction spectra for the
fenofibrate−SiO2 composite with unprocessed crystalline fenofibrate
powder direct from the manufacturer. The primary Bragg peaks in the
composite scattering do not match up with those in the crystalline
active, indicating they do not have the same crystal structure. A broad
amorphous SiO2 peak dominates the composite sample signal.
Figure 4. T2 distributions of drug solidifying in (a) bulk and (b) tablet. The full signal amplitude is shown in the right corner of the plots. There is a
large signal loss in both samples due to the solidification of the drug. Once solidification is complete in the bulk sample, most of the signal from the
longer T2 population, ∼10−3 s, is lost, and most of the signal results from the shorter T2 population, ∼10−4 s. In the tablet sample, there is an equal
signal distribution between the ∼10−3 s and the ∼10−4 s T2 populations when the solidification is complete, confirming the results seen in the
frequency spectra suggesting that the drug in the tablet is in a more amorphous mobile state than that of the highly ordered bulk sample.
porous matrix over the 14 day period. The two populations
representing the drug in the tablet decrease in amplitude over
the 14 day period but do not experience a significant shift in T1
or T2 relaxation time.
This is more readily observed in Figure 7, where the 1D T2
distributions from day 1, 3, 7, and 14 are presented. The T2
distributions are normalized to the maximum population
amplitude recorded on day 14 so that the growth of the
water peak can be visualized. The population with highest
signal amplitude in the T2 distribution for day 1 (Figure 7a)
represents the signal originating from both the highly restricted
mobility water population and the drug population with 10−3 s
T2 from Figure 6a. On day 1, these two populations exhibit
similar T2 times and can therefore not be distinguished in the
T2 distribution. On day 3, the water population has experienced
a large enough increase in T2 times to appear as a third peak
with a T2 of 5.6 × 10
−3 s in Figure 7b. This very short T2 time
indicates water in highly restricted thin films. The water peak
keeps increasing in signal amplitude and shifting toward longer
T2 times until a T2 of 2.3 × 10
−2 s is reached on day 14 (Figure
7d). This T2 time indicates highly restricted water relative to
bulk, but the near order of magnitude increase from day 3
suggests significant changes in the restriction length scale. The
two populations representing the drug have experienced a slight
decrease in signal amplitude by day 14 but no significant shift in
Figure 5. T1−T2 correlation data from the model drug in a tablet kept
at 0% RH. The T1−T2 data indicate that there are two populations
with different T2 relaxation times present in the tablet at 0% RH. The
two populations have identical T1 relaxation times and T2 times on the
order of 10−4 and 10−3 s.
Figure 6. T1−T2 correlation data of drug in a tablet stored at 100% RH for (a) 1, (b) 3, (c) 7, and (d) 14 days. It can be seen already on day 1 that a
third population has appeared closer to the T1 = T2 parity line (gray dashed line) when compared to the T1−T2 data at 0% RH. This population is in
liquid form and is attributed to water building up in the porous matrix of the tablet. The peak representing water is increasing in amplitude with time
and is approximately 1 order of magnitude greater by day 14. The two populations representing the drug trapped within the pore space of the tablet
are decreasing in amplitude with time.
T2 times, indicating drug insolubility. Since the drug appears
insoluble and the water occupies larger length scale environ-
ments, the pore structure must be altered by the water through
tablet volume expansion or breakup. This effect is also observed
visually.
■ CONCLUSION
NMR frequency spectra and 1D T2 measurements were used to
characterize the solidification of fenofibrate in bulk and in a
porous matrix. The results indicate that fenofibrate takes on an
amorphous structure in a confined environment of ∼15 nm
compared to that of the highly ordered crystal structure formed
by the bulk drug, and they confirm the application of non-solid-
state NMR to distinguish and characterize crystalline and
amorphous solid structural phases in drugs. One-dimensional
T2 and 2D T1−T2 relaxation correlation experiments were used
to investigate the impact of RH on fenofibrate in a porous silica
tablet. Over a 14 day period, water was observed occupying
progressively more of the porous silica matrix, decreasing the
total drug signal detected, and occupying an increasing size
pore space. The results indicate the potential for drug quality
control by NMR relaxometry. The further development of low-
field NMR systems opens the way toward online observation
during storage as well as during dissolution studies under
varying conditions.
■ AUTHOR INFORMATION
Corresponding Author
*E-mail: jseymour@montana.edu.
ORCID
Joseph D. Seymour: 0000-0003-4264-5416
Notes
The authors declare no competing financial interest.
■ ACKNOWLEDGMENTS
J.D.S. acknowledges Bend, Lonza, for research funding, the U.S.
NSF MRI program and Murdock Charitable Trust for
equipment funding, and Dieter Gross at Bruker BioSpin for
construction of the high-power probe. J.N.W. acknowledges the
NSF (DMR-1455247) for research funding. The authors thank
S. L. Codd for assistance.
■ REFERENCES
(1) Murdande, S. B.; et al. Solubility Advantage of Amorphous
Pharmaceuticals: I. A Thermodynamic Analysis. J. Pharm. Sci. 2010, 99
(3), 1254−1264.
(2) Friesen, D. T.; et al. Hydroxypropyl Methylcellulose Acetate
Succinate-Based Spray-Dried Dispersions: An Overview. Mol.
Pharmaceutics 2008, 5 (6), 1003−1019.
(3) Wang, S. B. Ordered mesoporous materials for drug delivery.
Microporous Mesoporous Mater. 2009, 117 (1−2), 1−9.
Figure 7. T2 distributions of the drug in a tablet stored at 100% RH for (a) 1 day, (b) 3 days, (c) 7 days, and (d) 14 days. On day 1 (a), the water
population and the drug population with the longest T2 time, ∼10−3 s, overlap and appear as just one population. By day 3 (b), the relaxation time of
the water population increased, and the water population appears as the third peak with the longest T2 time at 5.6 × 10
−3 s in the T2 distribution.
The water peak continues to shift toward longer relaxation times, 7.7 × 10−3 s on day 7 and 2.3 × 10−2 s on day 14, as well as increasing in signal
amplitude.
(4) Vallet-Regi, M.; et al. A new property of MCM-41: Drug delivery
system. Chem. Mater. 2001, 13 (2), 308−311.
(5) Xu, Z. P.; et al. Inorganic nanoparticles as carriers for efficient
cellular delivery. Chem. Eng. Sci. 2006, 61 (3), 1027−1040.
(6) Vallet-Regi, M. Ordered mesoporous materials in the context of
drug delivery systems and bone tissue engineering. Chem. - Eur. J.
2006, 12 (23), 5934−5943.
(7) Skotnicki, M.; et al. Characterization of Two Distinct Amorphous
Forms of Valsartan by Solid-State NMR. Mol. Pharmaceutics 2016, 13
(1), 211−222.
(8) Pham, T. N.; et al. Analysis of Amorphous Solid Dispersions
Using 2D Solid-State NMR and H-1 T-1 Relaxation Measurements.
Mol. Pharmaceutics 2010, 7 (5), 1667−1691.
(9) Dempah, K. E.; Lubach, J. W.; Munson, E. J. Characterization of
the Particle Size and Polydispersity of Dicumarol Using Solid-State
NMR Spectroscopy. Mol. Pharmaceutics 2017, 14 (3), 856−865.
(10) Yuan, X. D.; Sperger, D.; Munson, E. J. Investigating Miscibility
and Molecular Mobility of Nifedipine-PVP Amorphous Solid
Dispersions Using Solid-State NMR Spectroscopy. Mol. Pharmaceutics
2014, 11 (1), 329−337.
(11) Callaghan, P. T. Translational Dynamics & Magnetic Resonance:
Principles of Pulsed Gradient Spin Echo NMR; Oxford University Press:
New York, 2011.
(12) Blümich, B. NMR Imaging of Materials; Clarendon Press:
Oxford, 2000.
(13) Buda, A.; Demco, D. E.; Jagadeesh, B.; Blümich, B. Molecular
dynamic heterogeneity of confined lipid films by H-1 magnetization-
exchange nuclear magnetic resonance. J. Chem. Phys. 2005, 122 (3),
034701.
(14) Fujara, F.; et al. Translational and rotational diffusion in
supercooled orthoterphenyl close to the glass transition. Z. Phys. B:
Condens. Matter 1992, 88 (2), 195−204.
(15) Mitchell, J.; Webber, J. B. W.; Strange, J. H. Nuclear magnetic
resonance cryoporometry. Phys. Rep. 2008, 461 (1), 1−36.
(16) Petrov, O. V.; Furo, I. NMR cryoporometry: Principles,
applications and potential. Prog. Nucl. Magn. Reson. Spectrosc. 2009, 54
(2), 97−122.
(17) Abragam, A. Principles of Nuclear Magnetism; Oxford University
Press: USA, 1983.
(18) Kubo, R. A Stochastic Theory of Line Shape and Relaxation. In
Fluctuation, Relaxation, and Resonance in Magnetic Systems; Ter Haar,
D., Ed.; Oliver and Boyd Ltd: Edinburgh, 1962; pp 23−68.
(19) Washburn, K. E.; et al. Simultaneous Gaussian and exponential
inversion for improved analysis of shales by NMR relaxometry. J.
Magn. Reson. 2015, 250, 7−16.
(20) Santyr, G. E.; Henkelman, R. M.; Bronskill, M. J. Variation in
measured transverse relaxation in tissue resulting from spin locking
with the CPMG sequence. J. Magn. Reson. 1988, 79 (1), 28−44.
(21) Song, Y. Q.; et al. T-1-T-2 correlation spectra obtained using a
fast two-dimensional Laplace inversion. J. Magn. Reson. 2002, 154 (2),
261−268.
(22) Heinz, A.; et al. Understanding the solid-state forms of
fenofibrate - A spectroscopic and computational study. Eur. J. Pharm.
Biopharm. 2009, 71 (1), 100−108.
(23) Amstad, E.; Spaepen, F.; Weitz, D. A. Crystallization of
undercooled liquid fenofibrate. Phys. Chem. Chem. Phys. 2015, 17 (44),
30158−30161.
(24) Szklarz, G.; et al. Crystallization of supercooled fenofibrate
studied at ambient and elevated pressures. Phys. Chem. Chem. Phys.
2017, 19 (15), 9879−9888.
(25) Dwyer, L. M.; et al. Confined crystallization of fenofibrate in
nanoporous silica. CrystEngComm 2015, 17 (41), 7922−7929.