Monitoring secondary (solute+water) crystallization by DSC, synchrotron X-ray diffraction, and in
vials using heat flux transducer
Qiming Wang (Millrock Technology Inc.), Michael Sztucki, T Narayanan (European Synchrotron Radiation Facility), Brian Ivin, Adrian
Marley, Kieran Joyce, Prem Mohanty, Evgenyi Shalaev (Allergan)
40
30
900
ice melt
-11 oC
20
<NaCl*2H2O>+ice melt
0
300
Ice cryst
0
-10
-300
-20
-600
-30
-40
-50
0
<NaCl * 2H2O> + solution
-10
ice + solution
-20
<NaCl> + <NaCl * 2H2O>
ice + <NaCl * 2H2O>
4-17-2016, run 3
-40
-1200
-50
-60
0
Fig 1. DSC cooling/heating curves of 10 wt% (left) and 0.17 wt% (right)
poloxamer solutions. Solute+water crystallization observed during
warming of 10 wt% solution between -45 and -35C. In 0.63 % (not
shown) and 0.17 % solution, crystallization event was not detected either
during cooling or warming.
10
-30
-900
Heat Flux
TP7
TSHELF_Surface
NaCl + solution
20
-21 oC
<NaCl*H2O>+ice cryst
Temperature, oC
10
solution
30
600
Heat flux, W/m2
Crystallization of excipients during freeze-drying could have a major impact on both the
manufacturing process (e.g., rate of drying) and quality of the finished product. In this
investigation, binary solutions of water+NaCl and water+surfactant (poloxamer) were
studied by DSC, low-temperature synchrotron small- and wide-angle X-ray diffraction
(XRD), and heat-flux transducers (HFT) which are incorporated into the shelf of a
Millrock Technology, Inc. MicroFD freeze-dryer with LyoPAT II. In the HFT experiments,
the heat flux between the shelf and vials was measured during cooling/warming cycles.
For NaCl-water solutions, primary ice nucleation and secondary solute+water
crystallization (equivalent to eutectic crystallization for a binary system) during cooling,
and secondary (eutectic) solute+ice and ice melting during warming were detected by
HFT. Behavior of poloxamer-water system was more complex, with at least three
consecutive transitions taking place during cooling, i.e., primary water crystallization,
formation of a liquid crystalline (tentatively, cubic) phase, and finally appearance of a
crystalline poloxamer phase with 3-dimensional translational order. The study confirms
that crystallization behavior as observed by laboratory-based tests (DSC and XRD)
could be different from the behavior of a formulation during real freeze-drying runs, and
also presents HFT as a useful non-invasive tool to monitor excipients crystallization in
vials.
RESULTS: HFT (NaCl 15%)
RESULTS: DSC
Temperature, oC
ABSTRACT
200
400
0
600
10
20
30
40
50
60
70
80
90
100
NaCl, wt%
Time, min
Fig 3. Temperature and HFT data for cooling/heating cycle of 15 % NaCl solution (left) and
phase diagram of NaCl-water system (right). Cooling: both primary ice formation and
secondary NaCl*2H2O+water crystallization were detected by HFT.
Warming: both eutectic melt and ice melt were detected by HFT. Transition temperatures by
HFT during warming (Te = -21C and TL = -11C) are close to that from the phase diagram (Te =
-21.4C and TL = -13.1C)
RESULTS: sSAXS/WAXS
RESULTS: HFT (NaCl 5%)
1600
10000
1400
40
1000
100
0
5
10
15
20
25
0
30
1
2
3
4
5
q[nm-1]
q[nm-1]
Fig 2a. SAXS (right) and WAXS (left) patterns of P188 obtained at room temperature.
Ratio of the peak positions in the SAXS pattern is 1.82, close to the characteristic ratio of
hexagonal phase of 1.73.
10
<NaCl*2H2O>+ice
0
crystallization
0
-100
-10
-20
-200
-30
Shelf_inlet
TP1
HEAT_FLUX
-40
-50
-60
-400
100
200
300
400
500
600
700
800
Time, min
250
500
P188
RT
10%, 0C
10%, -5C
10%, -10C
10%, -15C
10%, -20C
200
Scattering, 1/mm
300
-20C
10%, -25C
10%, -30C
10%, -35C
10%, -40C
10%, -50C
250
200
ice
150
RT
10%, 0C
10%, -5C
10%, -10C
10%, -15C
10%, -20C
10%, -25C
10%, -30C
10%, -35C
10%, -40C
10%, -50C
Scattering, 1/mm
350
150
RESULTS: HFT (P188, 10% and 5 %)
100
100
50
50
21
12
26
14
16
18
20
0.5
1.0
q[nm-1]
q[nm-1]
1.5
2.0
2.5
3.0
3.5
q[nm-1]
300
Fig 2b. Low-temperature synchrotron SAXS and WAXS of 10 wt% poloxamer solutions
obtained during cooling. Ice formation was observed between -15 and -20C (WAXS
patterns, left), whereas crystalline peaks of P188 (3-d structure) appeared at a lower
temperature between -25 and -30C (WAXS patterns, center) center. Development of longrange structure was detected by SAXS in the same temperature range as ice formation, at 15 to -20C (right), i.e., before the formation of the 3-d structure. Note that pure “as is” P188
has hexagonal structure whereas the frozen solution corresponds to cubic structure..
400
20
10
200
ice nucleation
0
0
-10
-200
-20
-400
P188+ice melt
-30
-600
-40
-800
Shelf T
HFT
TP7
2-27-2016
200
400
600
-1000
800
Time, min
Scattering intensity, 1/mm
200
150
100
13
14
15
16
q[nm-1]
17
18
0
100
0
P188+ice crystalization?
-10
-100
-20
-200
-30
-300
-40
-400
-50
-60
-500
0
100
200
300
400
500
600
700
800
Time, min
Fig. 5. Temperature and HFT data for cooling/heating cycle of 10 % (left) and 5% (right) P188
solutions. P188+ice melting during warming was detected in both cases, with temperature from HFT
in agreement with DSC and XRD measurements. Secondary crystallization, however, was not
detected by HFT
-50C
-25C
-20C
-15C
10%, -10C heating
SAXS
Scattering intensity, 1/mm
-50C
-25C
-20C
-15C
poloxamer
P188+ice melt
-17.8 oC
ice nucleation
10
0
250
200
20
-50
1000
30
Heat flux, W/m2
16
Temperature, oC
11
Heat flux, W/m2
6
Temperature, oC
Scattering, 1/mm
-300
Fig 4. Temperature and HFT data for
cooling/heating cycle of 5 % NaCl solution (7
vials). Both secondary crystallization during
cooling and secondary melt during heating
were detected by HFT, indicatively sufficient
sensitivity. Multiple secondary crystallization
events are detected, showing that controlled
ice nucleation does not necessarily assure
consistent secondary crystallization
-500
0
400
Poloxamer-water solutions with poloxamer concentration 10 wt % were studied by
both DSC and low-temperature synchrotron small- and wide-angle X-ray scattering
(WAXS/SAXS), and with poloxamer concentration of 0.63 and 0.17 wt% by DSC. The
SAXS/WAXS experiments were performed at the European Synchrotron Radiation
Facility (ESRF), Grenoble, France.
100
Heat flux, W/m2
.
400
METHODS
Binary model solutions water+NaCl (5 and 15 wt% solute) and water+poloxamer (5
and 10 wt% solute) were filtered using 0.22 filter, filled into glass vials, and loaded into
Millrock Technology, Inc. MicroFD freeze-dryer (photo below), on top of the HFT.
Either one or seven vials were used in different tests; a photo of experimental set-up
with 7 vials is given below. Product temperature was monitored by thermocouples
during freeze/thaw cycles. Controlled ice nucleation was used in the HFT
experiments. Total of 13 runs were performed.
<NaCl*2H2O>+ice melt
ice nucleation
20
600
200
Evaluate feasibility of using heat flux transducers to detect secondary solute+water
crystallization during freeze-drying.
200
30
1000
800
Temperature, oC
PURPOSE
Scattering, 1/mm
1200
100
30
1
2
3
q[nm-1]
Fig 2c. Low-temperature synchrotron SAXS and WAXS of 10 wt% poloxamer solutions
obtained during warming. The warming data shows that P188 melting proceeds in two
steps, first disappearance of 3-d positional order (WAXS, center, -20 to -15C), and then
long-range positional 2-d order (SAXS, -15 to -10C).
CONCLUSIONS
SAXS/WAXS: Crystallization/melting of P188 proceeds in two steps. During cooling, 2-d cubic
structure forms followed by development of 3-d order, and the sequence of the transitions is
reversed during warming, when the 3-d structure melts at lower temperature while 2-d cubic
structure is still intact.
HFT: Secondary crystallization and melting was detected in NaCl-water solutions. Multiple
secondary crystallization events are detected during cooling, demonstrating that controlled ice
nucleation does not necessarily result in a consistent secondary crystallization. For P188-water
system, while secondary melt is observed during warming, secondary crystallization is not
detected even in the most concentrated solution. The behavior of P188-water system appears to
be more complicated inside the vial and requires further investigation. Note also that, according to
DSC, P188 crystallization was observed during warming of the frozen solution, whereas
SAXS/WAXS detected crystallization of P188 during cooling.