Micropipette manipulation studies: Material characterization of multiphase, multicomponent systems

  • Utoft, Anders (Projektdeltager)
  • Needham, David (Vejleder)
  • Kinoshita, Koji (Bivejleder)

Projekter: ProjektPh.d-projekt



This thesis presents a study and analysis of the material properties and
phase behavior of liquid-in-liquid systems. Both pure liquids and solutions
containing solutes such as drug, salt, oil, protein and antibody have been
investigated. Chapter 2 presents a thorough description of the micropipette
manipulation technique used as the primary experimental method
throughout this thesis. Chapter 3 presents the theory associated with liquid
microdroplet dissolution and some of the considerations that have been
taken into account when applying the Epstein-Plesset model. Diffusion
coefficients of octanol-in-water, water-in-octanol, water-in-decane, DCM-in-water, and Ibp-in-water are presented. Chapter 4 focuses on NaCl as a solute
and presents a novel triple micropipette technique for supersaturation, 10.3
± 0.3 M, and saturation, 5.5 ± 0.1 M measurements. It is shown that the
extended Epstein-Plesset model successfully predicts diffusion controlled
microdroplet growth and that induction time and surrounding medium affects
crystal morphology. Chapter 5 presents Triolein parameters in an ethanolwater-
lipid system. Solubility, 31 mg/ml in ethanol, diffusion coefficient in ethanol, and interfacial tension, 31 mN/m vs water and 1 mN/m vs ethanol are presented. Effect on interfacial tension of POPC addition was measured, 1.7 mN/m vs water, and parameters were used to predict the critical nucleation sizes for ‘trapped’ Triolein nanoparticles, 20 nm. Chapter 6 presents micropipette investigations of protein, sugar and antibody systems as part of a commercial scale-up for protein and antibody microglassification formulation. Final concentrations of bovine serum albumin, 1143 ± 82 mg/ml, and bovine gamma globulin, 1388 ± 177 mg/ml, in the microglassified particles were measured. Bulk preparation of bovine gamma globulin were made and scanning electron
microscopy was used to compare the different microparticles.


The work described in this thesis have investigated the method of single
particle studies using the micropipette manipulation technique described in
chapter 2. The work in chapter 3 addressed the baseline systems of octanol-in-
water and water-in-octanol to obtain diffusion coefficients for further studies in chapter 4. The importance of the approach to steady-state previously discussed by Duncan was then investigated by comparing these two systems. For a system such as octanol-in-water, the long dimensionless time for total dissolution thus made the transient term neglectable, since the 95 % steady-state was reached
within the first 10 % of the total dissolution process. However, for the waterin-
octanol system, the 95 % steady-state was not reached until 90 % of the total dissolution process had occurred. Thus, the full EP model, must be used for the water-in-octanol system, while the simplified model, can be used for the octanol-in-water system.
DCM microdroplet dissolution was unaffected by the presence of SDS. Lastly, we measured the diffusion coefficient of Ibp dissolving into water, following the relatively fast loss of DCM into water from a DCM-Ibp microdroplet.
Chapter 4 showed how earlier modifications to the EP model can be ignored
if certain experimental procedures are followed, thus decreasing the
difficulty and time required for data processing. It was shown the new ‘two
pipette transfer method’ developed by Kinoshita allowed for visualization
from t = 0, during micropipette studies without losing sight of the
microdroplet at any point in the experiment. Based on this two pipette technique, a novel ‘triple pipette method’ was developed and was shown to
have great possibilities for investigating solubility of liquids or solids such as
NaCl crystals in a second liquid while using very low amounts of material
(nano gram range). This novel technique combined the measurement of
supersaturation of NaCl in water dissolving into an octanol system, 10.3 ±
0.3 M and saturation of NaCl in water using nano grams of material, 5.5 ±
0.1 M, in one single experimental setup. The extended EP model of
Bitterfield was successfully applied to diffusion controlled droplet
growth. When NaCl was present inside a water microdroplet at a
concentration where it influenced the microdroplet water activity to a degree
where it was lower than the water activity in the surrounding octanol
environment, the microdroplet underwent diffusion controlled growth by
imbibing water from the saturated octanol phase. Thus, resulting in a
lowering of the NaCl concentration inside the microdroplet, which in turn
increased the water activity inside the microdroplet and slowed down the
growth. The effect on water dissolution rate, rate of NaCl concentration
change, crystal structure and the timeframe of initial crystal growth by
changing the bathing medium from octanol to decane was investigated. The
slower dissolution in decane resulted in longer induction time, thus giving
the NaCl more chance to nucleate. The crystal structure is very different
between NaCl solution droplets dissolving in octanol or decane. Several
ideas as to the underlying factors of these observations have been
In chapter 5 we measured the solubility value of Triolein in ethanol to be 31
mg/ml. We then measured the diffusion coefficient of Triolein in pure ethanol
based on this solubility. These values were not previously found in literature. Assuming the value of the diffusion coefficient to remain unchanged upon addition of small amounts of water, we then measured the solubility of Triolein in 95 – 80 V/V %. Using these measured values, we then calculated the solubility of Triolein across the entire binary water-ethanol range using the log-linear model of Yalkowski for hydrophobic materials. The interfacial tension values between Triolein and water-ethanol mixtures were then measured, using a new
micropipette setup to avoid system drift due to loss of ethanol. The interfacial tension between Triolein and the binary mixture of water-ethanol was found to decreases from 31 mN/m against pure water to 1 mN/m against pure ethanol. The interfacial tension measurements containing 1 mM POPC showed a very low interfacial tension for the Triolein-water interface of 1.7 mN/m. Thus, indicating that a POPC monolayer is formed at the oil-water interface. Increasing the ethanol concentration, further reduced the interfacial tension from 1.7 mN/m to 0.6 mN/m at 31.6 mol % ethanol. At 41.9 mol % ethanol the interfacial tension rose to 5 mN/m, indicating that we passed the point where 1 mM POPC was soluble in the binary mixture. Thus, POPC was soluble in the mixture above 42 mol % ethanol and the interfacial tension was dominated by ethanol adsorption. We then observed and measured the supersaturation of Triolein in an ethanolic microdroplet dissolving into hexadecane. After a very short (seconds) ethanol dissolution and slight up concentration of Triolein, we observed Triolein nucleation at a supersaturation fraction of S < 1.1. In the same manner, a supersaturation fraction of S < 1.1 was also measured from a Triolein-saturated 95 % ethanol + 5 % water V/V microdroplet dissolving into hexadecane. The
supersaturation fraction measured for a diffusion controlled microdroplet
system was lower than the value of S = 1.2 - 1.4 obtained for the solvent
shifting method by Walke. Using the measured parameters of solubility,
supersaturation and interfacial tension in classical nucleation theory, eq. 68,
showed that the predicted critical nucleation size matched with experimental
data from the solvent shifting method. These measurements were achieved
by using POPC to ‘trap’ NPs at their nucleation size.
In chapter 6 the previous work of Bitterfield and Aniket on BSA
solutions was repeated and confirmed to get familiar with protein solutions
and how they behave in a micropipette setup. The BSA concentration in the
final glassified microparticle was measured to be 1143 ± 82 mg/ml, which is
in exceptional agreement with the values of Aniket et al. of 1147 ± 32
mg/ml. We then performed microdroplet dissolution experiments with
sucrose and measured the final concentration to be 1252 ± 45 mg/ml, which
was slightly lower than expected from the density of sucrose. Surprisingly
the sucrose never precipitated or solidified, so the final microparticle was
actually still liquid, although with a viscosity so high that it could no longer
be pulled into the micropipette. BGG was then used as a cheap antibody
test molecule in micropipette experiments, which showed that a ‘skin’
formed during microdroplet dissolution, but that the resulting non-spherical
microparticle still lost water through this skin until an average BGG
concentration of 1388 ± 177 mg/ml was reached. 1:1 w/w BGG:sucrose
solutions were then used in micropipette dissolution experiments and
revealed no significant difference on neither the ‘skin’ formation nor on the
shape and surface of the final microparticles. SEM imaging was used to
compare the surface structure of the microparticles and these also showed
no significant difference. Thus, showing that sucrose can be incorporated
into solidified or microglassified protein and antibody formulations. Bulk
preparation of BGG microparticles were then performed using a single
injection during vortexing or by using a homogenizer. Microscopy images
showed that the single injection method produced very heterogeneous
microparticles regarding both size and surface smoothness. However, the
homogenization method proved to be far superior and produced relatively
homogenous microparticles regarding both size and surface smoothness.
Thus, scale-up of the microglassification process actually showed an
improvement in the antibody formulation as compared to the microparticles
formed during micropipette experiments. The micropipette was, however,
still important for characterizing the system and informing the scale-up
process with parameter values such as microparticle concentration, level of
dehydration, solubility and time of dissolution and microglassification for
individual microparticles.
Kort titelMicropipette manipulation studies
Effektiv start/slut dato02/09/201331/07/2017


Udforsk forskningsemnerne, som dette projekt berører. Disse etiketter er oprettet på grundlag af de underliggende bevillinger/legater. Sammen danner de et unikt fingerprint.