Jeff James reports on research carried out at Nottingham University to establish the effects and mechanics of moisture ingress on suspension formulations for metered dose inhalers
The effect of moisture ingress through seals and valves on suspension formulations for metered dose inhalers has routinely been assessed via storage stability trials conducted at high humidity, and general testing using such techniques as the Anderson Cascade Impactor.
However, while this can show the effect of moisture ingress on characteristics such as the fine particle fraction, it does not answer fundamental questions as to how the water elicits such effects.
Research was conducted recently by the author at Nottingham University under the guidance of Professor Clive Roberts, and in conjunction with Dr Richard Toon, research specialist at 3M Drug Delivery Systems. The project's aim was to establish the exact nature of the effect of moisture ingress on suspension formulations for metered dose inhalers. The investigations used the colloid probe technique1 of atomic force microscopy (AFM) in a model hydrofluoroalkane (mHFA), deca-fluoropentane.
In the colloid probe technique, a particle of an active pharmaceutical ingredient (API) is attached to the AFM micro-cantilever. This particle is challenged against a chosen surface, such as some canister material or another particle. The adhesive force (Fadh) required to remove the particle from this material is then measured.
There are some fundamental questions with regards to where water will preferentially be located within suspension formulations. There are many unknowns concerning the effect of water on suspension formulations. Such unknowns include the effect of different canister surfaces such as coated surfaces, whether the API is hydrophilic or hydrophobic and the effects of co-solvents, such as ethanol, which can also hydrogen bond.
Ethanol is an important factor to consider in such investigations. It is added to HFAs to increase the solubility of excipients such as surfactants - oleic acid, for example. While water is quite insoluble in pure HFAs, the presence of ethanol allows far greater solubility.1
The investigations consisted of challenging the surfaces of various pMDI components with some chosen pulmonary APIs, in mHFA, in water saturated mHFA with and without ethanol and finally mHFA containing oleic acid and ethanol. The results given in figures 1 and 2 show that on the addition of water, the force of adhesion for the chosen APIs increases with a number of the chosen MDI components.
The force of adhesion in the presence of mHFA containing both water and ethanol is, in a number of cases, little different from the force of adhesion in mHFA containing water alone, even though ethanol is present in far greater excess than water.
hydrogen bonding
A potential hypothesis to explain such observations is that the strong hydrogen bonding capability of water may thermo-dynamically dictate that water will form layers at the surfaces of both the components and APIs (figure 3).2
Therefore, it is likely that in mHFA containing water, the observed increase in the force of adhesion is due to a water layer on the component surface interacting with a water layer on the surface of the API.
Upon mixing 2% (v/v) ethanol to mHFA containing water, a reduction in the force of adhesion between the APIs and surfaces is observed in a number of cases, compared with those observed using mHFA and water alone.
This reduction in the force of adhesion in mHFA in the presence of both ethanol and water could be due to a mixed secondary layer of both water and ethanol (see figure 3). Ethanol has a weaker hydrogen bond than water and is also more soluble in HFAs. As a consequence, it is likely that ethanol will preferentially remain in the bulk, forming dimers in order to aid thermodynamic stability.
It has been reported in the literature that mixtures of HFAs and ethanol do not mix in an ideal fashion and ethanol does indeed preferentially remain in the bulk compared with the interface.2
The differences observed between APIs could be due to their relative hydrophobic nature. While some components may become coated with a layer of water, a hydrophobic API, or component, will have a reduced tendency to be coated with water. A hydrophobic API, or component, may preferentially be coated with ethanol and thus, the interactions are predominantly between a combination of a water/ethanol coating on one surface and mainly ethanol on the hydrophobic surface.
This theory was put to the test by repeating the study using the relatively hydrophobic API mometasone furoate. This showed a marked reversal, with the addition of ethanol having a greater effect than water (figure 4).
This theory is consistent with the measured ranking of contact angles for each active. Of the three APIs, mometasone furoate has the highest water contact angle of 40.8° (1.4), followed by formoterol fumarate 38.0° (0.7), with salbutamol sulphate exhibiting the lowest water contact angle, 33.8° (0.9). This essentially ranks the hydrophobicity of the three actives as follows; mometasone furoate > formoterol fumarate > salbutamol sulphate. It can be seen from figure 4 that while ethanol has, in a number of cases, a greater adhesive influence on mometasone furoate than water, the reverse is true for salbutamol sulphate (figure 1).
surface roughness
A further possible cause for the observed adhesive effects associated with coated and uncoated canisters may be related to a component's surface roughness. It has been well documented that the roughness of a surface has an effect on the adhesion of a particle.3 The scale of roughness of the component surface in relation to that of the contacting API will determine whether the contact area between the surfaces (and hence the force of adhesion) will increase or decrease. The FEP coated canister has a much smoother surface than the uncoated canister as determined by AFM imaging (figure 5).
In simplistic terms, the smoother surface of the FEP coating may allow a greater area of contact for the two adsorbed layers to interact with each other, thus increasing the adhesive force required to pull the API off the surface (assuming the surface roughness of each API particle is similar).
Conversely, the rougher surface of the uncoated canister may reduce the contact area and thus the adhesive interaction. However, the observed results may be caused by a multitude of interacting factors and more investigational work would need to be carried out to determine the major contributing factors (figure 6). It should also be noted that in nearly all cases, oleic acid reduces the adhesive force in the presence of both water and ethanol. Clearly, it appears that a major benefit of surfactant use in pMDI systems is that it acts as a stabiliser against the ingress of water during the lifetime of the unit.
This work is important to our understanding of how suspension formulations behave in the presence of excipients and particularly water. Not only does the work provide the building blocks to understand such processes, it also allows us a rapid method to investigate theories that can then be proven by other methods.