Engineering a nanoscale oil-in-water carrier to rescue a poorly water-soluble lipophilic drug — from crystal to characterized nano-delivery system
An estimated 40% of marketed drugs — and up to 90% of new chemical entities in discovery pipelines — fall into Biopharmaceutics Classification System (BCS) Class II (low solubility, high permeability) or Class IV (low solubility, low permeability). For these compounds, the bottleneck to absorption is not crossing the intestinal membrane, it is simply getting enough drug into solution in the gut lumen in the first place.
Dissolution rate is described by the Noyes-Whitney equation:
dC/dt = (D · A / h) × (Cs − C)
Where D is the diffusion coefficient, A is the total surface area of drug exposed to solvent, h is the diffusion boundary layer thickness, Cs is the drug's saturation solubility, and C is the bulk concentration already dissolved.
For a lipophilic BCS Class II compound, Cs is intrinsically low (often micrograms per milliliter), so the driving-force term (Cs − C) is small from the outset. Worse, a macroscopic crystal or coarse powder presents a small surface area A relative to its mass — most of the drug is locked deep inside the crystal lattice, inaccessible to solvent. The result: dissolution proceeds too slowly to keep pace with gastrointestinal transit time (~2–4 hours), and a large fraction of the dose is excreted unabsorbed, undissolved.
The Noyes-Whitney equation contains the entire rationale for nanoemulsion formulation in one variable: A. Every downstream stage of this process — oil selection, emulsification, homogenization — exists to convert a low-surface-area crystal into billions of nanoscale droplets, multiplying total interfacial area by orders of magnitude.
It is tempting to assume that a lipophilic drug (high logP) should cross the lipid bilayer of enterocytes easily — and it does, once dissolved. BCS Class II compounds are defined precisely by this combination: high intestinal permeability paired with low aqueous solubility. The membrane is not the obstacle; the aqueous diffusion layer immediately adjacent to it is.
A drug molecule must first partition out of its solid form into the surrounding fluid before it can diffuse to, and permeate across, the intestinal epithelium. If dissolution is the slow step, no amount of membrane permeability can compensate — the drug simply is never presented to the membrane in molecular (dissolved) form at a clinically useful rate. This is why formulation scientists describe such compounds as having "dissolution-rate-limited absorption," and why solubility-enabling formulations (nanoemulsions, solid dispersions, cyclodextrin complexes, lipid-based systems) are such an active area of pharmaceutical development.
A nanoemulsion sidesteps the solubility problem rather than solving it directly: the drug is never required to dissolve in water at all. Instead, it is pre-dissolved in a biocompatible oil phase at high concentration, and that oil is then dispersed as billions of nanoscale droplets throughout the aqueous vehicle. Each droplet is a self-contained reservoir of dissolved drug with saturation solubility in its own oil microenvironment — vastly higher than the drug's solubility in water.
Upon administration, these nanodroplets present enormous cumulative interfacial surface area to the gut lumen, and lipid digestion/processing pathways (mixed micelle formation, chylomicron uptake, lymphatic transport) provide additional absorption-enhancing mechanisms unavailable to a simple aqueous suspension. The remainder of this walkthrough follows the engineering path from oil selection through droplet fracture to final characterization.
Before any shear is applied, the formulator must choose two chemical systems that will define everything downstream: an oil phase capable of dissolving the drug at high loading, and a surfactant / co-surfactant blend whose hydrophile-lipophile balance (HLB) is matched to stabilize an oil-in-water interface rather than a water-in-oil one.
The ideal oil for a nanoemulsion drug carrier must satisfy several simultaneous criteria:
• High drug solubilizing capacity: the oil should dissolve the target dose in a small oil volume, minimizing the total dose volume patients must take • GRAS / pharmaceutical-grade status: medium-chain triglycerides (MCT, e.g. Miglyol 812), long-chain triglycerides, oleic acid, ethyl oleate, and isopropyl myristate are common choices with established safety profiles • Digestibility: medium-chain triglycerides are rapidly hydrolyzed by pancreatic lipase, releasing free fatty acids and monoglycerides that promote mixed micelle formation and lymphatic uptake — actively assisting absorption rather than being inert • Compatibility with the chosen surfactant: the oil-surfactant pair must produce a low, stable interfacial tension
A solubility screening step (shake-flask equilibrium solubility in a panel of candidate oils) is performed early in development to rank oils by drug-loading capacity — often spanning a 50- to 100-fold range between the best and worst candidates for a given molecule.
The hydrophile-lipophile balance (HLB) scale (Griffin, 1949) ranks surfactants from 0 (fully lipophilic) to 20 (fully hydrophilic) based on the relative size and strength of their hydrophilic head versus their lipophilic tail. Emulsion type follows a simple rule of thumb:
• HLB 3–6 → favors water-in-oil (W/O) emulsions • HLB 8–18 → favors oil-in-water (O/W) emulsions — the target for an orally administered nanoemulsion • HLB 15–40 → solubilizing / detergent action, often too hydrophilic alone
Single surfactants rarely hit the required HLB exactly, so formulators blend a low-HLB and a high-HLB surfactant (or add a small-molecule co-surfactant such as ethanol or propylene glycol) to reach the target HLB by the additive mixing rule: HLBmix = f1·HLB1 + f2·HLB2. The co-surfactant additionally increases interfacial fluidity, reducing the energy required to bend the interface into nanoscale curvature — a key enabler for very small droplet sizes during emulsification.
Required HLB is not a fixed universal number — it is specific to each oil. A given oil (e.g. MCT) has an empirically determined "required HLB" for O/W stability, typically in the 10–12 range, meaning the surfactant blend must be titrated against that specific oil, not looked up from a generic table.
| Product | Indication | Trial Design | Key Result |
|---|---|---|---|
| Span 80 (sorbitan monooleate) | HLB ≈ 4.3 | Lipophilic; W/O stabilizer or co-surfactant | Lowers overall blend HLB, increases interfacial fluidity |
| Tween 80 (polysorbate 80) | HLB ≈ 15.0 | Hydrophilic; classic O/W nonionic surfactant | Strong steric stabilization, widely used, GRAS |
| Poloxamer 188 | HLB ≈ 29 | Triblock copolymer, strong steric brush | Excellent long-term stability, low toxicity |
| Lecithin (phospholipid) | HLB ≈ 4–10 | Natural amphiphile, forms flexible monolayer | Biocompatible, often paired with a co-surfactant |
With the oil and aqueous phases chosen, a coarse pre-emulsion is formed by simple high-shear mixing (rotor-stator), then driven through a high-energy process that supplies enough mechanical energy to overcome the Laplace pressure resisting droplet breakup, fracturing micron-scale oil globules into a swarm of nanoscale droplets.
In a high-pressure homogenizer, the coarse pre-emulsion is forced under high pressure (typically 500–1500 bar) through a narrow valve gap only micrometers wide. Three simultaneous physical mechanisms fracture droplets as they transit the gap:
• Intense shear stress: the velocity gradient across the narrow gap stretches and elongates droplets far beyond their stable radius, causing them to break into smaller fragments • Turbulent eddies: at the high Reynolds numbers generated, chaotic turbulent flow imposes fluctuating stresses that further deform and split droplets • Cavitation: as fluid accelerates through the gap, local pressure drops below vapor pressure, nucleating vapor microbubbles; their violent collapse immediately downstream releases intense local shock waves that shatter nearby droplets
Each droplet must pass through the gap fast enough (residence time on the order of microseconds) that only a single homogenization event occurs per pass — hence the need for multiple passes to progressively refine the distribution (covered in Stage 4).
Probe ultrasonication achieves droplet breakup through acoustic cavitation: a piezoelectric transducer drives a titanium probe tip to vibrate at 20–24 kHz, generating alternating high- and low-pressure acoustic waves in the liquid. During the low-pressure (rarefaction) half-cycle, dissolved gas nucleates into microbubbles; during the high-pressure (compression) half-cycle these bubbles implosively collapse, generating localized shock waves, high-velocity microjets, and transient temperatures approaching 5000 K in nanoscale volumes.
These cavitational shock events physically tear oil droplets apart at the microscale. Ultrasonication is favored at bench/lab scale for its simplicity and rapid method development, while high-pressure homogenization dominates at pilot and manufacturing scale for its superior throughput, reproducibility, and validated GMP scale-up pathways.
Both methods are classified as "high-energy" emulsification, contrasted with "low-energy" methods such as spontaneous emulsification (drug/oil/surfactant phase injected into water, spontaneously forming nanodroplets via rapid solvent diffusion) or phase inversion temperature/composition methods, which exploit surfactant curvature changes rather than brute-force mechanical fracture. Low-energy methods use far less equipment energy but offer less control over final droplet size and are more sensitive to formulation composition.
A single pass through a homogenizer rarely achieves target nanoscale size; instead, the coarse emulsion is recirculated through the high-shear zone multiple times, with each pass fracturing the existing droplet population further while surfactant races to coat every newly created interface before droplets can recombine.
Droplet size reduction is a competition between two opposing processes happening simultaneously inside the high-shear zone: droplet breakup (splitting large droplets into smaller ones) and droplet coalescence (two droplets colliding and re-merging). Each pass through the homogenizer shifts the equilibrium toward smaller mean size, but only if surfactant molecules can adsorb onto newly created oil-water interfaces fast enough to lower interfacial tension and provide a stabilizing film before a freshly split droplet collides with a neighbor and re-merges.
With each successive pass, the population of large droplets shrinks and a growing population of small droplets accumulates; the size distribution narrows (PDI falls) as the process converges toward a steady-state size set by the balance between the maximum shear stress deliverable at that pressure and the Laplace pressure resisting further breakup of increasingly small droplets. Beyond roughly 8–10 passes, further size reduction becomes marginal — the process approaches its size floor for a given pressure and surfactant system, and additional passes mainly risk fluid heating and localized surfactant desorption.
As each droplet is fractured, its newly exposed oil-water interface is initially bare and thermodynamically unstable — high interfacial tension drives it to minimize surface area by recombining with neighboring droplets. Surfactant molecules, present in molecular or micellar form throughout the aqueous phase, diffuse to and adsorb onto this new interface, orienting their hydrophobic tail into the oil and hydrophilic head into water.
Once adsorbed, the surfactant monolayer provides one or both of two stabilization mechanisms:
• Steric stabilization: bulky hydrophilic head groups (e.g. polysorbate PEG chains, poloxamer PEO blocks) physically project into the aqueous phase, creating a repulsive barrier when two droplets approach — their surfactant "brushes" resist interpenetration • Electrostatic stabilization: charged surfactants (e.g. anionic phospholipids, ionic surfactants) impart a surface charge (zeta potential) to droplets; like-charged droplets repel each other before contact is even possible
A sufficient surfactant:oil ratio ensures full interfacial coverage at the final, much larger total surface area generated by nanoscale droplets — insufficient surfactant leaves patches of bare interface vulnerable to coalescence, producing a broader, less stable distribution (higher PDI).
Total interfacial area scales inversely with droplet radius for a fixed oil volume fraction. Shrinking mean droplet diameter from 2 µm to 150 nm increases total interfacial area roughly 13-fold — the surfactant demand at the end of processing is far higher than at the start, which is why the surfactant:oil ratio chosen in Stage 2 directly limits how small and how monodisperse the final nanoemulsion can become.
The finished nanoemulsion is characterized by dynamic light scattering (DLS) to confirm droplet size and distribution width, and its performance is validated by comparing dissolution and absorption against the original free drug crystal — closing the loop back to the solubility problem that motivated the entire formulation effort.
DLS measures the time-dependent fluctuation of scattered laser light intensity caused by Brownian motion of droplets in suspension. Smaller droplets diffuse faster, producing more rapid intensity fluctuations; larger droplets diffuse more slowly. Autocorrelation analysis of these fluctuations yields the translational diffusion coefficient, which the Stokes-Einstein equation converts into a hydrodynamic droplet diameter:
D = kT / (3πηd)
Where D is the diffusion coefficient, k is Boltzmann's constant, T is temperature, η is solvent viscosity, and d is the hydrodynamic diameter.
The polydispersity index (PDI), derived from the width of the correlation function decay, quantifies distribution breadth on a 0–1 scale: PDI <0.1 indicates a highly monodisperse population, 0.1–0.2 is generally considered acceptable for pharmaceutical nanoemulsions, and PDI >0.3 indicates a broad or potentially multimodal distribution that may signal incomplete processing or emerging instability.
Recall the Noyes-Whitney relationship from Stage 1: dissolution rate scales directly with available surface area A. A nanoemulsion with 150 nm mean droplet diameter presents on the order of 10,000-fold more interfacial surface area per unit oil volume than a 1 mm drug crystal fragment. Even though the drug is already dissolved within the oil droplets (rather than needing to dissolve from a solid), the same surface-area logic governs the rate at which lipid digestion products and dissolved drug partition out of the droplet into mixed micelles for absorption — a small, high-surface-area droplet completes this partitioning far faster than a large one.
In vivo, this translates into faster, more complete, and less variable drug absorption. Published nanoemulsion formulations for BCS Class II compounds commonly report 2- to 8-fold increases in oral bioavailability (AUC) relative to the crystalline drug administered as a simple suspension, alongside reduced food-effect variability, because lipid-based systems partially bypass the need for bile-salt-mediated solubilization that dietary state would otherwise control.
A freshly manufactured nanoemulsion meeting size and PDI targets is not automatically shelf-stable. Two degradation mechanisms threaten nanoemulsions specifically because of their small size:
• Ostwald ripening: oil has finite (if very low) aqueous solubility. Smaller droplets, having higher Laplace (curvature) pressure, exhibit slightly higher local solubility than larger droplets. Oil molecules therefore diffuse away from small droplets and deposit onto larger ones through the aqueous phase over time, causing the mean droplet size to slowly but steadily increase — potentially over weeks to months. Adding a small fraction of a second, water-insoluble oil component ("ripening inhibitor," e.g. a small amount of a very high molecular weight or highly lipophilic co-oil) suppresses this mechanism by creating an osmotic penalty that opposes shrinkage of small droplets.
• Coalescence: if the surfactant film is incomplete, depleted, or disrupted (e.g. by temperature excursions, freeze-thaw cycling, or dilution below critical micelle concentration), droplets that collide may merge permanently, driving the distribution toward larger sizes and eventually visible phase separation.
Accelerated stability studies (elevated temperature storage, freeze-thaw cycling, centrifugation stress tests) combined with periodic DLS monitoring are used to establish shelf life and confirm the surfactant system chosen in Stage 2 remains protective for the intended product lifetime.
A nanoemulsion is a kinetically stabilized, not thermodynamically stable, system — unlike a microemulsion, which forms spontaneously and is thermodynamically stable at a given composition and temperature. The nanoemulsion's small droplet size and adsorbed surfactant film create a high kinetic barrier to phase separation, but given enough time, ripening and coalescence will eventually drive it toward the lower-energy, phase-separated state. Formulation and packaging must manage that timescale to exceed the product's required shelf life.