Aluminum-based adjuvants will continue to be a key component of currently approved and next generation vaccines. Their large adsorptive capacity allows the combination with other immunostimulatory molecules to create a toolbox of adjuvants for selective vaccine applications. The widespread use of aluminum adjuvants is due to their excellent safety profile, which has been established through the use of hundreds of millions of doses in humans over many years, low cost, and extensive experience with licensure through regulatory agencies.
New-generation vaccines are being increasingly prepared with highly purified antigens, which improves their safety and tolerability, whilst enabling more simple characterization, but also typically results in decreased immunogenicity. Consequently, adjuvants are added to these vaccines, particularly those comprising recombinant protein subunits, to enhance their ability to induce robust immune responses. Although several new adjuvants have been introduced for specific vaccines over the past 20 years, aluminum adjuvants remain the most commonly used approach, even though they were first discovered nearly 100 years ago 
. The two main types of aluminum adjuvants included in licensed human vaccines are aluminum hydroxide (AH) and aluminum phosphate (AP). The term ‘alum’ is often incorrectly used to refer to aluminum-containing adjuvants but can be an acceptable abbreviation if associated with a clarifying statement. Alum is a chemical solution of potassium aluminum sulfate, [KAl(SO4
O], which is not used as a vaccine adjuvant. While the term ‘alum’ may be used for reasons of simplicity, it is important to define whether this refers to AH or AP, as these adjuvants have very different physical and chemical properties 
. In addition to AH and AP, a product called ImjectTM
Alum (ThermoFisher Scientific, Waltham, MA, USA) is sometimes used in preclinical studies. However, ImjectTM
Alum comprises amorphous aluminum hydroxycarbonate and crystalline magnesium hydroxide 
. Although it has immunostimulatory activity, it appears to be less potent than AH 
and is not used in licensed human vaccines.
The molecular and physical nature of materials used in vaccine adjuvants is very diverse, often for historical reasons. In addition to aluminum salts, adjuvants include oil-in-water emulsions, liposomes, and various natural products, including saponins and Toll-like receptor (TLR) agonists. Consequently, the mechanisms by which these components enhance immune responses are very heterogeneous and often not completely understood, as discussed comprehensively elsewhere 
. In general, adjuvants act only at the site of injection and the draining lymph nodes, while systemic effects should be minimized to reduce the potential for poor tolerability. Adjuvants typically increase the recruitment and activation of innate immune cells, enhance antigen uptake or processing and presentation, increase accumulation of lymphocytes in draining lymph nodes, and can promote the development and persistence of germinal centers, which are necessary for the development of a robust immune response 
. Each individual adjuvant can contribute through some or all of these mechanisms, to a greater or lesser extent. The enhancement of the immune response by adjuvants typically requires coadministration of antigen and adjuvant, although physical association between adjuvants and antigens is often not required. Nevertheless, adsorption of antigens is particularly important for aluminum adjuvants. Hence, the adsorption of antigen to aluminum adjuvants requires the use of specialized analytical techniques. Once the need for an adjuvant in vaccine formulations is concluded, the choice of adjuvant in vaccine formulations is determined by multifactorial considerations and is highly dependent on the targeted disease, the chemical and physical nature of the vaccine antigens, and the population for which the vaccine is intended. Aluminum adjuvants should be the first choice, given their long history of safe and effective use in man, in addition to their wide availability, low cost, and extensive experience of licensure of aluminum-adsorbed vaccines through various regulatory agencies. However, inadequate performance or incompatibility with vaccine antigens when using aluminum adjuvants can necessitate the use of alternative adjuvants such as emulsions or combination adjuvants with TLR agonists.
2. Physicochemical Characterization of Aluminum Adjuvants
A number of established assays are available that can be used to characterize aluminum adjuvants and to ensure consistency between batches 
. Structural information can be obtained using X-ray diffraction (for AH adjuvants only), spectroscopy (Fourier transform infrared, nuclear magnetic resonance, Raman), and transmission electron microscopy 
2.1. Particle Size
Aluminum adjuvants are typically composed of primary nanoparticles that form irregularly shaped aggregates with dimensions of 1 to 20 μm 
. The particle size and shape have important implications for the efficiency of particle uptake by immune cells through phagocytosis 
. However, the reported size of the aggregates varies widely and depends on the methodology used to measure particle size, the tonicity and pH of the dispersion medium, and any dilution factor. The particle size can be determined using laser diffraction, dynamic light scattering, or microflow imaging. The surface charge of aluminum adjuvants is pH-dependent, and larger particle sizes are observed when the pH approaches the point of zero charge as the electrostatic repulsion is decreased 
. Similarly, the addition of sodium chloride can obscure surface charges and enhance particle aggregation 
. However, dilution of aluminum adjuvants in saline caused a decrease in particle size 
2.2. Surface Charge
The aluminum ions at the surface of AH nanoparticles are coordinated with a hydroxyl that can accept or donate a proton depending on the pH of the dispersion medium. As a result, AH has a pH-dependent surface charge. Its point of zero charge (PZC) is 11.4 and it is positively charged at neutral pH 
. In the case of AP, a proportion of the surface hydroxyls is replaced by phosphate because of the higher affinity of aluminum for phosphate. Commercial AP adjuvants have a P:Al ratio of 1.1–1.15:1 and a PZC of around 5, which gives them a negative surface charge at neutral pH 
. A greater proportion of surface hydroxyls results in a higher PZC. Thus, AAHS with a P:Al ratio of 0.3 has a higher PZC and is neutral at pH 7. The surface charge can be determined using a Zetasizer instrument.
2.3. Surface Area
The primary nanoparticles that make up the aluminum aggregates afford the adjuvants a very large surface area, estimated at 514 m2
/g for AH adjuvant, based on water adsorption measured using gravimetric FTIR spectroscopy 
. The surface area can also be determined using the Brunauer–Emmett–Teller (BET) theory following nitrogen adsorption. However, a smaller surface area was reported for AH compared with water adsorption 
, because dehydration of the samples results in agglomeration of particles with loss of surface area. While these methods cannot be used for AP adjuvants, the ultrastructure of AP, composed of 50 nm nanoparticles, suggests that AP also has a very large surface area.
The large surface area of aluminum adjuvants allows for a high adsorptive capacity for antigens, which can be used as a key tool to allow characterization of the adjuvants. Importantly, adsorption of antigens can impact the quality and magnitude of the immune response and may enhance or decrease the stability of antigens 
. It should be noted that the dose of antigens in vaccine formulations is typically low, and usually well below full adsorptive capacity. Adsorptive capacity is affected by the type of antigens, the buffer (pH, ionic strength, composition), and other excipients, including the presence of stabilizers or surfactants. The major mechanisms involved in adsorption are ligand exchange of phosphates on the antigen with surface hydroxyls on the adjuvants, along with electrostatic and hydrophobic interactions 
. Since electrostatic mechanisms play an important role in adsorption, the adsorptive capacity of AH is often determined using a protein that is negatively charged at neutral pH, such as bovine serum albumin (BSA) or ovalbumin, while the adsorptive capacity of AP is typically evaluated with a positively charged protein, such as lysozyme 
. However, recent studies have suggested that ligand exchange contributes to the adsorption of BSA in spite of the fact that only 0.6% of its serine and threonine residues are phosphorylated 
. Marked differences were observed in the adsorption mechanisms between BSA and two different types of AH as revealed by changing the pH and tonicity of the BSA solution 
It has long been recognized that the strength of adsorption of protein antigens to aluminum adjuvants can increase over time 
, which likely reflects structural changes in the adsorbed antigens as discussed previously 
. These structural changes may improve the immunogenicity of the vaccine formulation, but can also lead to deamidation and loss of epitopes, as demonstrated for recombinant protective antigen of the anthrax bacillus 
. The stability and immunogenicity of adsorbed vaccine formulations should be studied over time.
2.5. Elemental Composition
The presence of impurities in aluminum adjuvants can be determined using inductively coupled plasma mass spectrometry (ICP-MS) 
. Differences in the type and quantity of metal ions were reported between AH adjuvants obtained from different manufacturers and different batches, which are likely caused by differences in the sources of aluminum salts, chemicals, and water used during the production process 
. The presence of sulfur likely reflects the use of alum as the starting material as discussed above. Some contaminants such as copper may affect the stability of adsorbed antigens