Competitive Adsorption of a Monoclonal Antibody and Nonionic Surfactant at the PDMS/Water Interface

Interfacial adsorption of monoclonal antibodies (mAbs) can cause structural deformation and induce undesired aggregation and precipitation. Nonionic surfactants are often added to reduce interfacial adsorption of mAbs which may occur during manufacturing, storage, and/or administration. As mAbs are commonly manufactured into ready-to-use syringes coated with silicone oil to improve lubrication, it is important to understand how an mAb, nonionic surfactant, and silicone oil interact at the oil/water interface. In this work, we have coated a polydimethylsiloxane (PDMS) nanofilm onto an optically flat silicon substrate to facilitate the measurements of adsorption of a model mAb, COE-3, and a commercial nonionic surfactant, polysorbate 80 (PS-80), at the siliconized PDMS/water interface using spectroscopic ellipsometry and neutron reflection. Compared to the uncoated SiO2 surface (mimicking glass), COE-3 adsorption to the PDMS surface was substantially reduced, and the adsorbed layer was characterized by the dense but thin inner layer of 16 Å and an outer diffuse layer of 20 Å, indicating structural deformation. When PS-80 was exposed to the pre-adsorbed COE-3 surface, it removed 60 wt % of COE-3 and formed a co-adsorbed layer with a similar total thickness of 36 Å. When PS-80 was injected first or as a mixture with COE-3, it completely prevented COE-3 adsorption. These findings reveal the hydrophobic nature of the PDMS surface and confirm the inhibitory role of the nonionic surfactant in preventing COE-3 adsorption at the PDMS/water interface.


Section S1 Spectroscopic Ellipsometry (SE)
A common SE measurement procedure is as follow: the unpolarized light is converted to linearly polarized and then incident on the sample surface, converting to the elliptically polarized light and being detected. The ratio of complex Fresnel reflection coefficients, ρ, can be expressed as shown in equation S1. R P and R S are the Fresnel coefficients of s and p waves after reflection. ρ can be parametrized by the phase shift △ and the amplitude variation Ψ when the polarized light is reflected by the sample 1 .
The Ellipsometer M-2000U was purchased from J.A.Woollam Co. Inc and the data was analyzed using CompleteEASE. The analysis procedure applies a theoretical model which contains the information of layer thickness and optical constants. The experimental and theoretical data are compared regressively until the best match of both is found. The accuracy of this estimation process can be quantified by the mean square error (MSE). The layer thickness d and the refractive index n, which are coupled, can be calculated by this procedure. To determine the unique d, n is first calculated using the Cauchy's equation where λ is the wavelength, A and B are the Cauchy's coefficients. In this experiment, A and B are taken as 1.45 and 0.003 for both protein and surfactants, 1.4 and 0 for PDMS layer, respectively. The surface adsorbed amount, Γ, is calculated using the De Feijter's equation 2 .
where n and n 0 are the refractive indices of the sample layer and the ambient (S1) ρ = R p R s = tan (Ψ)e iΔ S3 environment, dn/dc is the refractive index increment of the sample and is taken as 0.18 mL/g for protein and 0.12 mL/g for surfactant 3,4 . A common issue of SE is the coupling of layer thickness and refractive index for an ultrathin layer (<15 nm), but the coupled n and d can lead to the estimate of the adsorbed amount by using equation (S3) without the need to decouple thickness from refractive index.
To check the thickness of the silicon oxide layer on the silicon wafer/block we take the refractive index of the pure SiO 2 layer without any pore and defect. For the wafers and polished silicon blocks, the native oxide layers usually are 143Å thick and the absence of any defects has been confirmed by neutron reflection.
The dynamic measurement of the protein adsorption process was undertaken under "in situ" mode in which situation the silicon wafer was assembled in a purposely built SE liquid cell with a pair of fused quartz windows with the incident and exiting angles fixed to 70 .°S

ection S2 Neutron Reflection (NR)
NR experiments were undertaken on D17 reflectometer 5 at Institut Laue Langevin laboratory, Grenoble, France. The PDMS modified silicon block was assembled in a liquid cell to facilitate NR measurement at the solid/liquid interface. An alignment process was undertaken to make sure that the beam could be exactly reflected by the sample surface and detected. A transmission process was made to normalize the NR data. NR experiments can unravel the interior structure of the sample layer by recording the neutron reflectivity R, which is the ratio of the intensity of incident/reflected beam, and is plotted against the momentum transfer Q where Q can be expressed as  Table S1: SLDs of all materials used in this experiment.   Table S7 and S8. S10 Figure S9: The adsorbed amount of COE-3 at concentrations of 10, 100 and 1000 ppm (in His buffer, pH 5.5, ionic strength of 25mM) onto the SiO2 surface coated with a self-assembled monolayer of hydrophobic C18, C8, C1, PDMS and bare SiO 2 layer, respectively. The results were measured directly from the solid/water interface.      Table S6: Best-fit parameters to the reflectivity profiles shown in Figure 5 Table S7: Best-fit parameters to the reflectivity profiles shown in Figure S6