In order to uniformly coat the inner surface of the pipette tips composed of polypropylene with FS groups, the tips were first oxygen plasma treated for 2 min and then CVD fluorosilanized for 2.5 h using TPFS solution. During this process, hydrophilic terminals (trichlorosilane) of fluorosilane molecule bind to the plasma induced hydroxyls on the surface of the tips, resulting in self-assembled monolayers (SAMs) of FS with an umbrella-shaped structure in a way that the fluorine terminals are exposed on the surface (Fig. 1).
To analyze the changes in the chemical composition of the FS treated pipette tips, XPS test was performed on the inner surface of fluorosilanized and untreated pipette tips (Fig. 2a). The emerged peaks at about 35 eV, 690 eV, and 830 eV in the fluorosilanized spectra are attributed to F2s, F1s, and F KLL bonds, respectively. These peaks indicate the formation of FS layer inside the tips using our developed CVD method. Figure 2b reveals the atomic concentrations of the elements present at the inner surface of untreated and fluorosilanized tips, using high-resolution XPS. After fluorosilanization, the concentration of F, Si, and O increased from 0 to 28.3%, 1.2 to 4.7%, and 7.7 to 13.8, respectively. This clearly illustrates the existence of fluorosilane tail groups as a result of FS SAM formation during the CVD process. Notable, the atomic concentrations of C and impurities such as Na was diminished due to the plasma etching of the tips and breaking apart the chemical bonds at the surface.
We also conducted FTIR test to further confirm the formation of FS layer inside the tips (Fig. 2c). The spectra of both fluorosilanized and untreated tips had a number of peaks in the absorption band between 800 to 1200 cm− 1, specifically at 850 cm− 1, 1000 cm− 1, and 1200 cm− 1 vibrations that are characteristic for isotactic polypropylene. These absorption bands can be interpreted as vibrations of C-H, CH2 and CH3 groups in the polymer chain 39. In the fluorosilanized tips, however, a broad absorbance band is present around the 3700 − 3200 cm− 1. This is believed to correspond to the Si-OH group bonds formed as a result of the hydrolysis of TPFS molecules. It should be mentioned that 3700 − 3200 cm− 1 is the range of absorbance bands that commonly corresponds to alcohols and phenols, including bonds to silicone compounds 39. This pattern is notably absent when looking at the untreated tip sample spectra, which produces a flat line over the 4000 − 3000 cm− 1 region (Fig. 2c).
After fluorosilanization of the tips, they were lubricated by simply pipetting in and out a fluorinated lubricant (PFPP). The excess amount of the lubricant was then removed by extensively washing the tips via DI water. The remaining thin layer of the lubricant could be locked-in within the FS SAM through Van der Waals forces. This lubricant layer provides omniphobic properties and repellency at the inner lumen of the lubricant-infused tips. The lube, however, should remain in the tip after pipetting chemicals to preserve the omniphobicity. We evaluated the stability of the lubricant-infused tips via weighing the lube. As can be seen in Fig. 1d, after lubrication of the tips, the weight of the lubricant was 6.9 mg. When the tips were washed and the excess lube was removed, the weight of the remaining lubricant inside the tip was reduced to 4.5 mg. This amount of lubricant was very stable inside the tip as after 3-time pipetting water, the weight of the lubricant in the tips remained unchanged. This also proves that the lubricant will not be introduced to the pipetted solution when the lubricant-infused tips are used.
To examine the relative hydrophobicity/hydrophilicity of the treated and untreated pipette tips, contact angle measurements were performed using a 5 µL droplet of deionized water. The static contact angle measurements of the untreated tips, fluorosilanized tips, and lubricant infused tips are shown in Fig. 3a. Untreated tips demonstrated a lower contact angle of 75 ± 11° implying low levels of hydrophobicity. It is worth noting that the standard deviation value of the untreated samples is relatively high due to the variation in the surface properties of the commercially available tips. Upon CVD treatment and before lubricant addition, the contact angle increased to 89 ± 3°. Finally, after addition of PFPP lubricant to the fluorosilanized tips, contact angles of water were demonstrated to increase to 101.7 ± 7°.
In the contact angle measurements, the tips were cut open and the measurement was done on the concave area to assess the hydrophobicity of the inner lumen. Due to the surface curvature of the pipette tip, measuring the contact angle of a liquid droplet on the tip using a tensiometer presents complications and might result in slight inaccuracy. As such, the contact angle of water, tween 20, and isopropanol as a low surface tension liquid were measured on treated and standard pipette tips using capillary rise method. Capillary rise can be utilized to compute the contact angle of the rising/dropping liquid using hydrostatics principles. Due to the close resemblance of the tips to a truncated cone shape, the capillary rise equation was modified to account for such capillary shape (Eq. 1):
$$\theta ={cos}^{-1}\left[\frac{\rho ghR}{2\gamma }\right]-\beta$$
where θ is the contact angle, ρ is the density of the liquid, g is the gravitational constant, h is the rise/drop height, R is the diameter of tube at the three-phase region, γ is the surface tension of the liquid, and β is the conical angle of tip.
As such, the contact angles of liquids with various surface tensions on the fluorosilanized and untreated tips were acquired to test the effect of surface modification on omniphobic properties of the tips (Fig. 3c). The contact angles of all three liquids were higher on the fluorosilanized tips compared to untreated tips. This is especially evident for the lower surface tension liquids such as the tween 20 and isopropanol. The increase in the contact angles of the treated samples highlight the higher surface repellency induced by the treatment. The contact angle of tween 20 was 30° higher on the fluorosilanized surface as compared to the standard while that of isopropanol was 28° higher. Since the tips exhibited hydrophobic properties even prior to surface treatment due to the properties of the material used to manufacture the tips, there was a smaller difference in contact angle of water on the treated surface with 6.9° increase using the capillary rise method and 40° increase using the optical tensiometer. Notable, the hydrophobic nature of the intact pipette tips was not fully appreciated in Fig. 3a which could be due to the fact that the surface properties of the tips somewhat changed when the tips were cut and flattened for the purpose of the contact angle measurement.
The drop retention force on the surface of the modified tips was assessed via hysteresis contact angle measurements (Fig. 3b). We exercised needle method to evaluate the dynamic contact angle due to the curvature of the tips. While the contact angle hysteresis of untreated tips was about 17 ± 2.5°, the contact angle hysteresis of lubricant-infused tips reached down to 3.5 ± 0.6°, showing great repellent behaviour of the developed tips.
In order to quantify the efficiency of lubricant-infused tips in preventing carryover residue and cross contamination during pipetting, 10-fold serial dilutions of 1 mg mL− 1 of a food coloring dye were performed in water containing 0.05% tween 20 using lubricant-infused and untreated tips. The same pipette tip was used to perform all the dilution steps to better illustrate the effect of dye residue inside the tip during the serial dilutions. Using absorbance measurements (at the wavelength of 512 nm), we quantified the results obtained from lubricant infused and untreated pipette tips in two sizes of 10 µL and 200 µL shown in Fig. 4a and b, respectively. The results of the dilutions indicate that the untreated tips result in have significantly higher carryover residue; as a result, the dilutions did not occur efficiently. Thus, the more amount of dye that can be seen at lower concentrations could be due to the dye residue that had remained onto the tip surface from the initial high concentrations and was carried to the low concentrations.
As can be seen from the Fig. 4a and b, the lubricant infused tips show a steeper decrease in absorbance per dilution number followed by a sooner levelling off to the diluent’s absorbance value (with zero concentration of the dye) that is indicated by dashed line in Fig. 4a. On the other hand, the untreated tips reveal higher absorbance levels even for higher dilution numbers which is a result of the sample adhering to the tip’s wall even after performing several dilutions. In addition, there is significantly higher variation observed with untreated tips compared to treated ones as indicated by large error bars shown in Fig. 4a and b. In Fig. 4c, the colour difference in the well-plate after performing the serial dilutions via the lubricant infused tips and untreated tips could be observed. We also performed a visual test at the tips after pipetting in and out the food coloring dye diluted in water containing 0.05% tween 20 at the concentration of 1mg mL− 1. As shown in Fig. 4d, the untreated 200 µL pipette tip showed a great amount of dye adsorption onto the tip, whereas the lubricant-infused tip maintained significantly lower amount of dye mostly at the bottom end of the tip after pipetting out the dye solution.
In the next experiment, we used lubricant infused tips at different sizes of 10 µL (P10), 200 µL (P200), and 1 mL (P1000), to pipette in and out recalcified citrated human blood. As it is demonstrated in Fig. 5a, the lubricant infused tips in all sizes could effectively suppress blood adhesion and clot formation inside the tip in comparison with the untreated tips. SEM images in Fig. 5b also exhibits the presence of blood cells and clot on the surface of the untreated tips, while the lubricant infused modification significantly reduced the cell attachment and clot formation. We quantified the amount of clot formed inside the tips by measuring the weight of the clot. Figure 5c shows that the clot weight was significantly lower in P1000 and P200 lubricant-infused tips compared to the untreated tips. Due to the limitation in sensitivity of the weighing scale, we were not able to measure the clot weight in 10 uL tips. Overall, the results show that lubricant-infused tips are capable of repelling all blood cells, clotting factors, and other bio-species present in human whole blood.
Finally, we utilized our lubricant-infused tips to pipette MRSA bacteria solution at a concentration of 107 CFU mL− 1. After pipetting up and down the bacteria suspension using both lubricant-infused and untreated tips, the tips were submerged in 5 ml TSB and incubated in a shaking incubator for 20 minutes at 37°C (Fig. 5d). Further, CFU assay was run using the TSB solutions to compare the amount of bacteria attachment to the lubricant-infused tips and untreated tips. The growth assay results (Fig. 5e) showed about 2-log, 3-log, and 6-log reduction in CFU/ml MRSA by using P10, P200, and P1000 lubricant-infused tips, respectively. The lubricant-infused tips could repel the majority of MRSA bacteria when the bacteria solution was pipetted out and only a few bacteria remained in the tip (Fig. 5d). The result of this study reveals that a large number of bacteria physically attach to the commercially available tips during pipetting bacteria suspensions, which can bring about variation in the bacteria concentration of the pipetted solution and cause experimental errors.