Dynamic light scattering (DLS) was performed with the zetasizer Nano ZSP (Malvern Panalytical, UK) for analysis of (F)NP size distribution. DLS samples were transferred to a quartz cuvette (ZEN 2112) at 173° (25 °C). Each sample was measured three times with the following material properties: EtOH (refractive index: 1.287, dynamic viscosity: 1.074 cP, dielectric cte: 25.3), HFE 7500 (refractive index: 1.290, dynamic viscosity: 1.240 cP, dielectric cte: 5.8), silica particles (refractive index: 1.540).
Transmission electron microscopy (TEM) was utilised as an additional tool for particle size analysis. For the TEM a 300 mesh copper grid was glow discharged for 15 s. Next 3.5 μL sample was transferred to the grid and incubated for 5 min. After blotting samples were imaged.
Energy dispersive X-ray (EDX) composition analysis allowed for analysing successful fluorination of the silica NPs. Here the samples were oven dried at 70 °C overnight and then transferred to carbon tape. Afterwards samples were imaged with the Quanta 200 ESEM FEG (FEI, USA) with EDX system.
1.3 microfluidic chip production, 1.4 materials and imaging conditions for droplets.
A volume of 10 μL of droplets containing only buffer was combined with another 10 μL of dye-encapsulating droplets. Droplet populations were imaged directly after combination and after one day of incubation by transferring 4 μL of droplet solution inside a cell-counting chamber from which brightfield and epifluorescence micrographs were captured with a Nikon Ti2 microscope.
For the on-chip leakage fluorescence imaging three dyes were selected: AF488 (a well retained dye), sulford (a moderately leaky dye), Rhod6G (a strongly leaky dye). These were imaged on the Nikon Ti2. Chip design 1 (20 μM thick, see Fig. S1 † ) with 1 μL min −1 (aqueous phase) and 2 μL min −1 oil flow rate utilised for droplet generation.
(1) |
With this fitting, the medians and standard deviations of two different normal distributions and the mixing parameter were calculated. If the mixing parameter was close to zero or one, the bimodal normal distribution approaches a normal distribution with only one bell-shaped curve. That indicates that there is only one droplet population instead of two indicating transfer of dye from the droplets encapsulating dye to the empty droplets. However, this parameter is not a perfect indicator of cross-talk on its own: if the medians of two bell-shaped curves are close to each other (in the case of cross-talk with two overlapping populations), the mixing parameter can still be close to 0.5. For this reason a calculation of the intersection area was used, since it provides a more accurate indication of the occurrence of cross-talk ( e.g. , the intersection area corresponds to the percentage of droplets that cannot be differentiated into two distinct populations). In the absence of cross-talk, the intersection area was close to zero. Based on our analysis and our experimental observations, an intersection of 10% was chosen for identifying cross-talk. For a few experiments, though, the bimodal distribution fitting was unreliable ( e.g. , R 2 < 70), likely due to some variability in the experimental conditions at different time points. In these rare cases, the final classification was made based on the visual inspection of the data and the observations made during the experiments.
To obtain the data for the fitting images were analysed with a custom Phyton script. The location and size of the droplets were detected on the brightfield images through a Hough circle transform. The fluorescence intensities at these location were measured and averaged across each droplet. The background was detected by measuring the intensities at three different locations outside of the droplets (image background) and in pure buffer solution (buffer background). The droplet size distribution and fluorescence data were extracted from the Phyton pipeline. Then, this data was fed to a MATLAB R2023a (statistics and machine learning toolbox) to form the histograms and do further statistical analysis by fitting the data to probability distributions as described above.
To check for the leakage of fluorophores from the droplets, the percentage of droplets with an average intensity value below the background value (image background) was detected. If more then 10% of the droplets had an intensity below the background the sample was categorised as very leaky (category 4, see below). To calculate the cross-talk between the droplets, buffer background was subtracted from the fluorescence data of droplets and the resulting data were normalised between 0 and 100. This scaling provided a more suitable way for comparing different populations. The data was fitted to a bimodal normal distribution with the function “fitgmdist” with two components to represent two normal distributions. Then, a probability density between 0 and 100 was calculated based on the predicted median, standard deviation and mixing parameter, and drawn as a linear curve. In order to check for cross-talk, the intersection area of the two normal distributions was calculated by “cumtrapz” function. This function calculates the integral of the minimum values of the two probability densities by trapezoidal numerical integration. Then, histograms were generated with a hundred bins based on the raw normalised data, finally the intersection area was calculated. Based on the calculation above dye leakage and cross-talk was divided into four categories (see Fig. 4 for a visual overview).
• Category 1: more than 10% of the droplets show a lower average fluorescence intensity than the background value on day 0 indicating severe leakage of the dye into the oil phase.
• Category 2: less than 10% of the droplets show a lower average fluorescence intensity than the background value on day 0 and the bimodal distribution shows a minimal overlap of 10% on day 0 and 1. These droplets show limited leakage and severe cross-talk.
• Category 3: less than 10% of the droplets show a lower average fluorescence intensity than the background value on day 0 and the bimodal distribution shows a minimal overlap of 10% only on day 1. These droplets show limited leakage and mild cross-talk.
• Category 4: less than 10% of the droplets show a lower average fluorescence intensity than the background value on day 0 and the bimodal distribution shows an overlap of less than 10% on day 0 and 1. The droplets show limited leakage and cross-talk in this category.
1.7 cell adhesion and encapsulation, 2. results and discussion, 2.1 fnp characterisation.
A) DLS data of 100 nm (F) NPs. (B): EDX spectrum of (F) NPs. (C): TEM images of FNPs (scale bar 500 nm). |
Dye | DPBS | DMEM | MiliQ | Tris | Charge (pH: 7.4) |
---|---|---|---|---|---|
Rhod 6 G | 5.65 | 5.65 | 5.64 | 5.65 | 0 |
NR | 4.62 | 4.62 | 4.62 | 4.62 | 0 |
Rhod101 | 4.18 | 4.07 | −2.85 | −0.13 | 0 |
RhodB | 3.01 | 2.89 | −4.29 | −1.39 | 0 |
EMA | 2.71 | 2.61 | 0.97 | 0.39 | −1 |
Cy7 | 2.45 | 2.34 | −1.68 | 1.22 | −1 |
Res | 0.63 | 0.63 | 0.63 | 0.63 | 0 |
AZ | −0.31 | −0.31 | −0.29 | −0.31 | −1 |
Sulford | −0.85 | −0.96 | −4.98 | −2.08 | −1 |
Fl | −1.34 | −1.34 | −1.07 | −2.00 | −2 |
Cy5 | −2.27 | −2.38 | −6.4 | −3.5 | −2 |
DAPI | −3.24 | −3.46 | −7.93 | −5.5 | +2 |
AF350 | −5.92 | −5.92 | −5.66 | −6.16 | −2 |
Pyr | −6.33 | −6.33 | −6.31 | −6.34 | −3 |
AF488 | −7.13 | −7.13 | −7.03 | −7.53 | −3 |
AF568 | −7.20 | −7.31 | −13.64 | −11.6 | −3 |
Comparison of dye partitioning from aqueous phase into oil phase (HFE 7500) containing RAN, Krytox or FNPs. Samples were compared with one-sided t-test assuming variance and normal distribution. Colours were given according to p-values green for p-value > 0.01 red for p-value < 0.01. Dyes were sorted based on the distribution coefficients in DPBS. |
For NR no significant enhancement of dye leakage was measured compared to the HFE samples, this results from the strong partitioning of the dye into the oil phase. 42 If the majority of the dye already leaks in the pure HFE case, additional surfactant will not be able to induce any additional leakage.
Examples of the 4 different leakage categories. (A) Category 1: RhodB in DPBS with Krytox, strong leakage of dye. (B) Category 2: Rhod6G in DPBS with RAN, retention of the dye but strong cross-talk. (C) Category 3: Fl in DPBS with Krytox, retention of the dye but cross-talk over time. (D) Category 4: Cy5 in DPBS with RAN, retention of the dye, no cross-talk after 1 day (100 μM scale bar). |
A colour-coded overview of the leakage categories in which all the different conditions tested were classified is presented in Fig. 5 . A first important observation that can be made from these results is that cross-talk was minimally influenced by the buffer solutions in which the dyes were dissolved, in line with previous observations by Janiesch et al. 23 Another clear trend that can be inferred is that, for all these conditions, the more hydrophobic a dye was, the more leakage observed.
Colour coded overview of leakage between droplet. Dyes are sorted based on distribution coefficient in DPBS (at the top the most hydrophobic dyes and at the bottom the more hydrophilic ones). |
Perhaps the most striking result from our experiments is that, while it was previously reported that FNPs in fluorinated Pickering emulsions create a protective shell around droplets that effectively deters leakage, 25 our observations clearly contradict these claims: our FNP-stabilised droplets indeed exhibited significant degrees of leakage. This phenomenon was particularly pronounced for highly hydrophobic dyes (typically also neutrally charged), which in all cases surpassed the leakage levels observed with commercially-available RAN surfactant. There was a single, notable exception to this trend, namely for the Res dye, which leaked out from RAN- and Krytox-stabilised droplets but not from FNP-stabilised samples. This exceptional behaviour might be the result of the neutral charge of resorufin, in combination with its intermediate distribution coefficient and small size. As explained in the previous section, leakage from FNP-stabilised droplets can be explained by a combination of dye-particle interactions with a consequent change in the surface energy of these FNPs.
The leakage of DAPI from Krytox- and FNP-stabilised droplets possibly results from the electrostatic interaction between the Krytox and FNPs charged groups and the many amine groups present in the DAPI molecules, providing it with a positive charge. Additionally, 23 showed that low molecular dyes with planar structures can cross the droplet barrier more easily, which could potentially further explain the higher leakage observed for the DAPI, Fl and Res dyes. The UV-vis plate experiments and the observation of leakage in droplets do show some difference. This can be attributed due to the higher sensitivity of the plate assay compared to the droplet experiments.
Finally, it also needs to be noted that the low leakage of RAN-stabilised droplets could be partially explained by the use of BSA in our inner aqueous media, which was previously shown to reduce leakage by adsorbing to the o/w interface. 42 We chose to perform these experiments with BSA since BSA was essential for the rapid stabilisation of droplets with FNPs. To allow comparison the same buffer conditions were used for Krytox and RAN, although here BSA was not necessary for droplet stabilisation.
On chip droplet production for three different dyes and three different surfactant, imaged with Nikon Ti2 (100 μM scale bar). |
The assessment of droplet stability upon PCR-like thermocycling is presented in terms of the evolution of the coefficient of variance (CV) values of the droplets' diameter before and after this treatment, and is depicted in Fig. 7 . The most striking outcome of this process occurred in droplets stabilised with Krytox, which acquired irregular, non-circular shapes (and were, therefore, not analysed for their diameter and CV after the thermocycling process). Remarkably, these droplets were still quite stable: no fusion events were observed during the imaging process when droplets collided with each other. The effects of temperature on surfactant solutions were previously studied with molecular dynamics simulations, which concluded that as the temperature increases, thermophoresis of water molecules causes the hydrogen bonding between water and surfactant molecules to fracture and reconnect. 43,44 As mentioned previously, Krytox is composed of a PFPE chain functionalized with a carboxylic acid group. 1,20–22 To explain these results, we hypothesise that the carboxylic acid groups of Krytox molecules are fracturing their hydrogen bonds with water molecules and reforming them with carboxylic acids of neighbouring surfactant molecules, a phenomenon that has previously been observed in carboxylated PFPEs. 45 In these interfacial conditions, this reaction possibly leads to the formation of some form of hydrogen-bonded (fluoro)organic framework 46 at the droplet interface that behaves more like a gel and, therefore, can sustain deformations and support the non-circular shape of such droplets.
A) Boxplot of CV values of droplet diameters measured before and after a PCR-like thermocycling process. (B) Representative images of the same droplets (scale bar 100 μM). |
For the other two stabilisation systems exposed to a PCR-like thermocycling, the original droplet size was around 60 μM with a CV below 5%. After thermocycling, the CV of RAN-stabilised droplets became three times larger, indicating that a significant portion of the droplets stabilised with this reagent fused under these conditions. On the other hand, FNP-stabilised droplets remained relatively stable with only a limited increase in their CV as was shown in previous research. 47 The lower stability of RAN droplets can be attributed to the thermoresponsive nature of the PEG molecules in this triblock copolymer, which potentially destabilises droplets during temperature changes. 48–50 Additionally, the production of RAN does not result in a perfectly pure product, with traces of precursors, ionically-coupled surfactant and diblock copolymers still present. As previously suggested, this small amount of contaminating species might decrease the stability of RAN droplets. 17,50
With regards to the long-term stability upon storage in different conditions, as illustrated by the CV values of the diameters of droplets (see Fig. 8 ), these always increased over time for RAN-stabilised droplets but for Krytox-stabilised droplets a CV increase was only seen after storage in the incubator. Conversely, for FNP-stabilised droplets, CV values remained constant in all storage conditions.
Long term stability of droplets stabilized by different reagents under three different storage conditions: in the fridge, at room temperature and in an incubator. CV values and standard deviations of droplet diameters are plotted over a time period of two weeks. |
From the results of the above treatment and storage conditions, we can conclude that FNP-stabilised droplets have a superior stability compared to surfactant-stabilised droplets. This can be explained by the higher energetic stability of the particles at the droplet interface. 41
For all cases the aqueous phase consisted of DPBS with 5 mg mL −1 BSA. As it was shown in previous papers, BSA promotes a more rapid stabilisation of droplets. 30 Alternatively, a long serpentine channel could have been added to the end of the design to specifically allow FNPs sufficient time to adsorb at the droplet interface. We did not make use of this strategy since for the creation of small droplets this results in a large increase in hydraulic resistance and the consequent increase in pressure to drive flow usually leads to the delamination of the PDMS chip from its sealing glass substrate and failure of the device.
Resulting graphs for the three different flow focusing designs are plotted in Fig. 9 . For the 20 μM flow focusing designs, Krytox- and RAN-stabilised droplet production behaved in a similar fashion. Similar flow rate ratios resulted in a similar droplet size. The FNP-stabilised droplets tended to be smaller for design 1 but in case of the other two designs droplet size was similar to that of the droplets stabilised by surfactants. In all of the above cases, the standard deviations are comparable, indicating a robust production of monodisperse and stable droplets in all systems.
Droplet size in function of the flow rate ratio (flow rate of the oil phase/flow rate of the aqueous phase) for three different flow focusing designs with 2% Krytox, 2% RAN and FNPs. |
For the 10 μM designs, production of RAN- and Krytox-stabilised droplets remained again relatively similar. For FNP stabiliser, though, droplets were slightly larger for the second and the third designs. In all cases the standard deviation are slightly higher compared to the 20 μM design, which could arise as a result of a higher sensitivity to pressure fluctuations of the pumps for devices with smaller channel sizes.
In the T-junction device, droplets could be produced for all three droplet stabilisers. The trend shows that the higher the flow rate ratio was, the smaller the droplets were, starting at around 35 μM for a flow rate ratio of 1 decreasing to around 20 μM for a flow rate ratio of 8, this can be expected upon inspection of the literature. 51 No significant differences between droplet size and monodispersity, as well as stability, could be observed for the three different stabilisers after these experiments.
With the two different step-emulsification designs droplets could be produced with 45 and 15 μM in diameter (see Fig. 10 ). Interestingly, when BSA was added to DPBS, no droplets could be produced with FNPs as stabilising agents. In these conditions, the droplets did not pinch-off as expected and only very large, polydisperse droplets were formed. When utilising MiliQ water without BSA, though, droplet pinch-off did effectively happen. This indicates that the compounds present in the aqueous solution that one attempts to encapsulate in droplets can critically influence droplet formation when FNPs are used as droplet stabilisers.
The diverse influence that BSA exerts on the production of droplets in these different droplet-generation systems can be explained due to the attraction between the positively-charged amino acids in BSA and the negative charges on these silica-based FNPs. For the flow focusing designs (if no long serpentine channel is present) BSA may play an essential role in promoting a rapid particle accumulation ( i.e. beyond their diffusion-limited rate) at the droplet interface. Additionally, hydrophobic–hydrophobic interaction between the fluorocarbon chains and the BSA molecules could also facilitate this more rapid stabilisation. 30 In the case of the step emulsification devices, since droplet formation is a relatively slower process as it is not dependant on high shear energy as in flow focusing devices, 52 there appears to be no need for the use of a stabilisation enhancer. In these circumstances, BSA seemed to actually negatively impact droplet formation, possibly by affecting the droplet necking and pinch-off processes due to its exertion of dynamic, localised changes in surface tension, which is a critical parameter regulating droplet formation in these systems. 52–54
Finally, in the device combining flow focusing with step emulsification, droplet production resulted in droplets of about 10 μM in size for all three different droplet stabilisers. Here, solutions both with and without BSA could be utilised without any detrimental effects in droplet monodispersity and stability.
In general droplet size and polydispersity was very similar for all the different stabilisers. For the use of FNPs for droplet stabilisation, though, it is important to highlight some important differences with respect to surfactants. During our experiments, we noticed that a good modification of the surface of the devices ( i.e. , with fluorinated silanes) is of paramount importance for all designs except for the step emulsification ones. In flow focusing chips that were silanised more than two days before droplet production, some of the droplets readily wetted parts of the channel where, presumable, the fluorosilane layer had become deteriorated. These wetting issues resulted in droplet fusion and splitting events that led to a final polydisperse droplet population. Secondly, FNPs aggregated when in direct contact with the silanising agent (probably due to the high reactivity of the silanising agent with silanol groups), resulting in channel obstruction. This could be avoided by immediately rinsing the devices with pure HFE 7500 upon finalisation of the silanisation treatment.
Cell adhesion investigated in 96 well plate for MCF-7 and MCF-10A cells on a tissue-culture treated plate, on HFE 7500, on HFE 7500 with RAN surfactant and with FNP after 48 h of incubation (scale bar 100 μM). |
In another study by Lin et al. 31 it was observed that MCF-7 cells could adhere to the droplet border when encapsulated in droplets stabilised by FNPs. Given the observed adhesion in the plate-based experiment, both cell types were encapsulated in droplets of approximately 100 μM in diameter, which were stabilised by 2% Krytox, 2% RAN and FNPs (see Fig. 12 ). Once more, since there was an interaction between the cell medium and Krytox (ESI S4 † ), only results of RAN and FNP stabilised droplets are discussed. After an incubation of 24 h and 48 h, it was observed that when imaging droplets stabilised with RAN, all cells were in the same focal plane, while when imaging droplets stabilised with FNPs, cells were in different focal planes. Although these results indicate adhesion of the cells to the droplet edge, the typical morphology of spreading cells ( i.e. when cells flatten and extend their membranes after adhering to a surface), could not be observed at the droplet interface, which might indicate that the adhesion is non-specific or that the adhesion does not result in spreading due to the high curvature of the droplets. Given the different results obtained here and in the work of Lin et al. 31 where spreading was observed, future research is needed to investigate different parameters that might influence cellular adhesion to the FNP stabilised droplets, such as droplet size, FNP concentration, FNP fabrication, FNP characteristics, etc.
Cells (MCF-7 and MCF-10A) encapsulated in microfluidic droplets stabilised by RAN surfactant and FNPs. Representative image after 24 h of incubation (scale bar 100 μM). |
The process of droplet production exhibits strong similarity between FNPs, RAN, and Krytox for droplets generated by flow focusing. Initially we hypothesised that the slow diffusion of the particles to the droplet interface might prove problematic for droplet formation. However, this can be effectively addressed by the addition of BSA to the inner buffer solution, with the exception of step-emulsification devices, where BSA appears to impede the droplet pinch-off process. Encouragingly, FNPs demonstrate superior droplet stability under all examined storage conditions when compared to RAN and Krytox-stabilised droplets. Finally the adhesion of anchorage dependant cell is enhanced in a plate-based assay but could not be observed inside the droplets.
Looking ahead, further modifications of silica particles through silane chemistry hold the promise of providing droplets with more functionally tailored surfaces that can interact in specific manners with the droplet contents. Currently, the absence of commercially available fluorinated particles presents an opportunity for future advancements in the field, potentially leading to greater reproducibility and facilitating broader applications within microfluidics.
This study provides a comparison between FNPs, Krytox and RAN, comparing droplet leakage, cross-talk, stability, production and the adhesion of anchorage dependant cells. The findings shown in this manuscript can help in the choice of the most suitable surfactant for the right applications, driving further microfluidic innovations.
Acknowledgements, notes and references.
† Electronic supplementary information (ESI) available. See DOI: |
IMAGES
COMMENTS
Demonstrate how refraction of light occurs in water with this simple water refraction experiment for kids. Printable project included!
How does this work? Refraction ( bending of light ) happens when light travels between two mediums. In the refraction experiment above light travels from the arrow through the air, through the glass, the water, the glass again and air again before reaching your eyes.
The science behind the water refraction experiment: Refraction is the bending of light. In this case, light traveled from the air, through the front of the glass jar, through the water, through the back of the glass jar, and then back through the air, before hitting the picture. Whenever light passes from one medium into another, it refracts.
In this experiment, kids learn about refraction. This is the process in which when light passes from one material to another, it bends or refracts. This is a...
Refraction of Light in Water - Cool Science Experiment for Kids | Kids Education by Mocomi MocomiKids 662K subscribers Subscribed 811
Background When light that is traveling through the air hits water, some of the light is reflected off the water. The rest of the light passes through the water, but it bends (or refracts) as it enters the water. The same thing happens when light hits glass, or any other transparent material. Some light is reflected off the object, while the rest passes through and is refracted. All materials ...
Light Refraction Science Experiment Instructions. Step 1 - Get a sheet of paper and draw two arrows on it. One arrow near the top and one arrow near the bottom. Make the arrows point in the same direction. Step 2 - Fill a glass with water.
Discover the magic of light refraction with a simple and enchanting science experiment. Learn how to make arrows change direction using just a glass of water and a piece of paper, exploring the fascinating physics behind light's behavior.
A simple science experiment for kids to learn about water refraction. Reverse the colored bars or reverse the arrows. Learn more: http://www.whatdowedoallda...
What is refraction? Refraction is the bending of light. This occurs when light travels through one material to another (ex: air, water, etc…) Unlike objects that reflect, objects that refract light look different. For example, when looking at a glass of water with a straw in it, the straw may appear to be broken. This happens because light moves more slowly in water than in air. As a result ...
Refraction is the bending of light (it also happens with sound, water and other waves) as it passes from one transparent substance into another.
This easy light refraction experiment is a cool way to show kids how light bends in water!
Refraction is the bending of light as it passes from one transparent medium into another 1 . Let's explore a simple water refraction experiment using everyday materials like a glass of water placed in front of a sheet of paper. In this experiment, we observe the phenomenon of refraction in action. This fun and educational activity will help ...
What is refraction of light? When light travels from one transparent medium (air) to another transparent medium (water), the speed of the light slows down and when it hits water it changes its direction slightly. This change in the direction of light is known as refraction of light. In a simple language, light refraction is bending of light.
Explore bending of light between two media with different indices of refraction. See how changing from air to water to glass changes the bending angle. Play with prisms of different shapes and make rainbows.
Now, we know from refraction indices, or just in general, that light moves slower in water than it does in air. So it's slower in water and faster in air. So let's think about what's going to happen. Let me draw two rays that are coming from this point on the straw right over here.
Courses on Khan Academy are always 100% free. Start practicing—and saving your progress—now: https://www.khanacademy.org/science/p... Refraction in Water. Created by Sal Khan.
Refraction and light bending. You might have heard people talk about Einstein's speed of light, and that it's always the same. The part that most people leave out is that this is only true in a vacuum—when there's no pesky molecules of air or water to slow it down. But when light moves through a more familiar medium like air, it moves ...
Simple experiment for kids to demonstrate the refraction of light. Help answer the question: Can Water Bend a Straw?
Are you looking to bring science to life in the most engaging way possible? We've curated a list of easy-to-set-up water refraction activities to ignite your student's curiosity. These experiments, which involve bending light and color spectrums, offer a practical exploration of the laws of physics. While providing endless fun, they will also help you …
This Old Experiment With Mice Led to Bleak Predictions for Humanity's Future. From the 1950s to the 1970s, researcher John Calhoun gave rodents unlimited food and studied their behavior in ...
Cool Light Refraction Science Experiment In this Video we show you the steps to conduct this cool science experiment and you can follow along at home using a few common items. Enjoy the video!
The acoustic refractive index and impedance distribution in the head of a short-beaked common dolphin (Delphinus delphis) were determined by computed tomography (CT) scanning and tissue experiments (see Note S1 for details). Figure 1B displays the CT images of the sagittal cross-section of the dolphin's head.
The excellent agreement of fully quantum theory and experiment found here for water collisions with helium at a collision energy relevant to that of the interstellar media place the theoretically determined potential energy surface and the collision cross sections extracted using this surface on a firmer basis.
Easy and Fun science experiment to do with your kids to see a "magical" happening how the color can disappear from the picture. Let's try this water refracti...
The effects of temperature on surfactant solutions were previously studied with molecular dynamics simulations, which concluded that as the temperature increases, thermophoresis of water molecules causes the hydrogen bonding between water and surfactant molecules to fracture and reconnect. 43,44 As mentioned previously, Krytox is composed of a ...
Refraction - oil, glass, water - Show Me The Physics Website Mr. Mangiacapre 7.58K subscribers Subscribed Like 96K views 14 years ago Highlights