Upside-Down Water

The internet has many demonstrations of the surface tension of water – floating paper clips and razor blades, soap powered paper boats and pepper particles, upside down drinking glasses with a card supporting the water inside. Another version shows a drinking glass filled with water but sealed with a plastic or metal mesh. No water flows out through the tiny mesh holes because it is held in by the water’s surface tension.

In this Instructable we show that surface tension can hold the water inside a container even if the holes are much larger. We will prepare a water-filled jar with a hole in it.

Glass jars with screw-lids

Contact glue or a hot-glue gun

Thin metal or plastic sheets, e.g.:

0.5 mm polystyrene modelling card (”Plasticard”)

Pieces cut from a water bottle (usually PETG)

Plastic ring-file separators (usually PVC)

PTFE sheet (0.25 mm available on eBay)

Aluminium sheet or cooking foil.

Drill press or electric hand drill and drill bits up to about 12 mm.

Water.

Drilling a glass jar or metal lid is tricky, so instead we prepare holes in thin sheet material to be fixed onto the jar’s lid. For the hole you can use almost any sheet material. Metal and plastic both work, of almost any thickness. I have the best results with thin sheets of hydrophobic plastics, which are poorly wetted. When wettable material is used, the water can creep out of the hole and along the surface, eventually causing leakage. For these demos I settle on 0.25 mm thick PTFE or 0.2 mm PVC file dividers.

The first problem is how to drill clean holes in thin sheets or foils. To do this we clamp the sheet between two thicker pieces, ideally of the same or similar material. I had some offcuts of 6 mm thick grey PVC which worked well. Clamp the sandwich hard using toolmakers’ clamps or strong bulldog clips, as in the figure. An undersized pilot hole in one backing sheet will help to position the drilled hole centrally on a small piece of the plastic sheet. A 6 mm hole is a good place to start.

Use plenty of drilling lubricant such as isopropyl alcohol or water, sharp drill bits and frequent chip removal – you don’t want to let it get hot and melt your way through. Once drilled, disassemble the clamps and inspect the hole. It should be clean and free from burrs.

The foils can now be glued over a larger rough hole drilled in the jar lid. A thick ring of hot-melt glue or contact cement such as UHU ”All Purpose Adhesive” should be applied to both lid and the foil and left to dry for 15 minutes. Pressing together forms a watertight seal, even with PTFE. PTFE is tough to glue without extra surface treatments, but ordinary contact adhesive is strong enough if handled carefully. 

It is important to rigorously clean the jar and the plastic sheets. Use washing-up detergent and hot water followed by rinsing in more hot water and then clean tap water. Assemble the jar and lid and fill it with cold water through the hole under a tap. Tilt the jar to remove as much air as possible. Place a finger or another piece of plastic sheet over the hole, invert the jar and then slide the cover away. The water meniscus should remain in place. If it is pendant below the hole, sweep away a little water with a finger – you should end up with an almost flat water meniscus, which is surprisingly stable. The videos show an 11 mm diameter hole in PTFE.

With a little practice you can avoid covering the hole during inversion – just quickly sweep the jar into the inverted position.

Try shaking the jar up and down. Some water may flow out but the meniscus will re-stabilize. Try tilting the jar. At some angle the meniscus will deform, down on the low side and up on the other, and water will flow out continuously. Return the jar to a vertical position and the stable meniscus will re-form and not leak further.

A small hole meniscus is even stable when oriented vertically. Here a 7 mm diameter hole in red PVC is leak-free when vertical. However, the meniscus will be distorted, bulging out in the lower half and in at the upper half. The water pressure is higher at the bottom of the meniscus than at the top.

If possible, clamp the upside-down jar so that you can observe it more carefully. Take a couple of wooden cocktail sticks or pieces of plastic insulated wire. These can be pushed through the meniscus without destabilizing it.

Now dip the end of a stick into washing up detergent and touch the meniscus. The reduced surface tension will cause water to leak, but removing the stick should allow the meniscus to re-form.

If you want to show control of the meniscus shape, add a small syringe or rubber bulb. These images show a container made from a 60 mm length of 50 mm diameter plastic pipe glued to a glass microscope slide and an 11 mm diameter PTFE hole. A 1 ml syringe allows to add or remove a little water. In this way you can adjust the meniscus to be convex (like a positive lens) or concave (a negative lens). Note that the “lens” is not of great quality as the meniscus is not a spherical surface. Rather, the pressure at the lowest point is higher and so the meniscus is a bit more pointed there. Nevertheless, you can read text through the meniscus:

View down through the 11 mm meniscus.

The surface tension we talk about (symbol usually gamma with units N/m, or force per unit length) is really the liquid-air (LA) surface tension, as every pair of materials has its own surface tension value. The liquid is continuously trying to reduce the air-water surface area to a minimum, in this circular geometry to a flat plane. The force acts parallel to the water surface, tugging at the solid and supporting the pendant droplet. When the meniscus is flat and horizontal, it provides no vertical force to stop water flowing out. So surface tension is only part of the story of keeping water in the jar.

You remember that we initially tried to minimise any air in the jar. In this state, with water being almost incompressible (also non-extendable), there is nothing trying to push water through the hole. Even modest shaking will not apply much force to the meniscus.

However, while air in the glass jar does not affect the surface tension forces holding the water meniscus in place, it makes it less stable. The air volume acts like a spring (gases are compressible) supporting the water mass. Hence any movement of the jar will tend to bounce the water on its air-cushion-spring, likely ejecting some water through the hole and letting in more air. This easily destabilises the arrangement, so start off with as little air as possible.

The next issue is symmetry. With the jar lid accurately horizontal the meniscus will be most stable. However, as shown above, tilting the lid allows the water to distort, up on one side and down on the other, eventually causing leakage. For the same reason only small holes remain leak-free on vertical surfaces.

Soaps tend to reduce the surface tension, so they reduce the maximum hole diameter that can be supported. The surface tension is not reduced to zero, so smaller holes will still be stable even with soap or other contamination. As pointed out in Wikipedia, common salt and sucrose can increase surface tension and hence support larger holes.

Upside-down water seems to be just a fun experiment but it has been put to scientific use. The liquid sample bounded by air surfaces above and below is one way to perform long-path optical transmission measurements such as spectroscopy, free from the inevitable slow contamination of conventional windows. The meniscus will collect and support debris, but the jar can be flushed and refilled automatically. In this way we performed long-term contamination-free measurements in environmental water samples that would be impossible with normal methods.