As we've all learned, water is simply H2O - meaning that its molecule comprises two hydrogen atoms attached to an oxygen atom. Water is the world's most common oxide. While there are many oxides (compounds made with elements attached to oxygen), water is the most familiar to us: H2O.
For example, rust is an oxide - it's iron attached to oxygen. Most metals, and many other materials can oxidize by exposing them to air and/or water - both of which contain lots of oxygen. In water, two hydrogen atoms are covalently bonded to an oxygen atom. By "covalently," we mean that the electrons in the compound are shared by both atoms, as we'll explain below.
An Oxygen Atom (8 Electrons + 8 Protons) and a Hydrogen Atom (1 Electron + 1 Proton)
Both Oxygen and Hydrogen are "neutral atoms," meaning that each has the same number of protons as it does electrons. Note that electrons in atoms reside in different energy "shells." All electrons are attracted to the nucleus, but only so many of them will fit in a particular energy level. The first level is the K Shell, which can hold 2 electrons; the next level is the L shell, which can hold 8 electrons. Because Hydrogen only has 1 electron, it sits in the K shell - with room for one more. And because Oxygen has 8 electrons, 2 of them sit in the K shell, and 6 sit in the L shell with room for 2 more.
Because there is room for 2 more electrons in Oxygen's L shell, Oxygen easily bonds to two Hydrogen atoms, sharing their electrons "covalently." This means that Oxygen naturally combines with two Hydrogen atoms into a neutral molecule in which the electrons are shared by all nuclei. Although there are 3 nuclei, there are 10 protons and 10 electrons, as shown in the figure below. Note that the 8 electrons in the L shell are shared by all nuclei, and the angle between the hydrogen atoms is 104.5 degrees (remember that this is a "sphere," not a circle).
Water: The 8 Electrons in Oxygen's L-Shell are Shared Covalently with 2 Hydrogen Atoms
While the hydrogen atoms are primarily attracted to their host oxygen atom, because of covalence, they are also attracted to all adjacent oxygen atoms. This "holds water together," and it is the reason for the meniscus that forms when you fill a glass to the very top. (The meniscus is the surface tension that allows the water to form a slight dome that's higher than the glass.) And the hydrogen bonding between molecules (the covalence) causes the water molecules to pull together about 15% more than what would be expected from a more neutral liquid. And because the angle between the covalent Hydrogen Atoms in a molecule is 104.5 Degrees, this fits naturally.
Because oxygen is negatively charged (-2), it bonds easily with two hydrogen atoms, each of which is positively charged (+1 each). As a compound, water is stable, but it reacts very easily with other compounds, especially when it is heated. This is important to understand as we see how "water" is used to make coffee.
The surfaces of metals like copper, aluminium, and titanium - metals commonly used in boilers - will oxidize readily when exposed to water and air. Note that a surface oxide is a good thing on these elements. It's a coating that protects the metal from other kinds of corrosion. And oxide is a passivation compound, which means that it's an electrical insulator (which usually helps, although it can "fool" the circuits sometimes).
Note that oxygen readily combines with many other compounds, and many things that we take for granted on Earth have oxygen in them. While the atmosphere contains CO2 (carbon dioxide), the earths crust contains lots of SiO2 (silicon dioxide) which frequently manifests as sand. But Earth's crust is made of aluminum oxide, silicon dioxide, iron oxide, and many other compounds such as silicates of magnesium and iron.
In addition to these many common elemental combinations of various things with oxygen, we make soaps, adhesives, fertilizers, painting compounds, cements, and many other things from water soluable silicates.
The point is that water (H2O) reacts easily with most things on earth, and the fluid that we call "water" is seldom merely H2O. The water that we drink and use for food preparation contains both salts and acids, and various minerals. These all react with coffee, and with the components of your espresso machine.
We had said that heat makes water more reactive. One result of this is that you will find most of your corrosion and mineral buildup at junctions within the machine at which there are large thermal gradients (e.g., the points at which things connect to the boiler).
Again, these things are not "bad;" they are natural. But you should soften your water, and periodically clean the pieces of your espresso machine that are subject to scaling, or your machine will not function at its best.
Your espresso machine also contains electrical sensors that it uses to monitor the water. Most notably, the controller that keeps the boiler filled does this with electrical sensors. The boiler of an espresso machine is not actually full. It is kept a little more than half full, and is pressurized to about 1.5 Atmospheres. There should be a gauge on your front panel showing the boiler pressure.
The reason that the boiler is only half full is that this is the steam source for frothing and steaming. This boiler is NOT the source of the drinking water used for your espresso, although the drinking water is run through insulated tubes that pass through the boiler to heat it. Inside the steam boiler is an electrode that comes down from the top of the boiler to the surface of the water. The height of the electrode tip is set at the factory.
There is a circuit that tries to pass a small current through this electrode. If the electrode is not touching the water (i.e., if the water level is too low), no current will flow. If no current flows (meaning that the resistance is very large), the controller will activate a solenoid to open the water input line. When enough water flows in to submerge the tip of the electrode, an electrical current flows, and the controller will shut off the water input.
There are two things to notice here. First is the fact that the electrical path (that senses the presence of water) is from the controller through a wire to the electrode, then through the water, to the tank wall itself (another electrode), then back through a wire to the controller. The wires and the electrodes all have very low resistance, so this circuit is not that sensitive (meaning that it works very easily over a wide range of possible values and qualities of electrical connections) provided that the water conducts.
And second, that as scaling builds up on the tip of the electrode, the water level will have to rise slightly over a period of time (as the scaling builds) for the circuit to work. This second point is easy to monitor. Most machines have a gauge on the front to show the water depth in the boiler. If this is higher than it should be after a year or two of use, then the electrode needs to be cleaned.
But why are we making the first point about the water? Isn't this obvious? Not always. Overly purified water (e.g., distilled water) can fool the sensors. While we don't want hard water because we don't want mineral buildup, we need some mineral content in the water both for flavor, and so that it conducts electricity good enough for the sensors to work.
But isn't water a conductor? Not quite. Pure H20 (i.e., distilled water) is a poor conductor. In fact, in laboratories, they usually call it "deionized water," because there are few free ions (which are needed to conduct electricity). If there were NO free ions, water would be an insulator. But in the liquid state, even deionoized water constantly undergoes auto-ionization as it sits, so it conducts electricity somewhat. Because of auto-ionization, the electrical resistivity of "pure" (deionized) water has been shown to be about 182 KiloOhm-Meters (we will use 200 Kohm as a round number).
But water is also an excellent solvent, so it is rarely "pure" unless it has been subject to ultra-filtering and deionization. It almost always has some solute - at least a salt - in it. Introducing small amounts of solute (as nature does) increases the conductivity of water quite a bit by liberating lots of free ions (which is why it's a solute). The dissolution of any ionic material (such as salt - which is in everything, or acid - which is in many things) will reduce the electrical resistance of water dramatically. Even adding a solute in concentrations as small as 100 parts per trillion will have an easily measurable effect on resistance.
As an upper bound, we'll start with sea water, which contains nearly everything - including lots of salt, and weak acids. It's a GREAT conductor. Its conductance is about 5 Siemens per meter, which corresponds to a resistance of merely 0.2 Ohm-Meters. As we had just said, deionized water has a conductance that is about a million times smaller - about 200,000 Ohm-Meters. This is a huge range: 0.2 to 200,000!
We had mentioned salts and weak acids. In real water, we'll find both. If we put only weak acids into deionized water, most of the conductance will take place only between the acetates formed (and the free hydrogen ions). The resistance of a solution of water with weak acids will only go down to 1.4 Ohm-Meters, which is 7X larger than the resistance of sea water - incorrectly called simply "salt water" - which we have just stated to be 0.2 Ohm-Meters.
And if we put only salts (but no acids at all) into deionized water, some of the salts split into ionized chlorides. In this case, the resistance will only go down to 0.6 Ohm-Meters, which is 3X more than that of sea water. The curves below show pure water mixed with salts, and pure water mixed with acids. We would expect "real water" (like sea water) to have a resistance of 0.4 Ohm-Meters.
This is shown in the diagram below. Weak acids in water have a resistanc of 1.4 Ohm-Meters, and salts dissolved in water will have a resistance of 0.6 Ohm-Meters. Together they should give real water a resistance of 0.4 Ohm-Meters.
The Resistances of Weak Acids and Simple Salts Dissolved in Water
The curves below show what we would expect to happen when salts (found everywhere on earth) and weak acids (as found in rain, dissolved vegetation, and many other things on the earth) are dissolved in water. As their concentrations increase, the weak acids asymptotically approach 1.4 Ohm-Meters, and the salts asymptotically approach 0.6 Ohm-Meters. Their combination in parallel should give the water a resistance of 0.4 Ohn-Meters.
The Asymptotic Resistances of Weak Acids and Salts as They Saturate the Water
But we had stated that fully saturated water (e.g., sea water) is known to have a resistance of 0.2 Ohm-Meters. So why does real water have a resistance that's much less - about half - of this? It's because water is a substance that readily enables constant simple reactions - especially when heated. Since real water will have both salts and weak acids, the two together will also produce STRONG acids (e.g., hydrochloric acid), which has a resistance of about 0.4 Ohm-Meters. The acetates and chlorides, together with hydrochloric acid provides THREE parallel resistances: the two already mentioned, and the new strong acid (HCl) as shown in the figure below.
Salt Together With Weak Acids Form STRONG Acids, Which Dramatically Reduce Resistance
So in real water, we'll have acetates which give us 1.4 Ohm-Meters, we'll have chlorides which give us 0.6 Ohm-Meters, and we'll have hydrochloric acid which gives us 0.4 Ohm-Meters. These three solutions together (i.e., "water"), cause the resistances to work in parallel. The resistance of 1.4 in parallel with 0.6 in parallel with 0.4 is 0.2 Ohm-Meters - which is what is a lower limit (and is typical of sea water). These reactions are shown in the figure below.
Newly Formed Hydrochloric Acid Drops the Resistance Considerably
If we used pure (highly-resistive) distilled water, circuitry in the espresso machine that uses conduction to sense the presence of water would not always sense water correctly. So what about "normal" water?
Normal drinking water (tap water) has a resistance in the range of 20 Ohm-Meters to 2,000 Ohm-Meters. Normal drinking water has many of the same elements as found in sea water, but in much smaller concentrations; hence the 2-4 order of magnitude difference. Granted, there is a factor of 100X within this range, but anything in here is easy for the circuits in your espresso machine to detect as "water."
Note that salts will tend to consolidate and build up plaques inside your machinery especially at points in which there are steep thermal gradients. And strong acids will corrode metal surfaces. The saving grace is that metal surfaces will oxidize too. Normally, we think of "oxidation" as a BAD thing (i.e., rust), but in small layers, it forms a protective coating. Think of a copper coin. It doesn't "rust," but it has an oxide layer that makes it dull, and protects it.
While anything within the range of 20 Ohm-Meters to 2,000 Ohm-Meters will allow the espresso machine to work correctly, we would prefer to use water on the softer side (nearer the 2,000 Ohm-meter range) to minimize calcification and acidity in your espresso machine. In fact, all manufacturers require that Water Softeners be installed in series with your water source for their warranties to be honored.
So it's a GOOD thing to soften your water supply. But you don't want to over-purify it. Not only will it lack flavor, but it may "confuse" the controls in your espresso machine.
...written by your friends at
The Coffee Brewers