My interest in Rochelle salt is due to its piezoelectric properties. Basically that means a crystal of the salt generates electricity when it is squeezed. If that doesn’t seem fascinating to you then you might have come to the wrong place. I’ll get to the specifics in a minute, but first I want you to understand that this is NOT intended to be a scholarly document. I’ve taken liberally from existing sources without citation or reference. Heck, I’ve probably plagiarized whole chunks of this without regard to giving anyone proper credit. So watch out, it’s up to you to Google the stuff you find here and figure out where it actually came from if you are going to use it for some kind of research.
The story of Rochelle salt starts in the old port city of La Rochelle, France about 1665 (Figure 1). Jehan Seignette (Figure 1A) and his son Elie (Figure 2) were living in La Rochelle working as pharmacists. At the time, pharmacy primarily consisted of extracting essences from plants much like herbal medicine is today.
Seems people have always had problems with becoming constipated and therefore the need for laxatives. The most common laxative of the day was something brewed up from the senna plant. I guess Rhubarb can be used in the same way. Apparently there are unpleasant side effects from these treatments so people were always looking for alternatives. Minerals were recognized for medicinal purposes and Elie Seignette figured that salt might work as a laxative.
Chemically salts are made by reacting acids and bases. I’ll be getting into more detailed chemistry, but for now consider the classic vinegar (the acid) and baking soda (the base) volcano science project. That reaction is generating bubbles of carbon dioxide as every school child can tell you, but it is also making a salt in the “lava” that pours down the side of the mountain.
In searching for a laxative salt, Elie eventually tried mixing cream of tartar (the acid) and sodium carbonate (the base). Cream of tartar is a byproduct of wine making process and there is plenty of that in France. Basically the wine tartrates form as crystals when the wine is subjected to cold storage. Figure 2B shows some crystals “wine diamonds” that formed on the cork of a bottle of red wine.
Sodium carbonate is a naturally occurring chemical you can actually dig up. Strangely, the ancient Egyptians used a substance they called Natron to preserve mummies and that was mostly sodium carbonate. As a side note here, the improbable chemical symbol for sodium is Na which owes its origin to Natron. The precise chemical name for the salt Elie made is potassium sodium tartrate.
Your guess is as good as mine for how the Elie figured out which salt “did the job”; maybe he experimented on himself. The important thing is that this one worked really well and apparently without negative side effects. An interesting characteristic of the salt is that when it dissolves in water, it absorbs quite a bit of energy. As a result, the liquid becomes noticeably cooler. Just like pharmaceutical companies do with drugs today, they came up with the fancy trade name; Polychreste Salt derived from the Greek language meaning “salt of many virtues.”
By the way, Elie’s son Pierre became a famous doctor in France and is often falsely credited with the discovery of Rochelle Salt. The whole Seignette family tree is very confusing because they tended to reuse first names a lot. For some reason Elie even had two brothers named Pierre as well as his son. One of the Seignette decedents who now lives in the Netherlands has assembled the Seignette family tree. There is a book, really a PHD thesis, written in 1910 that documents the history of le Sel Polychreste, but it is in French.
Polychreste was a huge success, and better yet, the Seignettes were able to keep the process for making it a family secret for over 60 years! They supposedly made a pile of money. Even after the secret was out, the drug was widely used for centuries. Figure 3 shows the paper wrapper from a package of Polychreste from 1760 or nearly 100 years after its discovery. The swan or Cygnus is an allusion to the name Seignette. Around the perimeter it says something like “All is done with salt and sun” in Latin. In Europe the drug was also known as Seignette’s salt. Figure 4 is a 1920 drug store label for Rochelle salt which is the name most commonly used in America. I’ll get back to 1920 in a minute.
Rochelle salt can be grown into large crystals and is a favorable material for scientific study. In 1824 Sir David Brewster noticed that the crystal produced electricity when heated or what is called pyroelectricity. Brothers Pierre (Madame Marie Curie’s husband, Figure 5) and Paul Jacques Curie were the first to identify that Rochelle salt was also piezoelectric in 1880. In other words, it generated electricity in response to being mechanically distorted. You can imagine that heating something probably also causes it to physically distort all by itself.
In fact, Rochelle salt is orders of magnitude more piezoelectric than just about any other substance. It remains to this day one of the most piezoelectric substances ever found. Perhaps not surprisingly, the process works in reverse where applying electricity to the crystal causes it to physically distort. Truly this is a salt of many virtues.
Pierre Curie was a professor at the University of Paris (College de Sorbonne) and had a physics graduate student named Paul Langevin (see Figure 6). On a rainy day in 1906 Pierre was accidently killed while walking across a busy street. A few years later Marie started having a love affair with Paul. She goes as far as sending a letter to Paul suggesting that he dump his wife and run away with her. Problem was, Paul’s wife intercepted the letter and made it public. It became a huge scandal because the older Marie was cast as a foreign Jewish home-wrecker who was tarnishing the good name of Curie. Albert Einstein thought the whole thing was hogwash and even wrote to Marie to tell her to simply ignore the press.
World War I broke out in 1914 and Paul Langevin started developing a method to detect German submarines. Basically it is what we now call sonar and consists of making a sound and then listening for the echo off objects in the water. Langevin used quartz, which is another piezoelectric material, but the idea is quite elegant since the same device can be used to both produce the sound and then pick it back up.
I had never heard of Langevin, but apparently he is a big deal in France. Below is a photo of me at the entrance to the park named after him. The sign reads; Situated behind the buildings of the former Polytechnique School, is the square named after the physicist and magnetism theorist Paul Langevin (1872-1946), who was a contemporary of Einstein and Bergson. Magnolias and ivy are the main plant species in the garden.
Eventually the United States was drawn into the war in April 1917 and work began on sonar technology here in America. A professor at Wesleyan University, Walter Guyton Cady (Figure 7), followed the quartz approach of Langevin, while a Scottish researcher working at Western Electric (later became Bell Labs), Alexander McLean Nicolson (Figure 8), pursued Rochelle salt. The war ended in late 1918 before any significant development of sonar had taken place, but the effort permanently moved piezoelectric phenomena out of the laboratory.
Cady went on to advance piezoelectric science and technology for the next 50 years. His seminal 1946 book “Piezoelectricity”, (Amazon:Piezoelectricity) has several good chapters on Rochelle salt. Cady and Nicolson were caught up in a long patent battle over who was the first inventor of the crystal oscillator. Nicolson eventually won legally, but most people give Cady the real credit for the discovery.
Nicolson worked with Rochelle salt for a few more years by developing audio related inventions like microphones and speakers. Starting in 1919, Nicolson wrote, or was the subject of, several articles about piezoelectricity. He was also eventually granted many patents for the production and application of Rochelle salt. Figure 9 shows him in one of the many public demonstrations he gave to a bunch of obviously fascinated engineers.
Here is a list of some of the articles I know about:
1. Piezo-Electricity, Telegraph and Telephone Age, April, 1919
2. A Brief Resume of some of Our War Work, Western Electric News, June, 1919
3. Making Crystals Speak, Popular Science Monthly, Sept, 1919
4. The Piezo-Electric Effect in the Composite Rochelle Salt Crystal, AIEE, Oct, 1919
5. Speaking Crystals, Electrical Experimenter, Dec, 1919
6. Crystals that Speak, Scientific American, June, 1920
7. Rochelle Salts to Energize Electric Current in Voice Transmission, American Druggist, Jan, 1920
8. The Voice in a Lump of Salt, Popular Radio, Sept, 1922
A crystal is assembled from molecules that act like little building blocks. Normally it takes many days to slowly grow a crystal from a solution of water and salt. This solution is known appropriately as the “mother liquor”. A molecule floating around in the mother liquor must align and stick to the crystal surface in just the right way or the crystal will be imperfect. Beautiful large naturally occurring crystals of something like quartz are the result of hundreds of thousands of years.
Figure 10 shows some of the crystals Nicolson grew. The final product was described as “looking like a dirty chunk of something like paraffin melted up with find sand.” That was because he grew his crystals using a highly accelerated process. The result was a composite structure and it had many voids and faults filled with mother liquor that spoiled the piezoelectric effect. He found that by soaking the crystals in alcohol for a few hours and then baking them at 40C a lot of the unwanted solution could be removed in a process called desiccation. Later investigators preferred working with perfect clear crystals and were appalled by his desiccation process.
Electrical Experimenter was a magazine (Figure 11) from the 1910s that is now famous for publishing essentially science fiction articles by Nicola Tesla. Nicolson’s “Speaking Crystals” article contained illustrated step-by-step instructions on how to grow Rochelle crystals at home. It reveals just how accelerated his growing process might have been.
The ingredient proportions were given as 8 parts Rochelle salt to 5.33 parts water. For 100ml of water that works out to 150g of salt. That is about 50g more salt than the water can actually hold at its saturation point at room temperature. That is also about six times more supersaturated than what is called for in the classic text Crystals and Crystal Growing by Alan Holden (Figure 12). We’ll come back to Alan Holden a little later.
The method consists of two crystallizations. The first crystallization (Figure 13) is only used to obtain a seed crystal. The second crystallization grows the seed into a much larger crystal. Despite the illustration showing a pan of rapidly boiling water, I would avoid taking Rochelle salt much above 50C because it can breakdown into other compounds. I’d also use a flexible plastic container (like disposable Tupperware) to grow the crystals because they have a nasty habit of adhering to glass. When selecting the seed you should pick one that is as square and flat as possible since it will have started growing in the alignment for maximum piezoelectric effect.
The solution is reheated to 50C to dissolve all the salt left after seed selection. The seed is “planted” in the mother liquor (Figure 14) when the temperature has dropped below 38C. The key is to put the seed in when the temperature is not so hot that it dissolves and not so cool that putting it in causes a blizzard of spontaneous crystal growth. The seed crystal should be started lying flat on the bottom, rather than on end like the illustration seems to show. I’d also cover the container to limit water evaporation since this solution is already overly supersaturated.
You’ll notice that the illustration calls for floating the seed on a layer of Mercury. Back in Nicolson’s day, Mercury was a common laboratory material. We now know it is extremely toxic and should normally be avoided. I think he was only using it to prevent the crystal from sticking to the bottom of the vessel and that can be accomplished by using a plastic container anyway.
The crystal grows in a manner of hours as the temperature of the liquor falls to that of its surroundings. Figure 15 is a photo of Nicolson’s crystal growing lab. The containers look like they are reasonably large porcelain glazed steel pots. It may be a little hard to make out, but each pot has a square glass plate covering it to reduce evaporation and to keep out dust. The lab assistant proudly holding a crystal in some tongs reminds me a little of Alfalfa the child star from the old Our Gang movies.
Nicolson’s earliest Rochelle salt patent (#1,414,370 filed April 2, 1918) shows that he sometimes elevated the seed crystals (Figure 16) on inverted drinking glasses. The piezoelectric effect is usually greater for thinner crystals, and he reasoned that the crystal would grow thinner near the surface since the salt would be more depleted there.
At the same time Nicolson was producing composite crystals, Roy W. Moore at General Electric’s Research Laboratory in Schenectady, NY, was working on a method to produce large perfect crystals. Nicholson’s pots look pretty unsophisticated compared to Moore’s apparatus shown in Figure 17. His crystals were grown within an electrically heated temperature controlled crock. In fact, the temperature was accurately controlled to within 0.1C by way of a modified mercury thermometer thermostat. His patent for the process (#1,347,350) was filed February 26, 1918 which was actually a few months earlier than Nicolson.
To fully appreciate what Moore was up to, you need to understand the Rochelle salt saturation curve in Figure 18. This plot shows the amount of salt that can be dissolved in 100ml of water verses temperature. For example, with 22C as a room temperature, 100g of Rochelle salt will dissolve in 100ml or water. That is how I knew that Nicolson’s recipe called for 50g more salt than it could hold.
That is not to say that you can’t temporarily get more salt into solution than the saturation amount. If the liquid was at 33C then 150g would completely dissolve in 100ml of water. Cooled slowly, the higher amount of salt could remain in the solution and it would become supersaturated. Eventually an impurity of some kind would trigger a spontaneous crystallization and the solution would become simply saturated over time.
The seed crystal also serves as a place where the extra salt can go to get out of solution. That is how Nicolson’s method worked, but the process can go too quickly for the molecules of salt to precisely buildup on the crystal surface. Starting with a less supersaturated solution slows the process down and that is why Holden and Morrison only call for 9g of extra Rochelle salt.
Another way to slow the process down is to slow the rate of cooling. When things naturally cool, the rate of temperature change is greatest in the beginning. Unfortunately that is exactly when you don’t have crystal surface to accumulate the salt. Cooling just 1C from 33C to 32C causes about 10g of salt to leave the mother liquor which is actually a decent sized crystal. Ideally the temperature should drop very slowly at first and then get faster as the crystal builds.
By accurately controlling the temperature, Moore was able to pace the amount of salt coming out of solution to the growth rate of the crystal. Starting at 40C, he would only let the temperature drop 0.1C in the first day and then increase the drop slowly every day. He produced crystals about the same size as Nicolson, but it took a month not hours.
Charles F. Brush was a minor Thomas Edison like inventor who made a lot of money in the late 1800s improving the carbon arc street light among other things. He had a namesake son who attended Harvard, MIT, and served in the army during World War I. When Charles Jr. returned from Europe, his father financed a startup research company for him. The intention of Brush Development Labs was to commercialize the metal Beryllium and Rochelle salt. His partner was a college friend, Charles Baldwin Sawyer (Figure 20) a brilliant chemist that had graduated from both Yale and MIT.
Sawyer’s first patent was filed in 1922 for the accurate thermostat he needed to produce perfect crystals like Roy Moore. However, Sawyer planned for much larger scale production than Moore. In 1927, just when Sawyer and Brush had perfected Rochelle salt mass production techniques, Charles Brush Jr. died due to a botched blood transfusion while attempting to save his dying daughter. Sawyer went on to manage Brush Development Labs which dominated Rochelle salt production for the next 30 years.
Sawyer used the controlled temperature drop technique of Moore, but added an agitation element to the process. Long seed crystals were put into tanks of mother liquor that slowly rocked back and forth. The result was a very long flat crystal. Figure 20 is a patent drawing showing the cross-section of the equipment while Figure 21 shows the seed crystal.
We even have a photograph of two young women working with the equipment in Figure 22. The reason they are so lightly dressed is the entire room would be started out at the elevated temperature of 40C (106F). It appears that they are ducking down between the steel rafters of an industrial building because the floor was probably raised to limit the air space that needed to be heated.
The results were spectacular large clear crystals (Figure 23) that were cut into much smaller pieces and used for phonograph pickups and microphones. Figure 23B shows one of the crystals cut and polished in cross section and the square seed is visible sticking out of the bottom. Parts of the crystal were also cut to make the seeds for the next batch. They were even used to make underwater transducers for sonar; the technology that brought attention to Rochelle salt many decades earlier.
Getting back to Alan Holden (Fig. 23C), the guy who wrote the classic book on growing crystals mentioned before. He was a chemist working at Bell Labs in the late 1930s and made some improvements to Rochelle salt by using something called heavy water instead of regular water. That is a special kind of water with Deuterium atoms in place of Hydrogen. This gave the material a little higher operating temperature, but not enough to save it commercially. Other materials with more stable properties were eventually developed and Holden worked during WWII to grow large crystals of one of the alternatives called ADP.
Getting Rochelle Salt
There are two ways to get Rochelle salt; buy it or make it. I highly recommend buying because you get high quality material and it is actually cheaper than making it. However, there is satisfaction in making it from scratch from chemicals you can find in most grocery stores. You will find the internet is full of instructions on making Rochelle salt this way, but just about every one is flawed in some way.
Rochelle salt is used in an alternative photographic printing process known as a Kallitype. I don’t know anything about that, but because the process involves Rochelle salt, a company called Photographers Formulary sells it (Figure 24). Better yet, it seems to be high quality stuff and is reasonably priced. Their website is www.photoformulary.com and you should probably buy a pound of it since it is way more economical that way. B&H Photo in New York City also sells this exact material from their website. You might be able to find other sources, for example you can buy it from Amazon:POTASSIUM SODIUM TARTRATE 1LB
Sodium Bicarbonate (Baking Soda)
The traditional recipe for making Rochelle salt calls for reacting Cream of tartar and sodium carbonate. Sodium carbonate (Na2CO3) is usually available as washing soda in the laundry isle at the grocery store since it’s used to “soften” water to make detergents work better. However, it isn’t as common as it used to be and I’m not sure how chemically pure it is. Some websites ironically give detailed instructions on converting baking soda (Figure 25), or sodium bicarbonate (NaHCO3), into sodium carbonate by baking it. However, that procedure is totally unnecessary. Sodium bicarbonate can be used in the reaction without any preparation and, at least the Arm and Hammer brand, is supposed to be 100% pure. If at all possible get a new box because once opened it can absorb moisture and that will throw your weight measurements off.
You never want to use water directly from the tap for chemistry. It can contain a lot of dissolved minerals and added chemicals used to make it safe to drink. Only use distilled water which is nearly pure H2O. Grocery stores carry it in same isle as other water products usually in gallon sized bottles. Distilled water is used for steam irons and humidifiers because it won’t leave deposits behind as it evaporates.
Cream of tartar
A hard crystalline crust called argol forms on the sides of barrels used to store wine for long periods. It is primarily Potassium bitartarate (KHC4H4O6) but there are contaminates and color picked up from the wine as well. It can be purified into Cream of tartar through a boiling and cooling cycle and then carbon filtration. Fortunately, Cream of tartar can also be purchased in the grocery store in the spice isle because it is used in cooking. McCormick (Figure 26) is probably the most common brand and it comes in a 74g bottle. Cream of tartar has a long history that dates back thousands of years and even has some interesting ties to the ancient art of alchemy.
If you are only going to create a single batch of Rochelle salt, then buying Cream of tartar in the grocery store makes sense. However, that is a pretty expensive route to take if you plan to make a lot. The more economical method is to buy it in bulk on the internet. For example, Amazon caries many brands in addition to larger containers from McCormick. I’ve tried the Frontier Bulk Cream of Tartar Powder, 1 lb. package (Figure 27) with good success, but the Sauer’s brand (Figure 28) is a waste of money because it is just too impure.
Actually the problem with most commercially available Cream of tartars is that they are not pure enough to make really good Rochelle salt. Manufactures add chemicals that keep it from clumping and deliberately dilute the product to make it cheaper. Using any Cream of tartar right out-of-the-package can result in a useless mess. Fortunately, it can be easily purified by a crystallization process similar to the one originally used to extract it from argol. Unfortunately, it means the loss of about 10% of an already expensive material.
Just like Rochelle salt, more potassium bitartarate can be dissolved in hot water than cold. You can get 6.2g into 100ml of water at 100C, but only 0.5g at 10C. Basically the purification process consists of boiling water with impure Cream of tartar, pouring off just the liquid, refrigerating it, and then collecting the pure material that crystallizes out. Sadly, you lose the potassium bitartarate still dissolved in the cold water but you do eliminate most of the nasty contaminants.
Here are the steps I use to purify a 74g bottle of Cream of tartar:
- 37g Cream of tartar + 600ml distilled water in a Pyrex container
- Bring to a rapid boil in a microwave oven
- Let stand 1 minute and then stir
- Bring to a second boil
- Let stand 1 minute and then stir
- Bring to a third boil
- Let stand 1 minute but do not disturb it this time
- Gently poor off just the liquid into a large bowl leaving any solid residue
- Repeat step 1 adding to the residue already in the container
- Repeat steps 2 to 8 but you can now discard all the residue
- Allow the bowl of combined liquid to come to room temperature
- Refrigerate the liquid for at least 10 hours down to at least 10C
- Occasionally stir just to knock down any crystals floating on the surface
- Gently discard just the liquid
- Keep the residue which is pure Cream of tartar
- Completely air dry, any moisture will throw off your weight measurements
The pure Cream of tartar forms as tiny beautiful glittery crystals. The amount you recover depends partly on how good your tartar was to start with. I typically lose about 12% by weight since contaminates are thrown out in the solid residue and also dissolved in the discarded liquid. Starting with 74g you might only end up with something like 60g of pure dry Cream of tartar, but that is still enough to make about 90g of Rochelle salt.
Before we start making Rochelle salt we should cover some basic chemistry. Chemists describe the reaction with the following equation:
- KHC4H4O6 + NaHCO3 + 3(H2O) = KNaC4H4O6*4(H2O) + CO2
- KHC4H4O6 = Cream of tartar (mwt 188)
- NaHCO3 = Sodium Bicarbonate or Baking Soda (mwt 84)
- 3(H2O) = Three water molecules (mwt total 54)
- KNaC4H4O6*4(H2O) = Crystalized Rochelle salt (mwt total 282)
- CO2 = Carbon dioxide (mwt 44)
The equation above means that for every molecule of Cream of tartar we need a molecule of Sodium bicarbonate and three of water. We will end up with one molecule of Rochelle salt crystal and one of Carbon dioxide. Obviously we can’t count molecules like this because they are too small, but we can measure weight.
Molecular weight is used to figure how much of each chemical we have and the wmt numbers are the individual molecular weights. Mixing 188g of Cream of tartar, 84g of Sodium bicarbonate, and 54g of water will make exactly 282g of Rochelle salt and 44g of Carbon dioxide. Any quantities with the same proportions work too. For example, if you started with half the Cream of tartar (94g), just half the rest of the numbers to balance the equation. If you have 60g of Cream of tartar then you need 60/188*84=27g Sodium bicarbonate and you make 60/188*282=90g of Rochelle salt.
It seems odd that each molecule of crystalline Rochelle salt includes four molecules of water to form the solid. This water is locked up in the crystal structure, but when heated to only about 56C (132F) is acts to dissolve it. Unfortunately this temperature is not all that hard to find on planet earth and anything employing Rochelle salt would be permanently ruined if ever exposed to it. That is the primary reason alternatives were eventually found that could survive higher temperatures. The other reason is that it readily dissolves in water which is also not hard to find on earth.
If you tried to perform the reaction exactly as written above you would have two problems. First off, the reaction needs energy and the chemicals need to be heated to about 70C to completely react. Secondly, that is the absolute minimum water needed to make the reaction happen. About five times as much water would be needed to provide a place for the reaction to properly develop. Taking all that into account, here is my recipe for making Rochelle salt from scratch:
- 60g pure Cream of tartar
- 27g Sodium bicarbonate or Baking Soda
- 82g Distilled Water
The reaction needs to take place at an elevated temperature, but definitely not boiling hot. The easiest way to do this is with a double-boiler (Figure 29) which is a small container setting inside a larger container of very hot water. Mix the Cream of tartar and distilled water together in a 250ml Pyrex container. Only a small amount will actually dissolve. Then set this into a large pot of water that is being heated on a stove. The level of water in the larger pot needs to be as high as possible without making the smaller container float. The temperature of the water bath should be kept below boiling.
When the temperature of the Cream of tartar mixture is up to 70C, start adding the Sodium bicarbonate a little at a time. Constantly stir the mixture and wait till the generation of Carbon dioxide bubbles dies down before adding more Sodium bicarbonate. Each time you add some, the temperature will drop a little and then recover. Adjust the stove heat to keep the temperature of the mixture at or below 70C because if it becomes too cool the reaction will take too long but too hot will allow unwanted chemical reactions to take place. Keep stirring and cooking the mixture till it becomes clear which can easily be half an hour.
You might find a small amount solid material that just won’t disappear. Allow it to settle to the bottom and gently poor off just the liquid into a plastic container. For example, I use the disposable plastic Tupperware package lunch meat comes in. The recipe above has been carefully designed to end up with a mixture that is about the same concentration of water and salt as the one in Holden and Morrison’s saturated solution.
For now, let me give you with a PDF of an article written by William J. Millard published in Elementary Electronics, May-June 1968 about growing crystals and making a small speaker. He titles his article “Singing Crystals” I assume in tribute to an earlier article written on Rochelle Salt in July, 1938, Popular Mechanics with the same title. Like most things about Rochelle Salt you find on the internet, other than this one of course, there is a significant mistake in Millard’s article that I should point out.
He has the mechanical direction of the piezoelectric effect wrong. It is not a twist as he shows but rather a shear. Figure 30a shows half of a crystal grown on a string, or a whole crystal of one grown on the bottom of a container. If you cut or grind down the crystal to the block shown as a dotted box, then you get the plate shown in Figure 30B. When the electric field is applied on the X axis (which is in and out the page), the plate shears to the right or left depending on the polarity of the field. In other words the crystal lengthens along one diagonal and shortens along the other. You could further cut a plate like the rectangle shown that would simply lengthen or shorten when the field is applied.
Someday I’ll get around to publishing my techniques for growing and cutting crystals to make a speaker.