Electronic Calcium Monitoring by Randy Holmes-Farley

Aquarists are often looking for ways to simplify aquarium husbandry. Among the less interesting tasks involved in keeping aquaria is the testing of water parameters. Nevertheless, testing for such attributes as the aquarium's pH, calcium and alkalinity levels is frequently useful, and sometimes critical. Among these big three chemical parameters, pH is among the easiest to measure, as high quality pH probes and meters have been around for many years. Not only are they highly accurate, they can be left to read the pH in the aquarium in real time, even when the aquarist is absent.

Alkalinity measurement has not become this automated, and likely won't be. Alkalinity isn't a chemical in the water, but instead is a value resulting from a pH titration that gives information about bicarbonate and carbonate levels in reef aquaria. Perhaps someday alkalinity testing will be replaced by a more direct measure of bicarbonate, but in the meantime alkalinity testing is simple and straightforward using commercial test kits.

That leaves calcium. Most aquarists measure calcium with a traditional calcium test kit. These are sold by a number of companies, including Hach, LaMotte, Salifert, and Seachem. The technology to measure calcium with ion selective electrodes has been around for many years, but it traditionally has been very expensive, and its implementation has suffered from concerns over calibration and interferences. Recently, American Marine has added an electronic calcium monitor (using ion selective electrode technology) to its Pinpoint line of electronic monitors.

In this article I put this monitor through a few tests that should be helpful in assisting aquarists determine whether or not they want to use such a monitor. The article is broken up into several sections:

Ion selective electrodes consist of a membrane that responds electrically to the presence of specific ions in solution. In a sense they are similar to pH electrodes that are sensitive to H⁺, but these respond to other ions instead. In the case of calcium selective electrodes, the membrane material is typically PVC (polyvinylchloride) that has been modified to be selective for calcium. That is, only calcium ions can enter and potentially pass through the membrane. This modification typically adds calcium-binding organic functional groups to the membrane. The calcium is attracted to these groups much as it is attracted to organic small molecule chelators. These functional groups can be designed to be specific for certain cations based on size and other properties, so the membrane can be tailored to be calcium selective. It should be noted here that each manufacturer may choose to make the membrane differently, so do not assume that all calcium selective electrodes have similar selectivity.

As with pH electrodes, the molecular level details of how these electrodes function are surprisingly complicated and poorly understood. When in use, calcium enters the membrane and may pass through it. This movement sets up an electrical potential between the outside of the membrane (where the calcium enters it) and the inside (where there is no calcium to start with). Since the solutions on both sides of the membrane are initially electrically neutral, and only calcium enters and passes through the membrane, the inside becomes positively charged relative to the outside. This electrical potential is related to the number of calcium ions that move into or through the membrane, and the number of calcium ions moving into it is related to the calcium concentration in solution. Because of these relationships, it turns out that the electrical potential is directly proportional to the logarithm of the calcium concentration in the test solution.

Calcium selective electrodes are used with a meter which, when properly calibrated, converts this measured electrical potential into a calcium concentration. The calibration itself essentially assigns a calcium concentration to the measured electrical potential in the calibration solutions, and the meter then uses these values to assign a calcium concentration to all other possible electrical potentials. Like many pH meters, calibration consists of placing the meter into two solutions of known concentration. In the case of the Pinpoint calcium monitor, these are 100 and 1000 ppm calcium in a seawater mimic. The procedure is easy: put the probe into the first solution and push a button. Let it equilibrate until the value flashes (a few minutes), and then repeat in the second calibration solution. The meter then provides a real time reading of calcium levels in the solution, although it may take a reasonable amount of time to yield a completely stable reading (see below).

For those interested in more technical details of ion selective electrodes (including the whole issue of internal reference electrodes that are also necessary for measurement), there are a number of good online articles:

Figure 1. The meter portion (left) and the electrode (right) of the Pinpoint calcium monitor.

This sensing mechanism highlights some potential problems with this method. The first is in the membrane's selectivity for calcium over other ions of interest in the sample. Fortunately for reef aquarists, the ions that interfere most intensely with these sorts of calcium selective membranes (e.g. lead, mercury, iron, copper, ammonia) are normally present at much lower concentrations than calcium (calcium is typically 250-600 ppm in reef aquaria, while these others are usually much lower than one ppm). However, the two main cations in seawater, sodium at >10,000 ppm and magnesium at 1300 ppm, can also potentially interfere by getting into the membrane.

Additionally, materials in solution can impact the membrane in other ways. Organic materials may bind to it, changing its electrical potential. The ionic strength may also impact the electrical potential experienced by the membrane. Finally, the calcium in solution can itself be impacted by other ions in solution. Some of the calcium ions in seawater are paired to sulfate, chloride, carbonate, bicarbonate, phosphate, and fluoride. This ion pairing tends to decrease the likelihood of calcium entering the membrane by reducing the concentration of "free" calcium. Typically, only 10-20% of the calcium in seawater is ion-paired, but this amount may vary between aquaria with different concentrations of these other ions, and so can be a source of variability. This effect is significant enough that the calibration standards need to be made in a solution similar to the sample (that is, similar to seawater). The Pinpoint standards are made in just such a way, and in a later section of this article, I show that without such ions in solution, the calcium concentration reading provided by the meter is significantly off.

Organics also can bind strongly to calcium, producing a similar effect. In natural seawater the concentration of organic material is too low to impact the free calcium concentration significantly. In reef aquaria the concentration of organic materials may be higher, but I have not seen any detailed measurements of organic levels in such systems. In testing shown later in this article, however, the effect of added organic material was found to be insignificant.

Temperature is another possible source of variability. Temperature affects the penetration of the calcium ions into the membrane, as well as the speciation of the calcium in solution. In practice, the effect of temperature can be large, with the signal from many ion selective electrodes changing by more than 4% per °C. Consequently, it is very important to account for temperature somehow (such as by calibrating at the temperature of the water to be measured). The magnitude of this effect is shown experimentally in later sections of this article.

There are many ways to test an ion selective electrode for use in reef aquaria. The most obvious way is to make a standard solution and see if the measurement is correct. That is, however, much more easily said than done. Without knowing exactly what factors are most important to the proper or faulty functioning of the electrode, a simplified solution may not reflect real use.

For initial testing I chose to use as the "standard" a sample of artificial seawater that was mixed to an approximate salinity of S=35. I mixed a 44-gallon batch using Instant Ocean artificial salt mix and reverse osmosis/deionized (RO/DI) water to a conductivity of 52.7 mS/cm, and allowed it to settle for three weeks. I then proceeded to measure its calcium concentration by ICP-AES (inductively coupled plasma-atomic emission spectroscopy, an $80,000 analytical instrument. I was somewhat disappointed with my inability to use this sophisticated technique to get a precise answer. Despite taking five different samples and analyzing them at eight different emission wavelengths using two different calibration methods (five standard additions of known calcium concentrations to each sample, as well as comparison to a fixed 1000 ppm commercial calcium standard), I was unable to get consistent values. Some of the samples were acidified or filtered through submicron filter membranes to determine if solid materials were impacting the result (they were not). Overall, I took more than 200 measurements, each involving three replicate observations of the emission intensity. Nevertheless, the result was not very satisfying, with a substantial variation occurring between the different values. The average of every measurement taken was 336 ppm. With the uncertainty involved, however, I'd conclude that the true value was probably 340 ± 40 ppm. I also measured the same sample once with a Salifert brand test kit and got 330 ppm calcium.

I tested a variety of samples for calcium with two Pinpoint calcium monitors that had been calibrated within 48 hours of testing. These included all of the samples tested by ICP (the acidified ICP samples had a pH <0.7, and are not totaled below but are discussed later in this article). I also tested additional samples drawn from the 44-gallon container. I obtained the following values (after waiting sufficient time for the values to stabilize, five minutes to 24 hours):

Meter 1
380.2 371.5 363.1 371.5 363.1 355.0 380.2 346.7 363.1 363.1
Avg. = 366 ppm standard deviation = 10 ppm

Meter 2
346.7 380.2 380.2 371.5 354.8 354.8 346.7 346.7 331.1 331.1 363.1
Avg. = 355 ppm standard deviation = 17 ppm

It should be noted that despite reading to 0.1 ppm, the meter is clearly not that precise in its reported values. The concentration on the meter's readout moves in much larger units, more like 10 ppm. So in a solution with an increasing calcium concentration, the readout rises in blockwise steps, such as 340.0, 350.4, 360.8, 371.2, etc. Readers can draw their own conclusions based on the uncertainty in the standard and the variation between measurements, but in my opinion, in this test the monitor is sufficiently accurate for most reef aquarium purposes.

In order to determine if the calcium monitor is correctly responding to calcium levels, I also took a single sample and spiked it with high purity calcium chloride (99%) to ensure that it responded appropriately. I spiked a half-gallon (1.89 L) of the Instant Ocean salt mix with a standard solution containing 33,800 ppm Ca⁺⁺. I obtained the following values:

Plotting the measured calcium against the volume of solution added yieldsa slope of 18.3 ppm per mL (fit by linear regression). The actual slope should have been 17.9 ppm per mL. So the meter was slightly over-responding to calcium, but again, likely close enough for reef aquarists and probably within the noise of the testing (especially considering this was a single measurement and that the meter does not respond in small increments, but rather makes jumps of about 10 ppm).

It is not easy to decrease the amount of calcium in such a sample, but one way to do it is to dilute the sample with pure (RO/DI) water. Too much dilution will move the sample outside the range of salinity experienced by aquarists, but a 10% decline in salinity is easily within that range, and so is a useful test to determine whether salinity changes are important to such measurements. Toward this end, I took one liter of the same Instant Ocean standard used above, measured it with each of two meters, and then diluted it with 100 mL RO/DI water. I obtained the following results (after 43 minutes equilibration time):

Ideally, the samples would have declined in calcium concentration by 9.1%. Within the uncertainty of these measurements, especially the response time effects (see below) and the fact that the meter does not respond in small increments, but rather in jumps of about 10 ppm, the drop seems appropriate.

Figure 2. A scene from one of my reef aquaria, showing organisms that are sensitive
to calcium levels (calcifying corals, a clam, coralline algae), and others that are
less sensitive (fish, certain soft corals, etc).

A number of factors determine the electrodes' response time. As shown below, temperature effects are large, so if the electrode is moved between samples of different temperature, the response time may become very long. However, the electrodes also have a natural response time to chemical changes that is not insignificant. In the experiment above, when I diluted the sample with RO/DI water, I tracked the values over time, and got the following results:

Consequently, it is clear that the exact answer provided by the electrode is quite dependent on how long the electrode is allowed to equilibrate. This effect may be partially due to drift in the electrode, but mostly it is genuine equilibration, because samples that have been equilibrated for several hours rarely change much.

pH can significantly affect calcium selective electrodes. Some commercial calcium selective electrodes are claimed to cover a pH range from 3.5 to 11. I tested a commercial calcium standard (1000 ppm in 2% HNO3) whose pH was below 1, and found that indeed the electrode does not give accurate results at that pH (at least using the Pinpoint standards). The two meters I tested claimed the calcium concentration was 1.9 and 16.0 ppm, rather than the known 1000 ppm. Likewise, I tested the electrodes in saturated limewater, with a calcium concentration a bit above 800 ppm. One showed 1738 ppm, and the other 645 ppm. Consequently, limewater potency measurements are probably not a useful application for these meters, probably due to the high pH.

At pH ranges more likely to be encountered in reef aquaria, however, the effects are more modest. To test these effects, I raised and lowered the pH of the Instant Ocean standard above with 1 M (mole/L) hydrochloric acid (HCL), 0.05 M sulfuric acid (H2SO4), and 0.1 M sodium hydroxide (NaOH). I obtained the following results:

While there is some up and down noise, the values were consistently within ± 10%, regardless of the pH being between 7.0 and 8.63. From these results, I conclude that pH within the range normally experienced by reef aquaria is not a significant factor in calcium determinations by the ion selective electrode. This method may be effective to use when measuring the effluent of calcium carbonate/carbon dioxide reactors, where the pH is low but the bicarbonate level is very high. However, I have not tested it under those conditions.

One of the concerns with ion selective electrodes is the selectivity of the electrode for the ion of interest, especially relative to similar ions. For calcium measurements, the effects of magnesium and sodium should be considered, not so much because the electrode is not selective over these ions, but because they are present at such high concentrations. In general, it is important to use a seawater matrix that contains these other ions when making standards. This is clarified by testing a simple calcium chloride solution in pure (RO/DI) water. I made a 545 ppm calcium solution using high purity calcium chloride, and got 812.8 and 776.2 ppm from the two meters that I tested. Obviously, the other ions have some effect, and need to be accounted for in the standards.

But how much impact would "normal" ionic variation have in an analysis of reef aquarium water? To test this I spiked the Instant Ocean standard (1.89 L) with a variety of inorganic salts and got the following results:

That amount of sodium chloride effectively raises the salinity from 35 ppt to 46.6 ppt, well above what any aquarist would encounter. It is not clear whether the observed rise in the calcium reading is from the change in sodium and chloride concentrations, or whether the sodium chloride contains a trace of calcium, but in any case, the increase is reasonably small. With typical salinity variability being half that or less, the variation in calcium reading (about 15 ppm) is not significant, and salinity changes are not likely a concern for aquarists using this device.

I performed a similar experiment with magnesium sulfate (99.8% pure and containing only 0.02% calcium) and found the following:

This addition was sufficient to raise magnesium by 360 ppm (a 28% increase) and sulfate by 1400 ppm (a 52% increase). Since sulfate is one of the main ions that ion-pairs with calcium, it is not surprising that such a big rise in sulfate might decrease the observed calcium value. In fact, the observed drop is consistent with the known ion pairing constant of calcium with sulfate. In normal seawater, 10-15% of the calcium is paired to sulfate. If sulfate were to rise by 50%, the ion pairing to calcium ought to rise by a similar amount, and so one might have 15-23% paired. This increase in pairing of 5-8% drops the free calcium by a similar percentage, and is consistent with the observed decrease of 4.5%.

Another similar experiment was performed with magnesium chloride containing a portion of calcium chloride:

This magnesium chloride (manufactured by the Dead Sea Works) is used by some reefkeepers to supplement magnesium. By the manufacturer's specifications, it contains 11.5% magnesium and 0.7% calcium. This amount of calcium would raise the sample by about 37 ppm (after 10 g of addition), as well as raising magnesium by 608 ppm (a rise of 48%) and chloride by 1800 ppm (a rise of 9%). The observed rise of 44 ppm fits well, despite the large perturbation in the level of magnesium and chloride.

It turns out that temperature plays a very big role in many ion selective electrode measurements, including the calcium electrode sold by Pinpoint. In fact, when I first calibrated the Pinpoint calcium monitor and tested a sample of my reef aquarium water, I got a value of 700 ppm. For a number of reasons, I knew this result was unlikely. It turns out that if the calibration solutions and electrode are colder than the tank's water, the monitor will read artificially high.

To more carefully test this effect, I measured the calcium level in a number of samples that I warmed or cooled. For example, in one test I took a sample of aquarium water, put it in a plastic container, and kept it in my refugium overnight (to keep it at a stable, warm temperature) with the two calcium electrodes in it (both calibrated with solutions and electrodes that were at my ambient basement temperature in March of 60°F). Both read 660.7 ppm. I then removed the plastic container from the refugium and allowed it to cool, took measurements, and then returned it to the refugium and allowed it to warm again. I got the following results:

I repeated this sort of test several times, and found drops in calcium on cooling of 24%, 28%, 28%, 17%, 22%, and 17%, which all came back to about the starting point if the samples were warmed back up. The samples may not have all cooled down equally, as I did not monitor the temperature of all of them, so the differences above do not necessarily reflect testing variations but may simply be differences in the total temperature change.

This temperature effect is, in my opinion, the biggest concern with using such a system. Aquarists must somehow ensure that the calibration is done at a similar temperature to the measurement. If there are any temperature changes, allow a significant amount of time for the electrode's AND the solution's temperature to stabilize. This control can be accomplished by calibrating at tank temperature, or by allowing the tank sample to cool to room temperature before measurement (by sitting for a few hours to overnight, for example). A difference of 1°F is not likely to be significant, but 5°F may cause a mismeasurement by several tens of ppm calcium. A difference of 20°F can make a difference of 100-200 ppm calcium.

Many aquarists would hope to use such a calcium monitor the same way that many of us use pH electrodes, with the monitor in the tank water and possibly providing continuous readings. I have not tested whether there are problems with this procedure, but Lou Dell, owner of American Marine/Pinpoint, has recommended against it out of concern for fouling the electrode membrane. Some aquarists using the probe report it to be fairly stable for a week or so under such conditions, but it then needs to be recalibrated. If used continuously, plan to recalibrate it fairly frequently, and then to recalibrate it less and less frequently based on how much it seems to have drifted between calibrations, and how accurate the readings need to be.

These calcium selective electrodes are curiously sensitive to water movement. Perhaps the moving ionic solution puts some sort of charge on the membrane. Stirring a test solution with the electrode can drop the value by 50 ppm or more, and it takes some time (a few minutes) to stabilize. So do not place it in a highly turbulent location if it is placed directly into an aquarium system. Some water motion, however, is useful, especially at the start of a measurement to ensure that the water in contact with the electrode's surface represents the whole solution.

Organic materials have the potential to impact calcium sensing in two main ways. The first is by binding directly to the membrane, altering its potential, or fouling it so that calcium cannot enter. The second is by binding calcium in solution, making it less likely to enter the membrane. Strong binding of calcium by the chelator EDTA, for example, can make it completely unavailable for sensing by the membrane.

In order to test how likely these issues are to matter to reef aquarists, I started with the sample of artificial seawater made using Instant Ocean that resulted from the pH studies above, and added calcium chloride (part of another study). I then added skimmate from my aquarium's skimmer equaling about 2% of the liquid's volume, in order to simulate the possible effect on calcium sensing of a significantly higher level of organic materials. Two percent was chosen fairly randomly to simulate a high organic level without altering the water so much that it wasn't likely to be encountered by aquarists.

Over the course of 30 minutes, I noted no change in the calcium reading, indicating that any effects due to organics are likely to be unimportant for most reef applications.

My conclusion based on limited testing of two Pinpoint calcium monitors, and from the information that I have received from several other aquarists who have them, is that they can be a useful way to measure calcium, but are not without issues that aquarists must become familiar with. Given the time to calibrate and equilibrate them to each new solution, such a monitor is not faster than a test kit. Since I have not tested the accuracy of any calcium test kits, I cannot compare the accuracy of the calcium monitor to that of typical commercial kits.

I was pleasantly surprised by the fact that the response does not seem likely to be thrown off by the ranges of different water chemistries encountered in reef aquaria. Temperature seems to be the only big concern, and that is something that aquarists can control.

The monitor is warranted for two years, and it is recommended to replace the electrode after 18-24 months, or when it can no longer be calibrated. Pinpoint has replaced monitors for some folks that had problems (unable to calibrate). All of the measurements I took were using the 9-volt battery, although one of the meters did come with an AC adapter.

The section below gives some tips for using the Pinpoint calcium selective electrode in reef aquaria. Pinpoint includes many other recommendations in their User's Guide, which is well-written and easy to follow. I have listed here only those things not listed there, or which I consider most important:

Disclosure: American Marine/Pinpoint supplied at no cost the two calcium meters that I tested.