Ion Exchange: The Role of Cross-Linking and Aging
By C.F. ‘Chubb’ Michaud, MWS
The role of cross-linking (x-l)
The level of DVB (cross-linking) a resin has determines the tightness of the resin. Higher cross-linked resins have lower moisture content, meaning more plastic and higher capacity. Think of higher x-l as a neighborhood with bigger houses and more people with narrower streets—it can be more difficult to get around.
With the gel resins used in softening, the standard resin has a DVB level of eight percent. This has good kinetics (even in cold water) and decent resistance to oxidation (chlorine damage). Reducing the x-l to six percent gives excellent kinetics in cold water with moderate flows, adequate softening efficiency and easier iron removal on regeneration but not much resistance to oxidation from chlorine. The resin will soften and deform a bit at elevated temperatures, leading to higher pressure drop and bead breakage. They are okay for most well water applications but not too good for a chlorinated city supply.
Bumping to 10-percent x-l, the resin will exhibit about 10-percent higher capacity in warmer water (>80°F) but higher leakage on cold water systems (<40°F). The higher x-l resin has superior oxidation resistance for residential applications and is becoming more prevalent in areas that combine warmer water and higher chlorine or chloramine. Macro resins represent the ultimate in x-l and can be as high as 20 percent DVB. These resins are designed for higher-temperature applications (~250°F+) and highly oxidative environments. Mid-range DVB (15-16 percent) represents a good compromise for extreme conditions that might be encountered with a residential or commercial application (~150°F) with high chlorine. Figure 1 shows the relationship between the cross-link level of a gel resin and the resulting moisture and total capacity. The moisture content of macro resins is difficult to measure because of the residual free moisture that can hide out in the pore structure.
Because higher x-l means more plastic and less water within a given bead, it explains why the volumetric (wet) capacity for that bead is higher. As previously pointed out, however, the kinetics for such resins suffer because of the increased bead tightness. Operating capacity, which takes into account the kinetics of the resin, may actually be lower for a more highly x-l resin when used improperly. What is often overlooked with higher x-l resins is that they also have higher charge density and therefore, demonstrate higher selectivity; in other words, a stronger attraction to certain contaminants (see Figure 2).
In Figure 2, the selectivity for hydrogen (H+) has been equated to a value of 1.00 in order to show the relative increase for the selectivity of other ions as the x-l increases. That’s good to know for operating in the H+ form. But what happens when you are using a salt (Na+) form of the resin to remove heavy metals? Note that the selectivity for sodium increases as well. So unless the selectivity for heavy metals increases at an equal or higher rate than that of sodium, it would appear that lower x-l is better.
Let’s do the math. For lead (Pb++) as we go from 4 to 8 to 16 percent, the selectivity ratios (selectivity Pb++/selectivity Na+) go from 4.14 to 5.03 to 7.60. It would appear that the selectivity for lead increases with an increase in the DVB level. For copper, the ratios go from 2.08 to 1.95 to 1.88. It would appear that copper removal is better done with a lower x-l resin. Comparing copper (Cu++) to potassium (K+), the ratios decrease from 1.45 to 1.32 to 0.99. This tells us not to use KCl as the regenerant for the removal of copper. Increasing the x-l of a resin will increase its selectivity across the board but you have to be selective as to which form of the resin to run. In the case of copper versus potassium, the higher x-l example actually shows a slight selective preference for potassium.
How is cross-linking level defined?
Resin cross-linking is often referred to as its nominal cross-link level. The reference alludes to the actual level being somewhat of a range rather than an exact level. But what does 10-percent DVB actually mean? It means that in the ideal world of theoretical chemistry, every tenth styrene unit in the backbone is a DVB, evenly spaced in perfect order and distribution. That is the textbook version. Some manufacturers define a 10-percent cross-link as the amount of DVB added to a batch as a percentage of the styrene monomer; i.e, 10 pounds DVB per 100 pounds of styrene. The textbook, written by William Bornak, says no; it’s the ratio of DVB to total monomer or 10 pounds DVB per 90 pounds styrene. The ultimate authority, Dr Friedrich Helfferich (Ion Exchange, p. 35) says no; it’s the molal percent (meaning one-tenth of the molecules) are DVB. Since DVB has a MW of 134 versus styrene at MW = 104, that means 134 parts of DVB per 1,070 parts of total monomer. Doing the math, these three definitions calculate to 9.1, 10 and 12.5 percent. The corresponding moisture retention of these three resins would be approximately 46, 43 and 39 percent. Why would a manufacturer use a lower cross-link level? There may be idiosyncrasies in each manufacturing plant regarding process conditions that make variation of the DVB level necessary. This may be due to heat-exchange limitations, agitator speeds, batch size, etc. The fact that DVB costs 4.5 times as much as styrene, however, also has a bearing.
How about capacity versus particle size of the resin?
It’s understandable that larger, more uniform beads will have larger spaces in between them and contribute less to pressure drop than will a finer mesh resin. If resins of different size have the same x-l level, they will have the same moisture content and the same wet and dry (total) capacity. Why then is fine-mesh resin touted as having a higher capacity? The answer is kinetics. Since the operating capacity of a resin is based on use conditions, kinetics comes into play. It has been shown(1) that larger beads are both slower to react in service and slower to react in regeneration. Slow reaction rates mean decreased kinetics. This is shown in Figure 3.
Figure 3 shows the impact of bead size on kinetics. If you took a regular-size distribution resin and split it in half to make a coarser grade and a finer grade, the coarser grade (+16-25 mesh) would show about 20-percent lower operating capacity than would the finer mesh (+25-50 mesh) portion. In a typical Gaussian distribution of +16-50 mesh, the average bead size is about 600µ (microns). The largest beads are 1,200µ (twice as big) and the smallest are 300µ (twice as small). What is the relative kinetic ratio of these beads? In a paper(1) presented in 2000, it was shown that the kinetics for reaction rates of resins varies with the square of the ratio of the diameters. In other words, if it is twice as big, it is four times slower. In the case of the extreme spread of 1,200 to 300 m, the ratio is four and the kinetic difference is 16. The larger beads that sit on the bottom of the bed and see the weakest brine in downflow regeneration are 16 times slower to regenerate than the top of the bed, which also sees the strongest brine. One must take this into account by increasing the brine level, increasing the brine time, or switch to upflow regeneration. Low brine settings will hurt you. Cold water will kill you.
As resins age
Resins age in a number of ways. Some will lose capacity due to degradation of the reactive sites.(2) This is particularly true for Type I SBA resins—they simply break down. Some will go through a partial decomposition, such as the conversion of a quaternary amine to a tertiary amine (thus converting a strong base site to a weak base site). This happens with Type II SBA resins. Note that with Type IIs, the total capacity may not change by much but the salt-splitting capacity is lower. This makes Type IIs ideal for general purpose DI float. The higher initial capacity of the Type IIs is another reason. SAC resins are very stable over time and do not lose much capacity (on a dry-gram basis).
All resins, however, are subject to oxidation, which effectively lowers the level of cross-linking and allows the beads to swell over time. Swollen beads are weaker and subject to higher levels of breakage. Broken beads are subsequently washed out to the column during backwash. This resin loss may not be apparent because the lower x-l level allows the resins to swell. The column may appear full even though it has lost 20 percent of its functional volume.
A third method of degradation is from osmotic shock. When resins regenerate, they shrink and swell. This is much like bending the proverbial paper clip back and forth. With repeated flexing, some beads will break and get washed out. Mechanical breakage may also occur if the resins are subjected to mechanical shear with repeated transfers from one vessel to another or, heaven forbid, pumping the resin back and forth. Centrifugal pumps are resin grinders. You can use a pump with a recessed impeller or a diaphragm pump for moving resin but the best methods are hydraulically or with an eductor.
A final method of resin degradation is fouling. Contaminants are picked up by the resins that are not fully regenerated off without special effort. Eventually, they will build up and block ion-exchange sites. For SAC resins, the common ones are iron, barium or aluminum. SAC do not usually get organically fouled as do SBA resins. SBA resins can become hardness-fouled when run on hard water. This is a result of incomplete regeneration in waters containing high sulfates or carbonates. They can be silica-fouled when not properly regenerated in DI systems using warm caustic. SBA resins are good organic scavengers and can become organically fouled with natural organics such as tannins. These are large molecules that are slow to elute off the resin. There are specific remedies(3) to clean these resins and restore capacity that will not be covered in this article.
Do resins wear out?
The biggest threat to SAC softening resins is oxidation from chlorine or chloramine. As a general rule of thumb for 8-percent resins, the number 10 divided by the chlorine residual in ppm would predict the life expectancy. Example: 1.5-ppm chlorine: 10/1.5 = 6.67 years. A 10-percent resin might go 10-12 years under the same conditions. Note: life will be shorter in heated water (~140°F). The presence of iron or copper in the feed-water can catalyze the de-crosslinking rate of the resin and cut the life in half or more.(2) I have seen small softeners on new copper plumbing and used for softening heated water swell so much in six months’ time so as to block the flow of water through the system. This is a good application for macro resin.
We’ve said a lot about the benefits of higher x-l resins but what happens when the resin ages and starts to unzip? Moisture increases, size increases, beads begin to break, charge density decreases, leakage goes up, capacity comes down, weight density decreases (resin gets backwashed out more easily), it softens, pressure drop goes up, more beads break, more salt is needed to maintain capacity because of resin loss. Eventually, the extra cost of salt for regeneration surpasses the cost of re-bedding the system. Can we predict where we are on this slippery slope? Yes. A simple residual moisture test can give a good indication of the degree of x-l remaining (see Figure 4).
Starting with new 8-percent SAC softening resin with 47-percent moisture, follow the red curve to the right. As the residual x-l decreases to around six percent, the moisture increases to 54 percent and the useful life remaining is targeted at 4-5 years. By the time the moisture reaches 60 percent, the effective x-l is down to about four percent and the life remaining is 2-3 years. Is the resin still performing? Yes, but it is struggling. Figure 5 shows what the physical properties of the resin might be.
Let’s look at the nine-year mark outlined in red. The resin moisture is up to 57.1 percent. The total capacity has dropped over 21 percent. Note from Figure 6 that the operating capacity has dropped from 19.4 kgr/cu ft to 16.8, a decrease of 13.4 percent. If this softener was regenerated every day with 8 pounds of salt/cu ft, it would now require an additional 391 pounds of salt per year. At a nominal cost of $0.15/lb, this means each cubic foot of resin will cost an additional $58.65/yr to operate. This may be more than the cost of new resin but the performance is dropping in other ways (see Figure 6).
Not only has the capacity dropped but the average leakage has crept up to over 1 gpg. Bead strength is about half of the original value and pressure drop is starting to creep up. Note: with a bead strength of around 125 g/bead, you can usually crush beads between your fingers—not easily but it can be done. This suggests the resin is over 60 percent moisture and is near the end of its useful life. This is a field test that can save a lot of guesswork, time and money.
How long do resins last?
There is no point in the life of resin where it stops working altogether. Failure is gradual and over a fairly long period of time…even years. SBA resins in DI service are good for 4-6 years. The regenerant is more expensive than salt so the capacity loss shows up on the bottom line sooner. The resin will continue to perform for another 4-6 years but at increased cost and loss of performance. Salt-regenerated SBA (such as uranium and/or nitrate reduction) can easily perform well for 8-10 years in normal service. Periodic cleaning with citric acid to remove hardness scale will prolong the service life.
SAC resins in DI processes should give 6-8 years of service for two-bed use and 5-6 years for mixed-bed applications. Mixed-bed polishers might give only 3-4 years of life due to the loss of anion function and not the cation. SAC resins used in commercial softening, where leakage tolerance is minimal, should give 6-10 years of useful life and residential applications should perform well for 12-15 years, if chlorine levels do not exceed 1 ppm and iron is not an issue.
As resins age, they unzip and lose their zip. They cost more to run and produce less volume and lower quality. Like an older employee, there is a time to say goodbye. We have shown a simple test that can help predict when the resin should be replaced. Not everybody regenerates every day. Not everybody pays only $0.15/lb for salt. The results are site-by-site and case-by-case. But if you can squish the resin in your fingers, it is time to replace it.
- Michaud, C.F. and Sabzali, J. A shortcut to higher regeneration efficiency with shallow shell resins, a paper presented at the 2000 Cambridge University Ion Exchange Conference and published by Imperial College Press, 2000.
- Michaud, C.F. and Brodie, D. “Ion Exchange Resins-Methods of Degradation,” WC&P, January 1990.
- Michaud, C.F. “Causes and Treatment of Fouled Ion Exchange Resin,” WC&P, March, 2006.
- Fries, Bill. Purolite Company, Philadelphia, PA. Unpublished research data, 1999.
- Bornak, William E. Ion Exchange Deionization, p. 71, 2003.
About the author
C.F. ‘Chubb’ Michaud, Master Water Specialist is the Technical Director and CEO of Systematix Company of Buena Park, CA, which he founded in 1982. He has served as chair of several sections, committees and task forces within WQA, is a Past Director and Governor of WQA and currently serves on the PWQA Board, chairing the Technical and Education Committees. Michaud is a proud member of both the WQA and PWQA Halls of Fame, has been honored with the WQA Award of Merit and is a two-time past recipient of the PWQA Robert Gans Award. He can be reached at (714) 522-5453 or via email at AskChubb@aol.com.