March 31, 1997

Volume 69, Number 13

REPRINT

Analysis techniques for chelated minerals evaluated

The acceptance of chelated FIGURE
minerals by the feed industry
has long been hindered by
the lack of an assay, but a
new procedure, as presented
in this article, could change
all that and allow feed
companies to test their
purchases.
By G.A. LEACH and
RICHARD S. PATTON
Documentation of the effectiveness
of chelated minerals for animal nutri-
tion is not as daunting an assignment
today as in the past. Twenty years ago
chelates were considered to be lacking
in credibility by the mainstream intel-
ligentsia, and absurdly over priced for
production agriculture. Given the sup-
port data available at that time, this at- 1. Laboratory test method procedure
titude was not altogether irresponsible.
Less than 10 years ago an initial litera- FIGURE
ture review (Patton, 1990) was able to
locate only seven research journal ar-
ticles on chelates and nutrition, three
on human nutrition.
Insoluble fraction (%)

Today, it is possible to cite more than

50 refereed journal articles from the last
eight years that deal exclusively with
researching the topic of chelated min-
erals in animal nutrition. Additionally,
there are an equal number of abstracts,
and even more Extension Service re-
ports and company-sponsored reports Dry mix
of field trials.
Insoluble copper Insoluble protein
Chelate nutrition is becoming more
U,V,X,Y are metal proteinates, W is a metal amino acid chelate and Z is a metal amino acid complex
accepted, as more research comes to
light, but still various realities tend to
confuse a common- and broad-based 2. Copper feed supplements solubility in deionized water
understanding by the trade. Not all
companies market the same exact form in some cases, may be somewhat eso- research trials.
of chelate, with the Association of teric, and of more concern to the mar- On the other hand, some research
American Feed Control Officials defi- keters than the consumers. tends to bias results by using low
nitions (57.150, 57.151, 57.142, 57.29, Not all research is positive, which is bioavailable controls, such as copper
57.23) permitting different labels that, in keeping with the laws of biologic oxide or zinc oxide, for example, which
variance, and those with the courage to by comparison makes test treatments
■ Dr. G.A. Leach is a bio-inorganic publish negative results in a competi- look more acceptable. Lastly, in addi-
chemist with the Chelated Minerals tive market deserve our respect. Vari- tion to distinct species differences,
Corp. in Salt Lake City, Utah. Dr. Ri- ance between producers on quality/de- within any one animal, response to oral
chard S. Patton is a nutrition consult- gree of chelation may account for dif- mineral is invariably affected by cur-
ant who lives in New Mexico. ferences observed in different animal rent nutritional status; e.g., deficiency
© 2005 Feedstuffs. Reprinted with permission from Vol. 69, No. 13, March 31, 1997.
■ 2 — FEEDSTUFFS, March 31, 1997
or excess, recent repletetion, other min- FIGURE
erals in the diet or drinking water.
Acceptance of chelates has long been
hindered by the lack of an assay. A
buyer simply had no way of testing his

Fraction (%) of total copper

purchase, having to rely exclusively on
the reputation of his supplier and sub-
jective feedback from the field. The fi-
nal determination is almost solely in-
fluenced by cost per pound. Regula-
tory agencies, in turn, had no way to
protect the consumer or police charla-
tans. The reluctance of feed control of-
ficials, concerning chelates, was not
driven by a negative attitude toward
chelates as a nutritional concept.
T is a metal polysaccharide; U,V,X,Y are metal protenates; W is a metal amino acid chelate, Z
Their frustration has been that they is a metal amnino acid complex. Cu(Lys)2 & Cu(Met)2 are lab standards as per Hitchman (1987)
could not determine if the high price
one paid for a copper chelate was in 3a. Copper feed supplements bound and free copper fractions
actuality inexpensive copper oxide and
meat meal. Of course, feed officials need
to have just such an attitude but the FIGURE
absence of a test procedure was not in
any way a conspiracy as chelation
chemistry is a genuinely sophisticated
and intricate problem. Of significant
benefit to the industry is the emergence
in recent months of reliable test meth-
odology, a development that has been
long overdue.
In January of 1994, Brown and
Zeringue reported the first effort in the
technical literature of the feed industry
aimed at scrutinizing solubility and
structure of chelates. While adding in-
sight, the technique had low applica-
bility for the trade as it required chro- A,C,E are metal proteinates; B is a metal amino acid chelate; D is a metal polysaccharide and F is
matography equipment uncommon for a metal amino acid complex
feed mills.
In October of the same year, Parks and 3b. Zinc feed supplements bound and free zinc fractions
Harmston published a more attractive
possibility, in that it was simplistic and ker. The mineral product that stays be- changed in percent mineral and protein,
used inexpensive equipment, but was hind on the filter paper is insoluble (in this is a clear indication that the metal
also of somewhat limited utility. Still, water) and any mineral product that and protein were not bound together
these workers were really the first to passed through the filter paper into the by very strong forces, certainly not
break the ice, as subsequent develop- beaker is soluble. Assume that the origi- strong enough to stay bound in the
ments that built on their initiative have nal mineral product is 15% metal and physiological conditions of the diges-
resulted in a screening method that is 30% protein. tive system. Under this scenario, there
easy and useful. These values, percent metal and per- would be good evidence that the pur-
The understanding of these new meth- cent protein, are easily determined by ported chelated metal was not chelated.
ods requires a clear description of the atomic absorption and nitrogen Assume, alternatively, that the in-
concepts of soluble versus insoluble Kjeldahl assays found in most feed labs. soluble residue on the filter paper re-
and bound metal versus free metal ions. The first question becomes, what is tained its metal-protein profile of 15%
the percent metal and protein in the resi- and 30%. The next step in the process
Solubility due; the insoluble mineral left behind is to assay the filtrate, the liquid phase
Considering first the topic of solubil- on the filter paper; and similarly, what in the beaker.
ity, Figure 1 illustrates a simple flow of is the percent metal and protein in the If the proportions of the residue are
events in a laboratory. A powdered che- filtrate, the solution in the beaker? unchanged, then the proportions of the
lated mineral product is placed in a At this point in the process, it is not mineral product that went into solution
flask with water and mixed. Next the necessarily a poor indication if mineral will also be unchanged.
mineral-water combination is passed product remained insoluble on the fil-
through a filter. ter paper. Remember that water is not a Binding
Some mineral product may remain on duplication of gastrointestinal tract Now, the second question — what is
the filter paper and some probably dis- solubility. the nature of the metal in the filtrate in
solve in the water and flow into the bea- However, if the filtrate is significantly the beaker? Is it bound or free? In more
 March 31, 1997 FEEDSTUFFS — 3 ■
technical terms, is it dissociated from FIGURE
its ligand, in this case the protein, and
therefore a free ion or is it still bound to
its ligand? This question is aiming at
the same concept as before.
Free metal, unbound and ionically

Pounds required per ton of feed

charged in a mild solvent such as wa- lbs. required = 500/(%elemental x % bound)

ter, must be assumed to be only weakly

bound or chelated to its ligand, and in
poor likelihood of staying bound in di-
gestive fluids. This second question
can be answered by the use of an ion
specific electrode, a method commonly
employed by the feed industry for so-
dium determination.
The ion electrode measures only free
metal ions, the ones not bound but free
in solution. The ion measurement pro- Source of Zinc
A,C and E are 15% elemental zinc, B and F are 10% and D is 22%
tocol was as per the electrode
manufacturer’s instructions. If the fil-
trate in the beaker reveals, say, 15% 4. The amounts of different manufacturers’ products
metal via atomic absorption, and very FIGURE
little of it free, say 2%, as determined
by the ion electrode, then one can be
fairly confident that the metal is still
bound to its ligand. On the other hand,
if the ion electrode reveals a consider-
able portion of the metal as free, there
Cost (4) per ton of complete feed

is an indication of inadequate binding

of metal and protein, and there is justi-
fication in being skeptical of the de-
gree of chelation of the metal.
Returning to the issue of solubility,
it can be considered another way. If a
metal is bound to a protein, it should
stay with the protein in either the
soluble or the insoluble fraction. Where
separation occurs, e.g., most of the pro-
tein being insoluble and yet most of
the metal being soluble, this suggests
weak or no binding. A,C and E are 15% elemental zinc, B and F are 10% and D is 22%
Figure 2 depicts the insoluble frac- A=$1.80, B=$3.00, C=$2.06, D=$0.845, E$3.10, F=$1.60
tion of actual products from the mar- Cost per pound of mineral product based on manufacturesr’s suggestioned retail price
for a 50-lb. bag minimum order
ketplace, along with a dry mix of cop-
per and protein, when placed in deion- 5. The cost of 25 ppm totally bound zinc per ton of feed
ized water for 30 minutes. Again, high
or low solubility of a product per se in above procedure but brings us to the of unbound, free metal ions from a prod-
this system of deionized water is not next test or piece of the puzzle. As men- uct dissolved in water. It can be seen
the issue. A product may be of low solu- tioned, total metal content of the that the soluble fraction of different
bility in water and still be properly che- soluble fraction of a metal product can products evidence quite different de-
lated. What is the issue is that the level be determined by atomic absorption. A grees of bound versus free metal. In the
of solubility of each fraction, the metal measure of the amount of unbound, free case of products W, X and Y, fully one
and the protein, be identical, both ionic metal is given by the ion specific third or more of the copper is no differ-
equally high or equally low. electrode. The amount of bound metal ent than that from copper sulfate. It
As can be seen in Figure 2, there are (sometimes called complexed) is the would indeed be unfortunate to have
products in the market of widely vary- difference of these two values, total paid chelate price for this portion of
ing and dissimilar solubility. The ex- metal content minus unbound free ions. these products.
treme case is illustrated by the dry mix Figure 3a contrasts seven products from Please note that metal solubility and
of a mineral salt and protein. This sepa- the marketplace with two laboratory degree of binding are not related. Solu-
ration shows no binding, as would be standard chelates. The two chelate stan- bility does not measure binding. For
expected of a dry mix. dards, bis-(L-lysinato) copper(II) and example, the chelate bis-lysine
bis-(L-methioninato) copper(II) were copper(II) is very soluble, while bis-
Solubility plus binding prepared using the methods cited by methionine copper(II), also a chelate,
A totally soluble product (protein and Hitchman et al. (1987). is poorly soluble, yet both have only a
mineral) would not be screened by the The focus of Figure 3a is the amount very small measure of free, unbound
■ 4 — FEEDSTUFFS, March 31, 1997
copper. Solubility alone is a function (1995) noted short comings of this their dollar to receive the most chelated
of ligand and not a determination of method. Hydroxide precipitation also mineral. Using the simplistic laboratory
binding. The point is this: metal that ignores the insoluble portion of the pro- tests described here, that question too
becomes unbound in water is not che- teinate. It has merit as an indicator in can be answered.
lated. simple systems but certainly cannot be Figure 3b shows the bound and un-
Of course, the foregoing discussion used to give a definitive measure in the bound fractions for six zinc products
is simplified in the interest of clarity, case of a complicated proteinate. from the marketplace.
and these methods do require familiar- Potentiometry (Holwerda, ibid) is a Figure 4 takes these products and, us-
ity with standard laboratory procedures form of electrochemistry that is a well ing the results of total chelated zinc de-
such as calibrations, titration and gen- established, physical chemistry tool termined by ion specific electrodes, de-
eral elemental analysis, but none of this used widely in university laboratories. rives the pounds of proteinate product
is beyond the capability of a modern Its drawback — that it totally ignores per ton of finished feed required to de-
feed analysis facility. While the solu- the insoluble fraction — is com- liver 25 parts per million of totally
bility and ion specific electrode tech- pounded by the high sophistication it bound zinc. As can be seen, there is a
niques outlined above provide useful requires. As a technique for a feed labo- range from 0.347 to 1.263 lb.
and much needed new tools for the in- ratory, it would require unusual com- Figure 5 takes these inclusion rates
dustry, they are not a perfect system. mitment from management and staff. and calculates the cost per ton of add-
The emphasis with these methods is to Some would differentiate chelates on ing these amounts of zinc chelate prod-
present a screening ability to flag poor the basis of molecular weight, which uct to the finished feed, using the
products in a fair, reliable and efficient may indeed prove to be important, but manufacturer’s suggested retail price for
manner. If a product fails these tests, it it is not standard feed lab concern, and a 50-lb. bag. Depending on which prod-
is a poor chelate. It is possible for a prod- there is an entire body of evidence that uct is selected, one can pay from $0.625
uct to pass these tests, yet still be infe- a wide array of molecular weight prod- to $3.637 to achieve the same level of
rior. In other words, the strength of these ucts have effect. Solubility as a test to totally bound zinc inclusion. This is
methods is not in assuring quality, but delineate chelate quality is probably over a five-fold difference, and one that
as a screen for poor quality. Their value the least satisfactory of all, as men- buyers and nutritionists need to under-
is that they are acting as pieces to a tioned above regarding copper-lysine stand.
puzzle giving a true product profile. and copper-methionine chelates. Elec- In conclusion, a simplistic and useful
tron microscopy is really nothing more procedure to evaluate chelated miner-
Other tests than a photograph, and as a means to als is presented here for the feed
The discipline of chelate assay is new rank chelates for quality is no more use- industry’s consideration.
and evolving, and further improve- ful than snapshots of nails to reveal Application of this technique by
ments are on the horizon. One such pos- carbon content. the professionals in the trade is en-
sibility is visible spectroscopy, where X-ray crystallography or x-ray diffrac- couraged, with the predictable result
preliminary investigations show prom- tion can only be used on pure carefully that further informed dialogue will
ise (Leach, unpublished data). This grown crystals and has no application lead to modifications and improve-
method has the advantage over infra- on production samples. It can reveal ments. Meanwhile, at long last, there
red spectroscopy in that it probes the proof of chelation, but it is a university is a serious start on a chelate assay
metal rather than the ligand. technique that requires days of prepa- procedure.
So as ligands become more in num- ration, thousands of dollars and very
ber and more complicated, and infrared costly equipment. REFERENCES
suffers in utility, visible spectroscopy Ultra-violet absorption is highly de- Brown, T.F. and L.K. Zeringue, 1994. Laboratory
evaluations of solubility and structural integrity of
is capable of ignoring this and focus- pendent on the proteins and amino ac- complexed and chelated trace mineral supplements.
ing simply on the metal geometry. Other ids present, and at best can only show J. Dairy Sci. 77:181.
techniques deserve mention. that products are different. Hitchman, M.A., L. Kwan, L.M. Engelhardt and
A.H. White. 1987. Electron spin resonance and elec-
Separation methodology, such as It cannot measure degree of chelation tronic spectra and crystal and molecular structures
membranes and chromatography or strength of bonding. of copper(II) amino acid complexes. JCS Dalton 457.
(Brown and Zeringue, 1994) can only Holwerda, R.A., R.C. Albin and F.C. Madsen. 1995.
Chelation effectiveness of zinc proteinates demon-
be applied to solutions. Summary strated. Feedstuffs, June 19:12.
More importantly, separation disturbs Finally, the one question of concern Parks, F.P. and K.J. Harmston. 1994. Judging or-
the metal/ligand equilibrium and to all — the plant manager, the nutri- ganic trace minerals. Feed Management 45:35.
Patton, R.S. 1990. Chelated minerals: what
causes further, rapid, dissociation of the tionist, the customer and the feed con- are they, do they work? Feedstuffs, February
soluble chelated fraction. Holwerda trol official — is where does one spend 26:14. ■