A Whitening Toothpaste Built on Oxalic Acid The Claim Turns on One Cleaning Number and a pH Floor

This application covers a whitening toothpaste that uses an oxalate salt as the stain removal agent instead of peroxide or polyphosphate. It was filed by The Procter & Gamble Company, the owner of the Crest and Oral-B brands, and names Michael David Curtis, Andrew Frederic Groth, Paul Albert Sagel, Samuel James St John and Marianne Zsiska as inventors. It is a divisional of Australian application 2024227094, which is itself a divisional of 2021268189, the national phase entry of PCT/US2021/030751 filed on 5 May 2021. The claims are considerably narrower than the title suggests, and the narrowing is the interesting part.

The Problem

The specification opens with a survey of whitening agents and what is wrong with each of them. Oxidising agents such as peroxide are hard to keep from reacting with the other ingredients over the life of the tube, and they are simply not reactive with some surface stains. Abrasives can damage the tooth surface and cannot reach the places where stain sits, such as between the teeth. Polyphosphate anticalculus agents hydrolyse in the tube into orthophosphate, which does nothing, and in the presence of soluble fluoride the breakdown accelerates and produces insoluble fluoride. Bicarbonate pastes, the specification says bluntly, taste like baking soda.

Underneath that list is a chemistry argument. Tooth stain is described as a matrix of proteins and polysaccharides cross linked by metal cations, which then acts as a reservoir for coloured porphyrins and other chromophores. Break the electrostatic cross links and the stain lifts. That makes the job a ligand problem rather than an oxidation problem, and the oxalate anion is a strong bidentate ligand for exactly those cations.

The trap is that the same property cuts the other way. Oxalate pulls calcium out of the enamel mineral to form insoluble calcium oxalate, and until that insoluble phase forms it lowers the local saturation of the enamel surface with respect to calcium. At low pH and low calcium, the specification states, applying oxalate can measurably soften the enamel. The inventors say this demineralisation was discovered unexpectedly and had not previously been disclosed in the art, and it is why the claim carries a pH floor.

What This Invention Does

Claim 1 is a whitening dentifrice composition comprising a dicarboxylic acid that comprises oxalic acid or a salt of it, sodium monofluorophosphate, an abrasive that is a silica abrasive or a calcium abrasive or both, and a pH of from about 4.5 to about 6, wherein the composition has a pellicle cleaning ratio of at least about 140.

That is four requirements, and three of them are tighter than anything in the title or the abstract. The title says only “oral care compositions comprising dicarboxylic acid”. The abstract and the summary describe a pH of about 4 to about 6 and a pellicle cleaning ratio of at least about 120, and they also describe two part kits and a whitening regimen that the claims of this divisional do not cover at all. The claim as granted in this document requires a specific fluoride source, monofluorophosphate rather than the stannous or sodium fluoride used in most of the worked examples, and it lifts the cleaning threshold from 120 to 140. Readers should follow claim 1, not the abstract.

The pellicle cleaning ratio, or PCR, is the industry laboratory measure of how much stain a paste lifts off a stained bovine enamel chip compared with a calcium pyrophosphate reference slurry. The specification uses the Stookey method as refined by Schemehorn. Higher is better.

The numbers in Table 4 are what carry the case. Crest Pro-Health Advanced Whitening, used as the positive control, scored 140 plus or minus 22. Crest Cavity Protection, the negative control, scored 84 plus or minus 10. A paste with stannous fluoride and stannous chloride but no oxalate scored 109 plus or minus 16. Adding 3.14 per cent potassium oxalate monohydrate to that same paste lifted it to 141 plus or minus 20. Sodium fluoride plus oxalate reached 163 plus or minus 23 and oxalate with no fluoride at all reached 146 plus or minus 21.

The monofluorophosphate example, Example 5, is the one that matters for claim 1, because it is the only formulation in the document that contains the claimed fluoride source. It is 1.15 per cent sodium monofluorophosphate, 3.00 per cent potassium oxalate, 12.00 per cent silica abrasive, 1.50 per cent silica thickener, 5.50 per cent sodium lauryl sulfate solution, 0.55 per cent phosphoric acid and 49.90 per cent sorbitol solution. Its PCR is 139 plus or minus 12. That sits just under the “at least about 140” the claim requires, which the word “about” presumably has to absorb.

Key Features

• Oxalate as the whitening agent. The compositions of Table 1A carry 3.14 per cent potassium oxalate monohydrate in a base of 45 to 48 per cent sorbitol, 17.5 per cent silica and 7 per cent sodium lauryl sulfate solution. The oxalate replaces peroxide and polyphosphate rather than supplementing them.
• A pellicle cleaning ratio floor of about 140. Claim 1 defines the composition partly by its performance, not only by its ingredients. Note that dependent claim 31 recites a pellicle cleaning ratio of at least 120, which is broader than the claim it depends from and appears to be a drafting slip left over from the parent.
• A pH window of about 4.5 to about 6. Stain removal improves as pH falls, so the formulation wants to go low, while enamel safety sets the floor. The specification states plainly that at pH of about 4.5 an oxalate toothpaste was found to damage enamel relative to the water control, and it still nominates 4.5 as the lower bound.
• Sodium monofluorophosphate specifically. Most of the worked examples use stannous fluoride with stannous chloride or plain sodium fluoride. Claim 1 requires monofluorophosphate, which is matched by only one example in the whole document.
• Silica or calcium abrasive. Where a silica abrasive is used the claim restricts it to silica gel, precipitated silica, amorphous precipitated silica or hydrated silica. The calcium option in dependent claim 20 covers calcium carbonate, calcium pyrophosphate, calcium phosphate and hydroxyapatite.
• Stain prevention measured separately. In the in vitro pellicle tea stain model the oxalate plus sodium fluoride paste showed a 24 per cent improvement over the fluoride-only control, and the oxalate-only paste a 34 per cent improvement, while oxalate made no difference to the stain caused by tin ions themselves.

Who Is Behind It

The applicant is The Procter & Gamble Company of Cincinnati, the owner of Crest, Oral-B and Blend-a-med, and the controls used throughout the experimental section are its own retail products. One inventor is worth naming. Paul Sagel is a P&G Research Fellow and the lead inventor of Crest Whitestrips, which P&G describes as the largest product introduction in the company’s history, and he was awarded the Society of Chemical Industry’s Gordon E. Moore Medal in 2007. This application sits in the same whitening programme.

The chain of filings is set out in the specification itself, unusually clearly. This application is a divisional of Australian application 2024227094, which is a divisional of 2021268189, which is the Australian national phase entry of PCT/US2021/030751, published as WO 2021/226153 and filed on 5 May 2021. The specification does not name the earliest priority document. From the published family record, rather than from this specification, the PCT claims priority from United States provisional application 63/020,033 filed on 5 May 2020, which makes the priority country the United States.

The supporting data is all laboratory work. There is no clinical trial in this document: the PCR values come from brushed bovine enamel chips, the stain prevention numbers from in vitro biofilms grown on hydroxyapatite discs, and the safety numbers from extracted human enamel cores measured with a Vickers hardness indenter under a 50 g load.

Why It Matters

Whitening toothpaste is a large and crowded category in which most of the chemistry is a century old. Peroxide, abrasives and polyphosphates all have known limits, and formulators have been stuck choosing among them. A cheap food acid salt that works as a chelating ligand rather than an oxidant is a genuinely different lever, and the data here suggests it does the job: in the same base, adding oxalate moved the cleaning ratio from 109 to 141.

The more interesting content is the safety boundary, because the specification publishes the result that cuts against the invention. In the enamel softening study the monofluorophosphate and oxalate paste at slurry pH 4.56 produced a change in surface microhardness of minus 42.42 against minus 15.02 for water and minus 175.68 for 1 per cent citric acid. In other words, it was gentler than citric acid but statistically worse than water. The powder stain removal model tells the same story from the other direction: a 3 per cent oxalate solution at pH 4.5 gave a brightness change of 10.79 on tea stain, falling to 2.92 at pH 5.5 and 2.78 at pH 7. Cleaning power and enamel safety pull in opposite directions along the pH axis, and the claimed window of 4.5 to 6 is where the applicant has decided to stand.

That is also why the performance limitation is in the claim. A composition that hits a cleaning ratio of 140 without dropping below pH 4.5 is a narrow target, and defining the claim by the test result rather than only by the recipe is how the applicant fences it. For anyone tracking the dental pellicle whitening field, the numbers in Tables 3, 4, 5 and 7 are more informative than the claim language.

Related Concepts

• Dicarboxylic acid – the chemical class the claimed whitening agent belongs to, with oxalic acid as its simplest member.
• Tooth whitening – the product category this composition competes in, dominated by peroxide and abrasive systems.
• Tooth enamel – the mineral surface whose softening sets the lower pH bound in the claim.
• Hydroxyapatite – the calcium phosphate mineral used as the stained substrate in the powder stain removal model.
• Chelation – the ligand effect the specification relies on to explain why oxalate lifts stain.

---

AU 2026202215 was published in the Australian Official Journal of Patents on 9 April 2026 and is open for public inspection. Patent applications represent inventions that are sought to be protected and do not necessarily reflect commercially available products.