Yes, solubility can be used to identify matter, but only as a supporting property rather than a fingerprint. A substance's tendency to dissolve in water, ethanol, or hexane narrows the candidate list fast, sometimes from thousands of possibilities to a handful. It cannot confirm identity alone, because dozens of unrelated compounds share nearly identical solubility values, and the same compound can shift tenfold with temperature, pH, or crystal form. Used as the first screen before melting point, infrared spectroscopy, or mass spectrometry, solubility is one of the cheapest and most discriminating filters available to a chemist. It is a class test, not a name test.
Why Like Dissolves Like: The Mechanism Behind the Test
Dissolution is an energy trade. A crystal holds itself together with lattice energy, and the solvent must pay that cost before it can surround the freed particles. For sodium chloride the lattice energy is about 787 kJ per mole, while hydrating the released Na+ and Cl- ions returns roughly 769 kJ per mole. The 18 kJ per mole gap is covered by entropy of mixing, which is why the salt still dissolves freely at room temperature.
Molecular solids follow the same logic through hydrogen bonding and dispersion forces. Sucrose carries eight hydroxyl groups, giving water plenty of handles to grab, so the sugar dissolves to about 200 g per 100 mL. Naphthalene carries no polar groups at all, so water has nothing to hold onto and the hydrocarbon stays out.
The shorthand "like dissolves like" is a heuristic, not a law. A sharper version uses cohesive energy density, reported as the Hildebrand solubility parameter in units of (MPa)^0.5. Water sits near 47.9, ethanol near 26.5, toluene near 18.2, and hexane near 14.9. A solute usually dissolves well when its parameter falls within roughly 2 (MPa)^0.5 of the solvent's value, which is why a δ of 20 predicts good solubility in acetone and poor solubility in water.
Dielectric constant supplies a second, independent handle. Water's dielectric constant is 78.4 at 25 °C, so it weakens the Coulomb pull between ions by that factor. Hexane's value is 1.9, leaving ionic lattices essentially intact. That single number explains why table salt reaches 35.7 g per 100 mL in water and practically nothing in hexane.
In plain terms: solubility measures a tug-of-war between how tightly a substance holds itself together and how strongly the solvent can pull it apart.
How Wide Is the Solubility Range? A Quantified Comparison
The span from "soluble" to "insoluble" covers nearly six orders of magnitude, which is exactly what makes the property a useful filter. At 25 °C, sucrose reaches about 200 g per 100 mL of water while silver chloride manages 0.00019 g per 100 mL. That is roughly a one-million-fold ratio in the same solvent at the same temperature.
Close chemical relatives diverge just as sharply. Barium sulfate dissolves to 0.00024 g per 100 mL, while calcium sulfate reaches 0.24 g per 100 mL, a thousandfold gap between two alkaline earth sulfates. One measurement therefore eliminates enormous sections of the periodic table before any instrument is switched on.
Solubility also separates substances that look identical in a vial. Iodine gives a faint brown tint in water at 0.03 g per 100 mL but dissolves freely in ethanol. Naphthalene reaches only 0.0031 g per 100 mL of water yet dissolves to roughly 5 g per 100 mL of ethanol. A solvent swap turns an apparently insoluble powder into an obviously soluble one.
| Solvent | Dielectric constant at 25 °C | Hildebrand δ (MPa)^0.5 | Dissolves best | What a positive result implies |
|---|---|---|---|---|
| Water | 78.4 | 47.9 | salts, sugars, amino acids | ions or strong hydrogen-bonding groups present |
| Ethanol | 24.5 | 26.5 | resins, moderate polar organics | polar neutral molecules with some hydroxyl content |
| Acetone | 20.7 | 20.0 | ketone- and ester-like organics | moderate polarity without strong hydrogen bonding |
| Toluene | 2.4 | 18.2 | aromatic and nonpolar organics | absence of strong polar groups |
| Hexane | 1.9 | 14.9 | waxes, fats, alkanes | hydrocarbon character, no ionic content |
For nonpolar organics, solubility is predictable enough to calculate. The Yalkowsky general solubility equation estimates log aqueous solubility from melting point and log Kow, with typical error near half a log unit. That accuracy is good enough to sort candidates into bins, and far too coarse to name an unknown.
The Official Scale: From Very Soluble to Practically Insoluble
Pharmacopeias converted this continuum into named bands so that a reported value means the same thing in every lab. The U.S. Pharmacopeia defines "very soluble" as less than 1 part solvent per 1 part solute, and "practically insoluble" as more than 10,000 parts solvent per 1 part solute. Converted to laboratory units, that band runs from above 100 g per 100 mL down to below 0.01 g per 100 mL, so 10,000 parts of solvent is simply shorthand for 0.01 g per 100 mL.
The middle bands carry most of the diagnostic weight. "Freely soluble" covers 1 to 10 parts solvent, or roughly 10 to 100 g per 100 mL, where sodium chloride sits at 35.7. "Soluble" spans 10 to 30 parts, about 3.3 to 10 g per 100 mL. "Sparingly soluble" covers 30 to 100 parts, and "slightly soluble" covers 100 to 1,000 parts, which places calcium sulfate at 0.24 g per 100 mL.
The bottom two bands are where confusion starts. "Very slightly soluble" runs from 1,000 to 10,000 parts solvent, or 0.1 to 0.01 g per 100 mL. "Practically insoluble" means more than 10,000 parts, below 0.01 g per 100 mL, and both silver chloride and barium sulfate land there.
Practically insoluble is not zero. A substance at 0.00024 g per 100 mL still produces a saturated solution of about 1 × 10^-5 moles per liter, which is enough to drive precipitation reactions and enough to detect by conductivity. Chemists who treat the label as "none" misread the scale by several orders of magnitude.
The Laboratory Workflow, Step by Step
The reference approach is the shake-flask method, described in the OECD 105 guideline from 1995 and still the benchmark against which faster screens are validated. Its accuracy depends on reaching true equilibrium rather than merely observing that a solid disappeared.
| Step | Operation | Time required | Controlled variable |
|---|---|---|---|
| 1 | Dry and weigh the sample | 10 minutes | moisture below 0.5 percent |
| 2 | Add solvent and seal the vessel | 5 minutes | 25 °C ± 0.5 °C |
| 3 | Shake or stir to equilibrium | 24 to 72 hours | 200 rpm, protected from light |
| 4 | Settle, then filter at 0.45 µm | 15 minutes | discard the first 10 percent of filtrate |
| 5 | Quantify the dissolved fraction | 30 to 90 minutes | HPLC with UV detection, or gravimetry |
| 6 | Repeat at a second temperature | 24 hours | 4 °C and 37 °C for comparison |
| 7 | Compare against reference band tables | 5 minutes | match within one pharmacopeial band |
Elapsed time from weighing to answer is typically 48 to 96 hours, not the 20 minutes a demonstration suggests. Skipping step 3 is the most common error, because fine particles that pass a 0.45 µm filter can inflate apparent solubility by 2 to 5 times. Filtering cold, at 4 °C, limits that artifact and also captures the temperature dependence in a single run.
Can Solubility Be Used to Identify Matter? Limits and Edge Cases
Solubility struggles wherever the solid phase can change without the molecule changing. Polymorphs are the classic case: identical chemical formula, different crystal packing, and aqueous solubility differences that routinely reach 2 to 10 times. Amorphous material can appear 10 to 100 times more soluble than the stable crystal, largely because it has no lattice to pay for.
Ionization bends the number even further. For a weak acid with a pKa of 4.5, solubility is roughly 11 times higher at pH 5.5 than at pH 4.5, and roughly 1,000 times higher at pH 7.5. Salt selection does the same for bases, so a hydrochloride salt may exceed the free base by two or three orders of magnitude in water.
Temperature and particle size add two more moving parts. Sucrose climbs from about 200 g per 100 mL at 20 °C to 487 g per 100 mL at 100 °C. Particles below 1 µm show measurably higher apparent solubility through the Ostwald-Freundlich effect, which matters for nanocrystals and milled powders.
The sharpest limitation is stereochemistry. Enantiomers have identical solubility in any achiral solvent, so a solubility test cannot distinguish (R)- from (S)-ibuprofen no matter how carefully it is run. Two unrelated compounds can also coincide within one band by chance, which is why the property filters rather than identifies.
Safety Flags and Misconception Checklist
- Never report a solubility-based identification as confirmed. Treat it as a class assignment that a second, orthogonal method must verify.
- Beware supersaturation. A solution holding two to five times the equilibrium load looks stable for hours and gives a falsely high value.
- Record pH, ionic strength, and solvent grade alongside the number. The same compound can move two bands with a buffer change alone.
- Watch the common ion effect. Sodium chloride in 0.1 M brine falls well below its pure-water value, so the solvent is never just water.
- Do not compare values measured at different temperatures. A 10 °C shift moves many sugars and salts by 10 to 30 percent.
- Treat concentrated acids, chlorinated solvents, and fine powders as hazards. Weigh dusty solids in a fume hood and quench acid dilutions slowly.
What to Actually Do: A Practical Identification Protocol
Start with three solvents chosen to span the polarity range, not one. Water, ethanol, and hexane separate ionic, polar neutral, and nonpolar character in a single pass, and the pattern of three results is far more informative than any single number. A substance soluble only in hexane points toward hydrocarbons and waxes, while solubility in water plus insolubility in ethanol suggests a salt.
Control temperature at 25 °C ± 0.5 °C and state it in your notes. Run the measurement for at least 24 hours with agitation, and confirm equilibrium by sampling at 24 and 48 hours; if the two values differ by more than 5 percent, the system is still dissolving. Report the result in g per 100 mL alongside a pharmacopeial band, so the figure survives translation between labs.
Pair the result with one spectroscopic method before drawing any conclusion. Melting point narrows the field within a class, and infrared or mass spectrometry supplies the molecular fingerprint that solubility cannot. When the stakes are high, add a second solvent system and a temperature run, which together distinguish polymorphs and salt forms from genuinely different compounds.
Use the general solubility equation as a sanity check on nonpolar unknowns. If the predicted log solubility and the measured value disagree by more than one log unit, suspect an ionizable group, a hydrate, or an impure sample rather than an exotic new substance. That single comparison catches most bad measurements before they reach a report.
Finally, keep the property in its proper place. Solubility answers "what kind of substance is this" reliably and "which substance is this" only weakly, and the difference is worth stating plainly whenever a result is handed to someone else. A solubility test costs under USD 30 in consumables and an afternoon of bench time, which is why it remains the first move rather than the last.