The Cyclohexanone vs Cyclohexanol: KA Oil Ratio & Separation problem is controlled by two interacting variables: the upstream ketone-alcohol ratio generated in the oxidation or hydrogenation reactor and the downstream distillation train required to put a 99.8 wt% ketone cut into storage. KA oil, a condensed mixture of cyclohexanone and cyclohexanol, emerges from cyclohexane air oxidation with residual cyclohexane, water, organic acids, esters, and heavy ends. The normal boiling-point gap is narrow: cyclohexanone at 155.6 °C and cyclohexanol at 161.1 °C, a difference of 5.5 °C. Caprolactam-grade cyclohexanone is routinely specified with cyclohexanol below 0.05 wt%; adipic acid operations may accept the mixed KA-oil stream because nitric acid oxidation converts both components. The separation is therefore stage-limited rather than azeotrope-limited in most commercial designs. A shift in the KA oil ratio alters the load on the cyclohexanone column, the alcohol-dehydrogenation unit, and the heavy-ends purge, which is why the ratio is controlled before the first distillation tower rather than left to downstream blending.

How Does the KA Oil Ratio Respond to Oxidizer Design and Catalysis?

Cyclohexane oxidation is deliberately operated at low per-pass conversion, typically 4–8 wt%, to preserve selectivity to cyclohexyl hydroperoxide and its alcohol-ketone decomposition products. The reaction runs at 140–170 °C and 0.8–1.5 MPa with a cobalt naphthenate or cobalt octoate catalyst. Under these conditions, the mass ratio of cyclohexanone to cyclohexanol in the resulting KA oil can be shifted from approximately 0.4:1 to 0.8:1 by adjusting oxygen partial pressure, catalyst loading, and gas-residence time. Boron-assisted oxidation, in contrast, stabilizes the peroxide intermediate and raises cyclohexanol selectivity to 85 wt% or more of the KA oil, which suppresses direct ketone recovery and increases the need for downstream dehydrogenation. The cyclohexanol dehydrogenation reactor operates over copper-zinc or copper-chromium catalysts at 220–300 °C with single-pass cyclohexanone yields of 70–80%. Process control uses gas chromatography as the primary measurement; ASTM D4052 density and ASTM D1209 color are insufficient for ratio control because the two components are close in density and colorless. The unconverted cyclohexane is flashed, condensed, and recycled to the oxidizer, while the crude KA oil passes to saponification for ester hydrolysis and acid neutralization.

The saponified KA-oil stream enters a three-column purification sequence. A cyclohexane recovery column strips the low-boiling solvent overhead; a water-removal column removes water and light oxygenates; and the main cyclohexanone column takes the ketone overhead while cyclohexanol exits as bottoms. The main split is difficult because the relative volatility of cyclohexanone to cyclohexanol at atmospheric pressure is approximately 1.15–1.25. For a distillate of 99.9 wt% cyclohexanone and a bottoms of 0.1 wt% cyclohexanone at α=1.2, the Fenske total-reflux minimum is 76 theoretical stages. At α=1.15, the minimum rises to 99 theoretical stages. Commercial columns are therefore specified with 100–150 theoretical stages and reflux ratios of 10:1 to 20:1 to account for tray efficiency and feed-composition swings. Structured packing rather than random packing is often used in the upper section because its lower pressure drop reduces the reboiler temperature and its higher efficiency per unit height shortens the column. The alcohol bottoms are then directed either to a dehydrogenation reactor or to a second vacuum column for cyclohexanol recovery. Published data for this specific configuration is limited; the stage count above is a design estimate from standard Fenske analysis rather than a guarantee from a single operating plant.

Thermal sensitivity of cyclohexanol in low-pressure rectification

Cyclohexanol undergoes acid-catalyzed dehydration to cyclohexene at elevated reboiler temperatures, and cyclohexanone undergoes alkali-catalyzed aldol condensation when residual caustic from saponification is not neutralized. The acid number of crude KA oil before saponification can reach 2–5 mg KOH/g; after saponification it is typically reduced to 0.01 mg KOH/g or lower. The main column is operated under vacuum, with overhead pressure maintained at 10–30 kPa so that the reboiler film temperature remains below 130–150 °C. Above 180 °C in the hot zone, cyclohexanol conversion to cyclohexene increases measurably and creates a recycle impurity that is difficult to purge. Falling-film or forced-circulation reboilers are specified instead of kettle reboilers to reduce residence time and film temperature. The pH at the column feed is controlled between 6.5 and 7.5; operation above this band has been associated with rapid fouling of the stripping-section packing by aldol resin, which raises pressure drop and reduces stage efficiency. The same fouling mechanism can shift the effective reflux ratio and cause off-spec cyclohexanol breakthrough into the overhead cyclohexanone product. Heat-recovery exchangers on the feed and bottoms are designed for low wall temperature rather than maximum energy recovery to avoid thermal degradation.

When Phenol Hydrogenation Sets a Different Ketone-to-Alcohol Baseline

In the phenol hydrogenation route, the KA-oil ratio is inverted toward the ketone. Supported palladium or nickel catalysts hydrogenate phenol at 120–180 °C and 0.2–2 MPa hydrogen partial pressure, yielding cyclohexanone with reported selectivity above 95%. The crude product contains 2–5 wt% cyclohexanol and trace unconverted phenol, so the separation sequence can be a single high-purity distillation column with an alcohol purge, rather than the full KA-oil train required after cyclohexane oxidation. The operating risk shifts from distillation stage count to catalyst deactivation: as hydrogenation selectivity falls, the cyclohexanol content in the reactor effluent rises and the column must be run at higher reflux. This route still requires the same thermal care as the cyclohexane route because the cyclohexanol purge stream can dehydrate in the bottom loop. The alcohol purge is usually sent to a copper-based dehydrogenation unit operating at 220–300 °C, where single-pass conversion to cyclohexanone is 70–80%, with hydrogen recycled to the hydrogenation step. The comparison between cyclohexanone and cyclohexanol in the phenol route is therefore not a separation-limited problem but a selectivity-management problem.

On-line ratio measurement is complicated by the melting point of cyclohexanol at 25.4 °C. Sample lines and analyzer cabinets are heat-traced above 40 °C to prevent solidification. Process gas chromatographs provide a cycle time of 3–5 min and report cyclohexanone, cyclohexanol, cyclohexene, water, and light-ends simultaneously. ASTM D4052 density is used as a secondary check, but it cannot resolve small composition shifts because the two components differ in density by less than 0.02 g/cm³. The measured KA-oil ratio is used as a feed-forward signal to the cyclohexanone column: an increase in cyclohexanol content raises the bottoms viscosity and reboiler duty, while a decrease raises the overhead distillate rate and may pull alcohol into the ketone cut if the reflux is not adjusted. Batch-to-batch variance from the oxidizer can shift the column feed composition by ±5 wt%, so control loops that use only tray temperature are insufficient. Production sites therefore place the process analyzer on the saponified, water-free feed, not on the raw oxidizer effluent, because entrained water and organic salts interfere with GC injection and shorten column life.

Vapour-liquid equilibrium data and column stage requirements

The table below consolidates the physical-property inputs used to specify the cyclohexanone–cyclohexanol split. All values are from standard reference property data; the test methods are those commonly applied in release and in-process control.

The relative volatility calculated from these vapour-pressure data lies between 1.15 and 1.25 over the high-purity portion of the column. Applying the Fenske equation to a 99.9 wt% overhead and 0.1 wt% bottoms at α=1.2 gives 76 theoretical stages at total reflux. Reducing α to 1.15 increases the minimum to 99 stages. At a tray efficiency of 0.7 and finite reflux, actual stage counts of 110–150 are expected. This is why a simple two-column cyclohexane-to-KA-oil distillation unit cannot produce caprolactam-grade cyclohexanone from a high-alcohol feed; the stage demand is beyond the economic height of a single tower. The column overhead pressure is selected to keep the reboiler below 150 °C. For a cyclohexanol-rich bottoms, that usually means an overhead pressure of 20 kPa or lower, depending on tray pressure drop and reboiler fouling allowance.

Release specifications for caprolactam-grade cyclohexanone are written around direct impurity limits, not the bulk KA oil ratio. Representative limits are shown in the following table.

Adipic acid units may accept KA oil with cyclohexanol in the feed, but storage and transfer facilities must still account for the freeze point of cyclohexanol. Tank temperatures below 25 °C can create a separate solid phase and alter the circulating KA oil composition. The cyclohexanone vs cyclohexanol ratio in the final product is therefore maintained by a combination of upstream conversion control, high-stage vacuum fractionation, and tight release limits on the alcohol impurity. When the ketone column feed is not analyzed in real time, the first measurable symptoms are usually a rising reboiler steam demand or a falling light-cut purity, both of which indicate that the separation has moved away from the design KA-oil ratio.