Hydrogen peroxide in alkaline high-yield pulp systems undergoes a pH-dependent dissociation to the perhydroxyl anion HOO⁻, with reported pK_a values near 11.6 at 25 °C. In a pulp suspension at pH 10.8–11.2, the equilibrium produces sufficient active perhydroxyl anion for chromophore-selective oxidation but also increases the sensitivity of hydrogen peroxide to transition-metal-catalysed decomposition. Mechanical, thermomechanical and chemithermomechanical pulps retain the greater part of wood lignin, so oxidative bleaching must proceed through chromophore modification rather than delignification. This constraint makes peroxide decomposition losses particularly severe because hydroxyl radicals and molecular oxygen produced by side reactions consume alkali and oxidant without contributing to brightness gain. Manganese originating from wood fibre and process water is commonly present in acid-soluble form at loadings between 30 mg/kg and 300 mg/kg in spruce and pine high-yield systems, while iron and copper are typically present at 5–30 mg/kg and 0.5–5 mg/kg respectively. The dominant decomposition route involves alternating oxidation states of manganese and iron in Fenton-like cycles through which superoxide anion and hydroxyl radical intermediates evolve molecular oxygen. Uncontrolled decomposition of a 4.0% hydrogen peroxide charge can consume more than 40% of the applied oxidant before measurable chromophore oxidation occurs, and the resulting oxygen gas accumulates in fibre flocs to create gas hold-up, channeling and brightness variation in high-consistency retention towers. Stabilisation in industrial practice combines sodium silicate at 2.0–4.0% on oven-dry pulp, chelation of manganese and iron before alkaline peroxide addition, and magnesium sulphate at 0.05–0.20% on oven-dry pulp. Silicate functions as a buffering and metal-passivating colloid, DTPA preferentially complexes iron and manganese under acidic conditions, and magnesium ions modify radical-chain propagation and precipitate metal hydroxides. Brightness measurements are reported according to ISO 2470-1:2016; residual peroxide is determined by iodometric titration according to TAPPI T 556 om-15; fibre viscosity is monitored by ISO 5351:2010. The operational boundary for high-yield peroxide bleaching is narrow: pH below 9.9 suppresses perhydroxyl anion generation, while pH above 11.5 accelerates alkali darkening and peroxide decomposition. Temperature is similarly constrained because thermal decomposition increases above 85 °C and oxygen gas formation becomes severe above 90 °C.

Kinetic data from alkaline peroxide liquor in the absence of stabilisers show an apparent activation energy near 50 kJ/mol, while manganese-catalysed decomposition lowers the apparent activation energy to 30–40 kJ/mol and increases the rate constant by one to three orders of magnitude at pH 11.0. This sensitivity explains why mills that operate peroxide stages with dissolved manganese above 1.5 mg/L in filtrate observe residual peroxide values below 20% of initial charge after 90 minutes at 85 °C. In continuous high-yield bleach plants, the stabiliser package is typically split between the chelation stage, which removes metals before alkaline conditions are established, and the peroxide stage, where silicate and magnesium salts maintain residual stability. A twin-screw press after the chelation tower raises pulp consistency to 30–35% before high-consistency bleaching, and a medium-consistency pump transfers the fibre at 10–12% consistency to the bleaching tower. The peroxide liquor is added in a steam-heated mixing screw where temperature is maintained at 75–85 °C. In this configuration, peroxide charge efficiency is strongly influenced by the transition-metal concentration remaining in the fibre; as a result, acid-soluble manganese is frequently used as the controlling parameter rather than total manganese. Published data for this specific configuration is limited to mill-specific correlations, but the general trend is documented in technical literature.

Silicate Stabilisation Thresholds in CTMP Bleaching

Commercial sodium silicate solutions with a SiO₂:Na₂O weight ratio of 3.22 and density of 1.38–1.42 kg/dm³ are added directly to alkaline peroxide liquor. In CTMP bleaching of spruce and pine, sodium silicate charge below 2.0% on oven-dry pulp leads to rapid peroxide depletion, while charge above 4.5% increases ash content and deposit formation without a proportional brightness gain. The table below presents a representative laboratory evaluation of sodium silicate charge in a CTMP peroxide stage at pH 10.9, 4.0% H₂O₂, 85 °C, 120 minutes retention and 25% consistency. At a silicate charge of 0%, the final residual peroxide is typically below 15% of initial charge, while brightness gain is limited to 2.1 ISO points. At 2.5% silicate charge, residual peroxide exceeds 50% and brightness gain increases to 6.3 ISO points. Above 4.5%, brightness gain approaches a plateau while ash accumulation and deposition rates rise sharply. The data indicate a practical silicate threshold between 2.5% and 3.5% for CTMP lines that must balance peroxide stability against downstream ash and scale constraints.

Sodium silicate stabilises residual peroxide through at least three simultaneous mechanisms: it buffers the alkali charge and limits pH excursions; it forms colloidal silicic acid species that adsorb onto catalytically active hydrous oxide surfaces; and it precipitates or co-precipitates transition metals as silicates and hydroxides. The buffering effect is particularly important in high-yield pulp because the fibre itself releases organic acids during peroxide treatment, which can lower local pH at the fibre surface and suppress perhydroxyl anion formation. Silicate also increases the ionic strength of the bleach liquor and alters fibre surface charge, which changes the distribution of dissolved manganese between fibre and aqueous phase. At pH above 11.5, however, silicate can contribute to alkali darkening and can accelerate yellowing of peroxide-bleached mechanical pulp if residual silicate remains in the sheet. The ash increase measured by ISO 1762:2019 is therefore a direct quality parameter for silicate-stabilised grades. Deposition becomes severe when silicate polymerises in low-pH zones or when the solution is heated above 90 °C in plate heat exchangers. Cleaning of silicate scale requires acidic washing with 5% sulfuric acid at 45 °C or inhibited phosphoric acid blends, and clean-in-place cycles are commonly scheduled every 14–21 days on bleach plant heat exchangers and filtrate lines.

In thermomechanical pulp mill processing Norway spruce (Picea abies), a DTPA chelation stage operated at pH 5.8, 65 °C and 20 minutes reduces acid-soluble manganese from 180 mg/kg to 12–18 mg/kg before the alkaline peroxide stage. The chelation stage commonly uses 0.4% DTPA on oven-dry pulp and is followed by pressing to 30–35% consistency. DTPA forms stable complexes with iron(III) and manganese(II); however, its selectivity changes with pH, and at pH above 7.0 manganese removal falls because hydrolysis products compete with chelation. The optimum pH for DTPA chelation is 5.0–6.0 for manganese and 4.0–6.0 for iron. Below pH 4.0, acid hydrolysis of polysaccharides becomes measurable under hot conditions; above pH 7.0, transition metals remain as colloids that are not removed by pressing. The chelation stage is therefore a narrow-window process. In mills with calcium hardness above 180 mg/L as CaCO₃, a chelating agent with lower calcium affinity such as DTPA is preferred over EDTA because EDTA consumption by calcium increases chemical demand and impairs manganese removal. The chelation filtrate is sent to effluent treatment; residual DTPA in pulp entering the peroxide stage must be below 0.05% on pulp because uncomplexed chelant can reduce peroxide efficiency by forming mixed-ligand complexes with iron that retain catalytic activity.

Process control of the chelation stage uses ICP-OES analysis of filtrate according to ISO 11885:2007; total manganese, iron and copper are monitored, and the target for acid-soluble manganese before the peroxide stage is below 20 mg/kg pulp. When total iron exceeds 15 mg/kg, DTPA alone may not be sufficient. A small addition of sodium bisulfite at 0.2% on pulp during chelation reduces Fe(III) to Fe(II), which is more mobile under acid conditions. This reduction step is incompatible with alkaline peroxide addition if not washed effectively because residual sulfite consumes peroxide. In a TMP mill with a twin-wire press achieving 30% consistency after chelation, carryover of dissolved metals in the pressate is lower than that from a screw press achieving 35% consistency, but the twin-wire press requires higher wash-water flow. Published exact mill data for this configuration is limited, but the general operating boundaries are well documented: chelation temperature between 60 °C and 70 °C, residence time between 15 minutes and 30 minutes, and pH between 5.0 and 6.5. Operation outside this window leads to either insufficient metal removal or excessive acid hydrolysis, both of which reduce final brightness and increase peroxide consumption.

What Limits Peroxide Charge Efficiency When Dissolved Manganese Exceeds 25 mg/kg Pulp?

Dissolved manganese concentrations above 1.5 mg/L in filtrate or acid-soluble manganese above 25 mg/kg pulp in the feed to the peroxide stage limit peroxide charge efficiency. The mechanism involves Mn(II)/Mn(III)/Mn(IV) cycling that catalyses hydrogen peroxide decomposition to oxygen and water. Under typical alkaline peroxide conditions at pH 10.8 and 85 °C, manganese loadings from 5 mg/kg to 80 mg/kg produce a nonlinear decline in residual peroxide after 90 minutes. The threshold for effective peroxide stabilisation is 20–25 mg/kg acid-soluble manganese. Above this level, residual peroxide falls below 40% of initial charge and brightness gain drops below 6 ISO points. To control manganese, acid washing at pH 4.5–5.5 with sulfuric acid or sodium bisulfite followed by DTPA at 0.4% is used. Acid-only washing removes a fraction of acid-soluble metals but can dissolve fibre-bound calcium and increase hardness in downstream filtrate. DTPA addition at pH 5.5 is then required to complex residual ions. A two-stage sequence consisting of acid washing at 70 °C for 30 minutes followed by chelation at 65 °C for 20 minutes is adequate for most spruce and pine high-yield pulps.

The table demonstrates that manganese loadings above 25 mg/kg produce a measurable loss in intrinsic viscosity, indicating that radical-driven carbohydrate depolymerisation accompanies poor peroxide stabilisation. Intrinsic viscosity is determined according to ISO 5351:2010 in cupriethylenediamine solution. The viscosity decline from 860 dm³/kg at 5 mg/kg manganese to 730 dm³/kg at 80 mg/kg manganese is significant for high-yield pulps, which depend on fibre length and wall integrity for strength development. The process bottleneck is often pressing: if press consistency falls below 28%, dissolved metals in the carryover pressate increase sharply and the peroxide stage must be run with reduced alkali and temperature to avoid gas formation. When acid-soluble manganese cannot be reduced below 20 mg/kg, the preferred response is to increase sodium silicate charge from 2.5% to 3.5% and to add magnesium sulphate at 0.2% on oven-dry pulp. This combination does not remove manganese but reduces its catalytic activity sufficiently for short-retention bleaching at 80 °C. The operational temperature must be limited to 80–85 °C because manganese-catalysed decomposition accelerates rapidly above 85 °C.

When Magnesium Sulphate Replaces Sodium Silicate in BCTMP Brightness Ceiling Operations

Where mill water hardness exceeds 180 mg/L as CaCO₃, magnesium sulphate replacement of sodium silicate must be evaluated against deposit formation. In BCTMP lines aiming for brightness levels above 78 ISO points, silicate-free peroxide stages with MgSO₄ at 0.15% on pulp and DTPA chelation produce acceptable peroxide stability only if acid-soluble manganese is below 15 mg/kg. The magnesium ion stabilises residual peroxide by precipitating as Mg(OH)₂ and by blocking catalytically active hydrous oxide surfaces; however, MgSO₄ alone does not provide the silicate buffering. In a silicate-free system, pH must be controlled within 10.4–10.8, and alkali charge is reduced by 15–20% relative to silicate-stabilised runs to avoid alkali darkening. Brightness ceiling is lower by 1.0–1.5 ISO points when silicate is eliminated. Ash content measured by ISO 1762:2019 decreases from 0.65% to 0.25%, and paper machine white-water fines retention improves because colloidal silicate is absent. This is a primary motive for silicate-free operation in BCTMP grades intended for absorbent hygiene products, where ash content and surface absorption capacity are critical quality parameters.

The operational boundary is pH: below 10.3, brightness gain falls; above 11.0, Mg(OH)₂ precipitation becomes excessive and oxygen gas evolution increases. If the water contains bicarbonate above 120 mg/L as CaCO₃, MgCO₃ or Mg(OH)₂ deposits on heat exchanger surfaces and tower walls, so a softener is required. Silicate-free systems are unsuitable for hardwood BCTMP when iron exceeds 10 mg/kg, because magnesium is less effective at iron passivation than silicate and iron-catalysed decomposition persists. Magnesium sulphate should be added to the bleach liquor after steam heating and before peroxide injection to avoid localised Mg(OH)₂ formation in hot pipelines. Avoid combination with phosphate-based scale inhibitors in the same dosing point because magnesium ammonium phosphate or struvite scaling can occur in downstream filtrate lines. Published data for this specific configuration is limited; mill-specific trials are required to define the brightness ceiling and the maximum acceptable iron loading. Where BCTMP mills operate with magnesium sulphate and DTPA only, the peroxide charge must be maintained at 4.0–4.5% to compensate for lower stabilisation efficiency, and retention time must not exceed 100 minutes.

At pulp consistencies above 28% by mass, gaseous oxygen released from uncontrolled peroxide decomposition occupies the void volume between fibre flocs and reduces bulk density. In a high-consistency tower with diameter 3.8 m and side height 14 m, gas hold-up above 5% by volume changes residence-time distribution from plug flow toward a dispersed regime and increases brightness variability by 2–3 ISO points over a production day. The tower discharge cone must be equipped with a deaeration cyclone and a gas vent line because accumulated oxygen causes bridging of fibre at the dilution screw and intermittent discharge. The oxygen release rate from a 4.0% H₂O₂ charge is low under stabilised conditions, typically below 0.2 L O₂/kg o.d. pulp/h, but can exceed 0.8 L O₂/kg o.d. pulp/h when manganese loadings are above 25 mg/kg or tower temperature exceeds 90 °C. Stabilisation therefore has direct process-control implications: it reduces gas hold-up, improves tower plug flow and narrows brightness distribution. In tracer studies on a pilot high-consistency tower, stabilised peroxide runs gave a dispersion number of 0.10 and a mean residence time of 112 minutes, while manganese-spiked control runs gave a dispersion number of 0.35 and a mean residence time of 97 minutes. This difference was attributed to oxygen-gas channelling and localised fibre compaction. The results are specific to the tower geometry and are not a universal predictive model.

Gas channelling in high-consistency towers is controlled primarily by minimising oxygen formation, not by increasing tower pressure. Vented towers at atmospheric pressure require careful control of oxygen release, because high gas hold-up can produce upward gas migration that disrupts fibre plug flow. Fibre flocs act as nucleation points for oxygen bubbles, and hydrophobic extractives on mechanical pulp surfaces increase bubble adhesion. At oxygen release rates above 0.5 L O₂/kg o.d. pulp/h, the resulting two-phase flow reduces effective tower capacity and may force a production rate cut of 10–15%. The peroxide stage should be operated with a final residual peroxide of at least 20% of initial charge to ensure that oxidation continues throughout the retention tower but not so high that excessive oxygen forms during storage. Excess residual peroxide entering paper machine white water also causes oxidative degradation of wet-end polymers and increases chemical oxygen demand. The closed water loop must therefore be monitored for dissolved oxygen and residual peroxide using on-line sensors calibrated against TAPPI T 556 om-15.

Elevated Post-Refining Temperatures Reduce Peroxide Half-Life in High-Consistency Retention

A temperature rise from 80 °C to 90 °C in high-consistency retention reduces hydrogen peroxide half-life by approximately 2.0–2.5 times when all other variables are held constant. This follows the general kinetic rule that reaction rate doubles per 10 °C increase in temperature; published data for this specific configuration is limited. In a TMP line with post-refining consistency of 30%, steam heating to 90 °C before peroxide injection causes rapid initial decomposition and a drop in residual peroxide from 55% to 31% of initial charge at 90 minutes. The operating window for high-consistency peroxide bleaching is 70–85 °C; below 65 °C chromophore oxidation is too slow, and above 85 °C oxygen gas formation becomes severe. Heat exchangers must be designed for silicate-stabilised liquor with a viscosity of 1.5–3.5 mPa·s at 80 °C; plate-and-frame exchangers require clean-in-place cycles with 5% sulfuric acid at 45 °C every 14–21 days to remove silicate scale. Temperature after the refiner should be reduced to below 85 °C by dilution-cooling or heat recovery before peroxide addition. Steam addition should be controlled by mass-flow meters and a final temperature probe loop; local hot spots above 92 °C cause fibre yellowing and local peroxide depletion.

Thermal stabilisation also relates to the addition sequence. If steam is injected after peroxide has been mixed with alkali, local high temperatures create zones where the perhydroxyl anion decomposes before it can diffuse into the fibre wall. The preferred sequence in high-consistency bleaching is dilution and temperature adjustment before peroxide injection, followed by alkali addition in a high-shear mixer. The high-shear mixer should operate at a tip speed of 25–35 m/s to distribute bleach liquor uniformly over fibre surfaces. At consistencies above 30%, the mixing energy requirement rises sharply, and insufficient dispersion produces localised alkali and peroxide concentrations that increase yellowing and oxygen formation. In medium-consistency systems at 10–12% consistency, the same temperature limits apply, but gas removal is easier because the suspension can be passed through a medium-consistency gas separator. The thermal half-life data therefore support a maximum bleach tower inlet temperature of 85 °C and a maximum pulp temperature during post-refining storage of 70 °C.

In bleach plant filtrate recovery, silicate polymerisation becomes irreversible when the pH drops below 10.0 or when the SiO₂ concentration in the aqueous phase exceeds the solubility limit at the filtration temperature. Filtrate from a silicate-stabilised peroxide stage contains 2.5–3.5 g/L SiO₂ and has pH 10.5–10.8. If this filtrate is mixed with acidic filtrate from the chelation stage without pH control, the pH drops to 6.5–7.5 and polymerised silica precipitates as a gel that blocks fibre fines and reduces paper machine dewatering. To avoid this, bleach plant filtrate must be maintained above pH 10.2 or diluted below 1.0 g/L SiO₂. Silica scale deposition in filtrate tanks increases cleaning frequency and can raise biocide consumption because biofilms adhere to rough silica surfaces. A laminar-flow shell-and-tube heat exchanger operating at a surface temperature above 60 °C accelerates silicate scale formation on the hot side; the exchanger should be designed with a fouling factor of 0.00035 m²·K/W and operated with a wall shear stress above 10 Pa to limit deposition. In mills that need a silicate-free system, this filtrate silica problem is eliminated, but the brightening ceiling and paper machine deposit profiles change. The choice between silicate and silicate-free operation is therefore a trade-off between peroxide stability in the bleach tower and downstream silica management.

In activated sludge treatment of high-yield bleach plant effluent, residual DTPA at concentrations above 5 mg/L can reduce metal bioavailability and complicate compliance with discharge limits for heavy metals. Chemical oxygen demand measured by ISO 6060:1989 is also elevated by dissolved organic fragments produced during peroxide decomposition. Silicate-stabilised effluents contain high suspended solids from polymerised silica, which increases the load on primary clarification and may require flocculant doses of 0.5–1.5 mg/L of cationic polyacrylamide. Magnesium sulphate raises conductivity, and at concentrations above 300 μS/cm in final effluent the discharge may exceed receiving-water limits under certain permits. DTPA is poorly biodegradable in activated sludge and persists in secondary effluent; EDTA is even more persistent. Where ecotoxicity tests are required, the whole-effluent assessment may use ISO 6341:2012 for Daphnia magna immobilisation. Process design must therefore integrate the stabiliser choice with effluent treatment capacity and permit limits, not only with bleach tower brightness and peroxide residual.