The sofa arrives looking like silk. Eighteen months on, somebody tips elderflower cordial down the outside arm, a cleaner turns up with a hot water extraction wand, and by the following evening the whole panel has gone the colour of weak tea. Nobody spilled tea. That brown came out of the fabric itself, dragged to the surface by water which had no business going near it. I take these calls across Greater London roughly twice a month, almost always after another firm has had a go and then stopped answering the phone. The fabric is viscose. The showroom ticket said faux silk. Those two facts explain each other.
What sits inside a “faux silk” sofa fabric?
Viscose never grew on anything. It starts as wood pulp – beech, eucalyptus or bamboo – dissolved in caustic soda, aged, treated with carbon disulphide until it becomes a honey-coloured syrup, then forced through a spinneret into an acid bath where it hardens into filament. Regenerated cellulose, on the technical data sheet. Cotton’s chemistry without cotton’s structure.
What survives that process matters more than what goes into it. Short-chain hemicelluloses. Fragments of lignin the pulping stage failed to strip out. Residual sugars. Mill sizing and optical brighteners. Softeners sprayed on at the finishing stage to give the fabric that liquid hand people pay for. All of it sits in and around the fibre. All of it dissolves in water. Most of it is yellow already or turns yellow the moment it oxidises at an exposed surface.
Then there is the strength problem. A viscose filament loses somewhere between half and two thirds of its tensile strength once it is saturated, because water penetrates the large amorphous regions left over from spinning and pushes the cellulose chains apart. Silk, which is protein, holds up far better under the same conditions. A wet viscose pile that gets brushed or leaned on will crush and splay, and the crush is permanent.
The label that costs people money
Art silk. Banana silk. Bamboo silk. Sabra silk. Vegan silk. Viscose rayon, modal, cupro, and the far better behaved lyocell. I have seen every one of those on a swing tag in a Tottenham Court Road showroom, and I have seen exactly one salesperson volunteer that the material behaves nothing like silk in the presence of water. The buyer goes home believing they own something delicate and precious. They own something delicate and cheap, and the care advice for the two materials points in opposite directions.
Why does the brown only appear once the sofa has dried?
The damage happens during drying. That is the bit almost every general cleaner gets wrong.
Water goes into the cushion or the arm panel. It dissolves everything soluble in the face fabric and in whatever sits below – the jute-based hessian webbing across the frame, the deck fabric stapled under the seat, hot-melt adhesive on a foam wrap, the tannin-heavy dye in a contrasting piping cord. Then evaporation begins, and evaporation only occurs at a surface. As the water leaves, capillary action drags the remaining moisture upward and outward towards the drying front, and every dissolved sugar and degraded cellulose fragment travels with it. The water goes. The brown stays, concentrated in a hard ring at the exact edge of wherever the wet patch ended.
Cleaners call this wicking. Clients call it a stain that appeared out of nowhere.
The rate matters as much as the volume. A panel that dries in ninety minutes moves very little dissolved material to the surface, because there is not enough time for capillary transport to do its work before the fibre stiffens up. The same panel left to dry over two damp days in an unheated Victorian front room will deliver every last sugar to the face of the fabric and set it there.
There are two separate mechanisms and they get muddled constantly. Cellulosic browning comes from the viscose itself, from sugars and oxidised cellulose in the face fibre. Jute browning comes from the backing, the webbing or the deck, and it is the same lignin-rich material that ruins jute-backed rugs on a wet job. On a badly built sofa you can get both at once, and the second one is worse because it keeps feeding the surface for as long as the frame stays damp.
Alkalinity accelerates all of it. Most upholstery prespray sits somewhere between pH 9 and 10.5, because that is the range which shifts body oil off a headrest. Cellulose hydrolyses in alkali. Chains snap, reducing ends form, and those reducing ends go on to develop the chromophores which show up as tan and gold. Heat compounds it – a 55°C solution in a machine tank is doing chemistry to a wood-pulp fibre, not rinsing it.
Something I saw in a first-floor flat off Fortis Green Road in Muswell Hill still stands as the best illustration I have. Three-seater, taupe viscose velvet, spot-cleaned by the tenant with a supermarket foam and a damp cloth. Six overlapping brown haloes across the seat deck, each one tracing the outline of a cloth. She had cleaned it perfectly. The cloth boundary is what browned.
The gap between the van leaving and the phone ringing
A job like that looks flawless when the technician packs up. The fabric is dark with moisture and the colour looks even. Browning shows at somewhere between eight and thirty-six hours, which is precisely long enough for the cleaner to have invoiced and moved on to the next property. Ask me how I know.
Can browning on viscose upholstery be reversed?
Sometimes. Not always, and the honest split is roughly two in three.
Chemical browning responds to acid. An anti-browning treatment based on citric or acetic acid, applied at low moisture and dried fast, will reduce the chromophores back to a colourless state and pull the fabric pH down to 4.5 or thereabouts, which is where viscose is happiest anyway. Stubborn cases need a reducing agent – sodium bisulphite or metabisulphite, cautiously, on a fabric already weakened. Every one of these carries dye-bleed risk, and a viscose dyed in a deep teal or a plum will surrender colour to an acid rinse faster than you can blot it off the towel. Ask me how I know. Test in a hidden spot, under the skirt or behind the outside back panel, and give it a full dry before judging the result.
Hydrogen peroxide is where I part company with a lot of the trade. It is an oxidiser applied to a fibre whose whole problem is oxidation. It brightens the panel today by attacking the chromophore and it leaves you with oxycellulose and a weaker, chalkier fabric which browns again more readily next time. I will not put it on viscose upholstery and I think the people who do are borrowing against the client’s sofa.
Physical damage has no chemical answer at all. Once a wet viscose pile has been crushed by a wand head or a technician’s forearm, that area reflects light differently for good. Fibrillation – the splitting of the filament into fine hairs – reads as a permanent white bloom or a dull patch, and no amount of grooming brings it back.
When the answer is that I should not touch it
I turned down a pair of viscose velvet armchairs on Lansdowne Road in Holland Park last spring. Water damage from a radiator valve, both seat platforms dried hard and stiff, browning right through the pile into the foam. The client wanted them saved. The realistic outcome was a partial improvement on chairs that would still look wrong, at a cost around forty per cent of what a re-cover through a Bermondsey trimmer would run. I said so and lost the job. That is the right call more often than the industry admits.
What do specialists do instead of wet extraction?
Identify the fibre before anything else touches it. Burn test on a thread pulled from a seam allowance underneath, or a scope if the sofa is worth the extra five minutes. Check the manufacturer’s cleaning code where one exists – S for solvent, W for water, SW for either, X for vacuum only – and treat an absent label as a reason for caution rather than permission. Plenty of imported frames sold through the big Croydon retail parks carry no code at all, or carry an S code stitched into a fabric the factory swapped out mid-production run.
After that the method choices narrow down fast.
Dry solvent is the default for anything with a serious viscose content. A volatile hydrocarbon or a low-residue dry cleaning fluid applied to a white towel and worked into the fabric, never poured in, never flushed. Solvent carries oily soil, evaporates without swelling the cellulose, and leaves nothing behind to wick. Ventilate properly and mind the foam beneath.
Where water cannot be avoided – protein spills, some food spotting – the approach is low-moisture encapsulation at acid pH, worked with a soft horsehair brush, then immediate drying. Air movers on the panel within minutes. A dehumidifier if the room is a basement conversion off Bellevue Road in Wandsworth in February. Under two hours to dry, or expect a ring. Ask me how I know.
Work seam to seam, always. A viscose panel cleaned in the middle will show its own boundary as clearly as any spill did, so the seams become the edge of the wet area and the drying front has somewhere sensible to stop. Groom the pile in one direction while it is still damp, then leave it alone.
The five minutes that saves a £2,400 sofa
Pull a thread from underneath and burn it. Viscose flares like paper, keeps an afterglow, leaves a soft grey ash and smells of a garden bonfire. Silk shrinks away from the flame, self-extinguishes, smells of burnt hair and leaves a black bead that crushes to powder between your fingers. Polyester melts into a hard black lump and smells of nothing you would want in your lungs. The difference takes about eight seconds to establish and decides everything that follows.
Most viscose sofas sold in this city should never meet a hot water extraction wand, and a good proportion of the ones I inspect in Bethnal Green warehouse conversions and Sydenham terraces have already met one. The brown ring on the outside arm is usually where the previous cleaner’s towel stopped.