Lubinpla

Why polyurethane sealant bubbles in the expansion joint

Summary

Bubbling in polyurethane joint sealant usually comes from substrate moisture, a hot or fast-warming substrate, a punctured backer rod or air trapped during gunning. Gas-pressure bubbles form while the sealant is still skinning over, so the cause is usually set by joint conditions in the first hours after application, not by a defective sealant (Henkel, 2022).

  • Henkel's OSI bulletin reports that bubbles usually form in the skin-over period, typically between five minutes and up to five hours after application, and usually need two or more factors together (Henkel, 2022).

  • In a traditional moisture cure, water reacts with the terminal isocyanate of the polyurethane prepolymer and generates carbon dioxide (Incorez, 2020).

  • One polyurethane data sheet gives a recommended ambient application temperature of 4 °C to 38 °C (40 °F to 100 °F) and warns that moisture vapour moving out of the substrate can cause bubbling (Sika Canada, 2026).

  • A silicone maker's manual states that closed-cell polyethylene backer rod may outgas if punctured and should then be left 20 minutes before sealing (Dow Corning, 2007).

This guide covers bubbles and voids in one-part, moisture-curing polyurethane sealant in building and concrete joints, seen in the first days after application. It does not cover hot-poured pavement sealants, two-part sealants, or blisters that appear years later from coating failures or adhesion loss.

What do bubbles in a polyurethane sealant joint look like?

In a polyurethane sealant joint, bubbles show as raised blisters or domes on the tooled face, small pinholes in the skin, or hollow spherical voids that appear only when the bead is cut open. They form early, in the skin-over period, while a skin is still forming over the uncured sealant. Henkel's OSI bulletin reports that bubbles usually form in that period, typically between five minutes and up to five hours after application (Henkel, 2022).

Where the voids sit in a cut section is the most useful clue. Voids along the bond line with the substrate point toward moisture or gas coming out of the joint wall. Voids on the backer rod point toward the rod itself or toward moisture at the joint base. In practice, irregular, flattened voids and gaps where two passes of sealant overlap point toward air trapped while gunning.

In a Wiss, Janney, Elstner Associates (WJE) failure review of two projects, sealant samples showed both surface bubbling and spherical voids inside the bead, with voids at the interface with the stone and at the backer rod (Construction Specifier, 2019). The samples were silicone, not polyurethane, so treat them as an example of how void position is read.

What are the most likely causes of polyurethane sealant bubbling?

Four causes are the most likely: substrate moisture, a hot or rapidly warming substrate, a punctured closed-cell backer rod, and air trapped during gunning. Henkel's bulletin, written for sealants in general, notes that two or more factors usually act together (Henkel, 2022).

Substrate moisture

Moisture in the substrate converts to vapor, builds pressure in the sealed joint and pushes into the uncured sealant (Henkel, 2022). In a polyurethane, water also reacts with isocyanate and releases carbon dioxide (Incorez, 2020).

Hot or rapidly warming substrate

Sun can raise exterior wall temperature by upwards of 39 °C (70 °F) within an hour, and field observations put bubbling most often on south and west exposures (Henkel, 2022).

Punctured backer rod

Closed-cell polyethylene backer rod may outgas, releasing the gas held in its cells, if punctured during installation (Dow Corning, 2007).

Air trapped during gunning

Air gets in when the nozzle leaves the joint bottom or when passes overlap; one polyurethane data sheet says to avoid overlapping of sealant to eliminate entrapment of air (Sika Canada, 2026).

Does substrate moisture make polyurethane sealant bubble?

Yes. Substrate moisture is the first factor in Henkel's bubbling bulletin and the cause found in the WJE failure review. Moisture held in the joint wall turns to vapor and pushes into the uncured bead (Henkel, 2022). A moisture-curing polyurethane adds a second route: the water that reaches the sealant also reacts with the isocyanate and generates carbon dioxide, and polyurethane sealant makers describe bubbling from this carbon dioxide as a known challenge (Incorez, 2020).

One polyurethane data sheet tells applicators not to apply when a moisture-vapour-transmission condition exists from the substrate, meaning vapour is moving out of it, as this can cause bubbling within the sealant (Sika Canada, 2026). The same data sheet sets waiting times for concrete: green concrete cured at least 24 hours at 24 °C (75 °F), and damp concrete sealed only 60 minutes after all standing water has been displaced (Sika Canada, 2026). The limits are product specific, so check the data sheet of the sealant actually used.

Dew and condensation count as moisture too. In the WJE review, sealant applied early in the morning to avoid heat may have met condensation that formed in the joints before or during installation (Construction Specifier, 2019).

Why does polyurethane sealant bubble on hot, sunny joints?

On hot, sunny joints, heating expands the air and moisture inside the joint and drives them into the sealant while the skin is still forming. Henkel's OSI bulletin reports that sun can raise the temperature of exterior walls by upwards of 39 °C (70 °F) within an hour of direct exposure, and that field observations confirm bubbling most often on southern and westerly exposures (Henkel, 2022).

The worst timing is a joint sealed on a cool, damp morning that then takes full sun. Henkel's advice for that case is to let cladding installed on a cool, wet morning sit a few hours before sealing (Henkel, 2022).

High temperature also takes the sealant outside its stated range. One polyurethane data sheet gives a recommended ambient application temperature of 4 °C to 38 °C (40 °F to 100 °F) (Sika Canada, 2026). A dark substrate in summer sun can sit well above air temperature, so measure the substrate surface, not only the air. The WJE review also tied its bubbling cases to work done near the upper and lower limits of the temperature range allowed by the sealant maker (Construction Specifier, 2019).

Can a punctured backer rod cause bubbles in sealant?

Yes, if the rod is closed-cell. Once its skin is punctured, the gas held in the cells escapes into the uncured bead above it. Dow Corning's technical manual states that closed-cell polyethylene backer rod may outgas if punctured during installation and should then be left for 20 minutes before the sealant is applied (Dow Corning, 2007). The manual is written for silicone sealants, but the gas comes from the rod, so the risk applies under any cold-applied sealant.

Dow Corning recommends sizing backer rod about 25% larger than the joint opening to give back pressure (Dow Corning, 2007), and an oversized closed-cell rod pressed in with a sharp tool can break its skin.

Rod type can rule the cause out. Nomaco describes bi-cellular rod as a mix of open and closed cells that will not outgas when the skin is ruptured, and shows bubbling above a punctured closed-cell rod but none above its bi-cellular rod with the same puncture pattern (Nomaco, accessed 2026). In the WJE review, the use of bi-cellular rod led the investigators to exclude backer-rod outgassing as a cause (Construction Specifier, 2019).

Open-cell polyurethane rod has no gas-holding skin, but it absorbs water (Nomaco, accessed 2026), which brings the moisture cause back in.

How can you tell the bubble causes apart in the joint?

To tell the causes apart, cut out a short section of the joint and read where the voids sit, then check the joint records for moisture, temperature, sun exposure and backer rod type. Each cause predicts a different void position and a different record, and each has a condition that excludes it. Cut one section where bubbling is worst and one where the bead looks clean, and photograph both before the sample dries.

The table maps each check to the cause it points to.

Check

What you see

Points to

Void position in a cut section

Voids along the substrate bond line, wall damp or dark

Substrate moisture

Void position and rod type

Voids sitting on a punctured closed-cell rod; excluded if the rod is bi-cellular (Nomaco, accessed 2026)

Backer rod outgassing (Dow Corning, 2007)

Face orientation and sun

South or west face sealed on a cool morning, then sunlit

Rapid temperature rise (Henkel, 2022)

Temperature at application

Air temperature above the data sheet range, such as 38 °C (100 °F) for one product (Sika Canada, 2026), or a sunlit substrate well above air temperature

Heat

Substrate moisture record

Rain, dew, green concrete under 24 hours, damp concrete under 60 minutes (Sika Canada, 2026)

Substrate moisture

Bead fill

Flattened voids, gaps at overlaps or along the nozzle path

Air trapped during gunning (Sika Canada, 2026)

Void position is a field reading, not a rule: in the WJE review, moisture produced voids both at the stone interface and at the backer rod (Construction Specifier, 2019).

The row that matters most is void position against rod type: voids on the rod with a bi-cellular rod in place point back to moisture, not outgassing. Before you cut the joint, write down what each cause predicts, then mark each one confirmed, excluded or open; root cause analysis for a chemical quality complaint sets out that method.

Do bubbles in cured polyurethane sealant need to be cut out?

Cut the bubbles out when they break the skin, thin the bead or sit in a joint that must stay watertight. Henkel's preferred repair is to remove the whole bubbled bead and reapply (Henkel, 2022).

Henkel and Incorez weigh bubbles differently. Henkel describes bubbles as normally not a performance concern but a cosmetic imperfection (Henkel, 2022), while Incorez links carbon dioxide bubbling in polyurethane sealants to poor durability and delamination from concrete (Incorez, 2020). A few surface domes on a sound bead differ from voids that run to the bond line.

Henkel's repair steps for sealant bubbles, in order:

  1. Fully remove the bubbled bead and reapply, after fixing the cause of the bubbling (Henkel, 2022).

  2. If full removal is not possible, cut the surface off the individual bubbles (Henkel, 2022).

  3. Fill the cut-open bubbles with sealant in direct and continuous contact with the substrate, and do not apply in very thin applications (Henkel, 2022).

Cured polyurethane can only be removed manually or mechanically (Sika Canada, 2026), and any solvent used on uncured material is handled under the solvent maker's warnings and the product SDS (Sika Canada, 2026).

How do you stop polyurethane sealant bubbling in the next joint?

To stop bubbling in the next joint, remove the conditions found in the failed one: dry the substrate, avoid sealing walls that will take full sun within hours, protect the backer rod and fill the joint in one continuous pass. ASTM C1193-25, the guide for use of joint sealants, covers substrate preparation, sealant backing installation and the environmental conditions known to harm a sealant joint (ASTM C1193-25, 2025).

A short pre-sealing check covers each cause:

  1. Moisture: no rain or dew on the joint, concrete past the data sheet waiting time, and standing water displaced (Sika Canada, 2026).

  2. Heat: air temperature inside the data sheet application range (Sika Canada, 2026), substrate surface measured as well, and a cool, wet wall allowed a few hours to dry and warm (Henkel, 2022).

  3. Backer rod: rod inserted without puncturing the skin; a punctured closed-cell rod replaced or left 20 minutes before sealing, per Dow Corning's guidance for its sealants (Dow Corning, 2007).

  4. Air: nozzle kept at the bottom of the joint with sealant flowing ahead of it, no overlapping passes, then tooled to full contact with the joint walls (Sika Canada, 2026).

When should you call a sealant specialist about bubbling joints?

Call a sealant specialist when voids reach the bond line, when bubbling returns after a careful reapplication, or when the cause cannot be narrowed to one of the four groups.

A dispute between applicator and supplier also needs an outside view, because the WJE review shows that the same symptom can come from site conditions rather than the sealant material (Construction Specifier, 2019). So does a substrate with unknown moisture behavior, such as new precast or porous stone.

Bring the evidence: photographs of the bead face and cut sections, the backer rod type and a sample, the product data sheet and batch, its storage conditions and whether the unit was opened the day before, and the weather and substrate temperature at application.

Lubinpla, an applied AI company for specialty chemicals technical reasoning, works such a case from your own records: it fixes the cause hypotheses and their predictions first, then checks them against the site records to give each cause a status and a confidence. That record lets the specialist start from the causes still open.

Frequently asked questions

Does a faster-curing sealant bubble less?

Henkel's OSI bulletin compares sealant technologies, not cure speeds within one: slower curing formulas such as latex and solvent-based sealants have a higher propensity to bubble, while faster curing hybrid and silicone sealants show far fewer occurrences (Henkel, 2022). It does not rank polyurethane.

Can you apply a second thin bead over bubbled sealant?

A second thin bead over bubbled sealant is not the repair Henkel describes. Its bulletin says to remove the bubbled bead or cut open the individual bubbles, fill them in continuous contact with the substrate, and not to apply sealant in very thin applications (Henkel, 2022).

Do porous substrates make bubbling more likely?

Porous substrates make bubbling more likely. Henkel's OSI bulletin names wood, cellular PVC and other porous materials as able to trap moisture, and recommends priming and sealing cut ends of cladding to prevent moisture absorption (Henkel, 2022).

Related posts