PLA filler interfacial adhesion without a compatibilizer

Poly(lactic acid) composites filled with wood flour and a thermoplastic starch were made with no compatibilizer at all, and the interfacial evidence was still there: the PLA carbonyl peak shifted from 1746 to about 1735 reciprocal centimeters, a 5 to 10 reciprocal centimeter move consistent with hydrogen bonding to filler hydroxyl groups. Adding wood flour raised apparent crystallinity from 49.50% to as much as 87.1%. For a formulator, the surprise is that the filler acted as a nucleating agent, not just a filler.

Why the standard answer is to add a compatibilizer

Polylactic acid is a compostable thermoplastic with real advantages over petroleum-based plastics, including lower production energy and good formability. It also has well-known weaknesses: brittleness, slow degradation, inherent hydrophobicity, and a lack of reactive functional groups along its backbone. That last one is the problem. PLA has nothing on the chain that will react with a hydroxyl-bearing natural filler, so the interface between the two rests on van der Waals forces and hydrogen bonding rather than a covalent bond.

The standard industry answer is a compatibilizer. Prior work on PLA/wood flour/starch blends typically used maleic anhydride-grafted PLA or silane coupling agents to overcome the interfacial incompatibility. Those work, which is the point, but they mask the question underneath. Once you have added a coupling agent, you can no longer tell whether the interface was already adequate and the additive was solving a different problem, such as dispersion or processability. Measuring pla filler interfacial adhesion first is what makes the distinction possible.

A recent study took the additive out and looked at what the raw system does. The design was deliberate, and that is why the results are worth reading. The stated purpose was twofold: establish the baseline interactions and property balances in a ternary system, and identify whether one filler enhances the function of the other. In practice, a compatibilizer-free baseline tells you what an additive still has to fix.

What the formulations were

The matrix was PLA, a general-purpose grade. The fillers were wood flour below 250 micrometers and a commercial thermoplastic starch compound. That starch compound is worth a second look, because it is not pure starch. Thermogravimetric analysis measured its composition at 63% starch, 31% polypropylene, and 6% plasticizer, and it degrades in two steps, starch at 250 to 350 degrees Celsius with 63% mass loss, then polypropylene at 400 to 470 degrees Celsius with 31%.

That polypropylene fraction matters more than it looks. It is why the starch compound is a processing aid rather than a filler in the ordinary sense, and also why the contact angle results in this system behave counterintuitively.

ComponentLoading rangeRole
PLA matrix30 to 50 wt%Continuous phase
Starch-based compound20 to 60 wt%Bio-based filler with a polypropylene fraction
Wood flour10 to 30 wt%Cellulosic filler under 250 micrometers

Nine formulations were produced by extrusion followed by compression molding, covering three PLA levels against three wood flour levels. Samples for water contact angle were measured at 15 measurements per formulation, from three specimens at five locations each, and reported as mean plus or minus standard deviation.

What pla filler interfacial adhesion looks like without a compatibilizer

The clearest evidence is in the infrared spectrum. Neat PLA has a sharp, intense carbonyl stretching peak at 1746 reciprocal centimeters. In every composite, that peak broadened and shifted down to roughly 1735 to 1740 reciprocal centimeters. In one formulation, 40% PLA with 40% starch compound and 20% wood flour, the band sat centered near 1735.

A shift of about 5 to 10 reciprocal centimeters downward in a carbonyl band is what hydrogen bonding looks like. The carbonyl oxygen of PLA accepts a hydrogen bond from a hydroxyl group on the wood flour or the starch, which pulls electron density out of the C=O bond and lowers its vibrational frequency. The peak also broadens, which is what happens when the same vibrational mode is distributed across a range of slightly different local environments. That is pla filler interfacial adhesion measured directly, without an additive in the loop.

The honest caveat is worth stating, and the study states it. Infrared spectroscopy alone cannot confirm hydrogen bonding without peak deconvolution, temperature-dependent measurements, or X-ray photoelectron spectroscopy. Alternative explanations for a shift this small include changes in the local dielectric environment, dipole-dipole interactions, or physical confinement from filler dispersion.

A second region invites over-reading. All composites showed increased absorption between 1600 and 1650 reciprocal centimeters, and it is tempting to attribute that to the carbonyl. The study pushes back, correctly. That region can hold the O-H bending of absorbed water, which hydrophilic natural fillers carry, aromatic C=C stretching from lignin, and overlapping starch signals. Attributing it to a drastically shifted carbonyl is not supported by how hydrogen-bonded esters behave.

The defensible conclusion is narrow and still useful: physical adhesion via hydrogen bonding does occur at the PLA-filler interface, and the evidence is the consistent small shift of the main carbonyl band itself.

Water contact angle says two things at once

Contact angle rose and fell in a way that only makes sense once you know what is in the starch compound.

SampleContact angle (degrees)Work of adhesion (mN/m)Spreading coefficient (mN/m)
Neat PLA80 +/- 2.185.4-11.9
30 PLA, 60 starch, 10 WF92 +/- 3.470.3-15.3
30 PLA, 50 starch, 20 WF85 +/- 2.977.5-13.4
30 PLA, 40 starch, 30 WF68 +/- 2.8100.1-6.6
40 PLA, 50 starch, 10 WF65 +/- 2.6103.6-5.5
40 PLA, 40 starch, 20 WF62 +/- 2.5107.0-4.6
40 PLA, 30 starch, 30 WF58 +/- 2.4111.4-3.7
50 PLA, 40 starch, 10 WF61 +/- 2.3108.5-4.2
50 PLA, 30 starch, 20 WF58 +/- 2.4111.4-3.2
50 PLA, 20 starch, 30 WF55 +/- 2.2114.6-1.9

Neat PLA measured about 80 degrees, described as weak hydrophilicity. Look at the first data row and the trend inverts: 30% PLA with 60% starch compound and only 10% wood flour gave 92 degrees, the highest angle in the set. More starch compound means more polypropylene, and polypropylene contributes hydrophobicity.

Wood flour pushes the other way, and hard. Every increase in wood flour at constant starch compound dropped the angle: at 30% PLA, moving from 10% to 30% wood flour took it from 92 to 68 degrees. At 50% PLA, the same move took it from 61 to 55 degrees.

Work of adhesion, calculated from the Young-Dupre equation using a water surface tension of 72.8 mN/m, tells the complementary story. Neat PLA was 85.4 mN/m. The starch-rich formulation was lowest at 70.3 mN/m. The wood-flour-rich formulation was highest at 114.6 mN/m.

All spreading coefficients were negative, from -15.3 down to -1.9. A negative spreading coefficient means water does not spontaneously wet the surface, which is the useful number for a composite meant to resist water. The more negative the value, the more reluctant the wetting. The formulation with the most wood flour and least starch compound, at -1.9 mN/m, was the least wettable in the set.

The two effects partly cancel. The polypropylene in the starch compound pulls the angle up and offsets the hydrophilicity the wood flour would otherwise add, which is why the contact angle drop with wood flour is smaller than the wood flour content alone would predict.

The finding nobody expected: the filler crystallized the matrix

This is the part that changes how you think about filler selection. Differential scanning calorimetry showed wood flour working as a heterogeneous nucleating agent, and the effect was large.

SampleMelting onset (degrees Celsius)Melting enthalpy (J/g)Apparent crystallinity (%)
Neat PLA169.946.3349.5
30 PLA, 60 starch, 10 WF152.038.1355.9
30 PLA, 40 starch, 30 WF153.242.7678.0
50 PLA, 40 starch, 10 WF153.451.7270.3
50 PLA, 20 starch, 30 WF149.652.4287.1

Apparent crystallinity rose from 49.50% in neat PLA to 87.1% in the formulation with 50% PLA, 20% starch compound, and 30% wood flour. The melting onset fell at the same time, from 169.9 degrees Celsius to as low as 149.6.

The proposed mechanism is heterogeneous nucleation by cellulose. Cellulose is semi-crystalline, and that structure may offer a favorable template for the epitaxial ordering of PLA chains through better topological matching. The lignin fraction, while more hydrophobic and therefore potentially helpful for interfacial compatibility, has an amorphous bulky aromatic structure less conducive to the long-range order effective nucleation requires. Mechanical processing during extrusion and molding likely exposes cellulose microfibrils at the interface, and the wood flour particle size below 250 micrometers gives enough surface area for these interactions.

There is an important measurement caveat, and it affects how you read the table. The polypropylene phase in the starch compound melts between 160 and 170 degrees Celsius, which overlaps the PLA melting endotherm near 150 to 155 degrees Celsius. Because of that overlap, the reported crystallinity is an effective value for the whole blend, not PLA-specific. The calculation used a reference enthalpy of 66.77 J/g, derived as 0.63 times 4.12 plus 0.31 times 207, from the measured starch compound composition.

So the honest reading is that total apparent crystallinity rose substantially, and the rise is consistent with wood flour driving nucleation of both the PLA phase and potentially the polypropylene phase. Confirming the mechanism would need polarized optical microscopy or X-ray diffraction to distinguish epitaxial growth from transcrystalline layer formation.

What the additive would actually have to fix

The baseline established here is the starting point for compatibilizer selection, and it points at a narrower target than expected. Density measurements showed relative density above 0.95, so the composites compacted well without an additive. Soil burial at 15 cm depth, pH 8 plus or minus 0.3, and 25 to 30 degrees Celsius produced 4 to 7% mass loss over 6 weeks on construction soil with no plants cultivated for at least 10 years.

The remaining weak points are specific. The contact angle data show the surface is still wettable, with a spreading coefficient of -1.9 at best, and the hydrophobicity of the polypropylene fraction in the starch compound works against the wood flour rather than with it. Thermal analysis showed phase-separated behavior, with two distinct derivative peaks confirming a phase-separated blend, and the increased thermal stability of the PLA phase in composites may be partly attributed to heightened crystallinity.

So the additive’s job is narrower than it looks. It is not rescuing a dead interface; pla filler interfacial adhesion is already there through hydrogen bonding. The additive improves the margin, particularly dispersion and the water barrier, without destroying the nucleation benefit that the wood flour provides for free.

Practical guidance for PLA filler and compatibilizer work

  • Measure pla filler interfacial adhesion before you add anything. A 5 to 10 reciprocal centimeter carbonyl shift is measurable evidence that hydrogen bonding already exists.
  • Check whether your filler nucleates. Wood flour moved apparent crystallinity from 49.50% to 87.1% here, and that changes the whole optimization.
  • Know what is in a commercial starch compound. 63% starch, 31% polypropylene, and 6% plasticizer is a very different material from pure starch, and the polypropylene fraction drives the contact angle up.
  • Do not over-read a broad spectral region. The 1600 to 1650 reciprocal centimeter band carries water and lignin signals, so the carbonyl shift is the only defensible evidence pla filler interfacial adhesion offers.
  • Use work of adhesion and the spreading coefficient together. The angle alone will mislead you in a system with a hydrophobic polymer fraction.
  • Confirm nucleation mechanism before designing around it. Polarized light or X-ray diffraction is the next step, not DSC.
  • Remember that more crystallinity changes more than strength. It raises glass transition temperature, improves dimensional stability, and reduces degradation rate.
  • Treat the compatibilizer-free case as a baseline worth publishing. It tells you which problems are real and which the additive was solving by accident.

Limitations

The crystallinity values are the weakest part of the dataset. Overlapping polypropylene and PLA melting endotherms mean these are blend-level effective values, not PLA-phase crystallinity, so the 87.1% figure should not be quoted as the crystallinity of the PLA phase. The nucleation mechanism is proposed rather than demonstrated, and the cellulose-dominant hypothesis is stated as a hypothesis.

The mechanical data are thinner than the thermal and spectroscopic data. Density, contact angle, spectroscopy, thermal analysis, and burial mass loss are reported across nine formulations, but no tensile, flexural, or impact strength results are given, so the mechanical consequence of the crystallinity jump cannot be quantified from this work. The reported degradation is soil-burial mass loss over 6 weeks, measured on disc specimens 2.5 mm thick, which is a mass-loss endpoint rather than a full degradation kinetics analysis.

The formulation space is also narrow. Nine formulations at three PLA and three wood flour levels, with the starch compound filling the remainder, do not explore the full ternary space, and no compatibilized control was run, so the incremental benefit of adding one cannot be quantified against this baseline. The burial test used a single soil type, depth, pH band, and temperature range, so the 4 to 7% figure is specific to that condition.

Finally, the contact angle values come from a water droplet on a molded surface, so they characterize the surface as molded, not the bulk interface. The good agreement between 3 specimens at 5 locations helps, but a contact angle is still a surface measurement and should not be used alone to conclude anything about internal adhesion.

Bottom line

Two findings here are worth carrying into formulation work. First, a polar interface formed without any compatibilizer, evidenced by a consistent 5 to 10 reciprocal centimeter carbonyl shift, so pla filler interfacial adhesion is something the additive reinforces rather than creates. Second, the filler was doing a job nobody assigned to it: wood flour acted as a nucleating agent and raised apparent crystallinity from 49.50% to 87.1%. Measure the interface first, check for nucleation, and then specify the compatibilizer for the margin it actually needs to fix.

FAQ

Can PLA form a strong interface with wood flour without a compatibilizer?

Yes, to a meaningful degree. The PLA carbonyl peak shifted from 1746 to roughly 1735 reciprocal centimeters in the composites, a 5 to 10 reciprocal centimeter move consistent with hydrogen bonding between the carbonyl oxygen and the filler hydroxyl groups. The authors describe this as indicative rather than conclusive, since infrared spectroscopy alone cannot rule out dielectric or confinement effects.

What does a 5 to 10 reciprocal centimeter carbonyl shift indicate?

It indicates the C=O bond is weakened by interaction with hydroxyl donors, which lowers its vibrational frequency, so the shift points to hydrogen bonding at the interface. The broadening matters too, because it shows the same vibrational mode spread across a range of local environments. Confirming it rigorously needs peak deconvolution, temperature-dependent measurements, or X-ray photoelectron spectroscopy.

Can a wood filler act as a nucleating agent in a PLA matrix?

Yes, and the effect reported here is large. Wood flour raised the apparent crystallinity of the blend from 49.50% in neat PLA to 87.1% at 30 wt% loading, while the melting onset fell from 169.9 to 149.6 degrees Celsius. Cellulose is proposed as the primary nucleating site because its semi-crystalline structure can template the ordering of PLA chains.

Why should you be careful quoting 87.1% crystallinity?

Because it is a blend-level effective value, not a value for the PLA phase alone. The polypropylene fraction in the starch compound melts between 160 and 170 degrees Celsius, overlapping the PLA endotherm near 150 to 155 degrees Celsius, so the two could not be cleanly separated. Confirming the nucleation mechanism would require polarized optical microscopy or X-ray diffraction.

How does work of adhesion differ from contact angle here?

They move in opposite directions, which is why both are needed. Neat PLA measured 80 degrees with a work of adhesion of 85.4 mN/m, while the wood-flour-rich formulation measured 55 degrees with 114.6 mN/m. Every spreading coefficient was negative, from -15.3 to -1.9 mN/m, so none of these surfaces are spontaneously wettable by water.

Why did adding starch compound sometimes increase the contact angle?

Because the commercial starch compound is not pure starch. Thermogravimetric analysis measured it as 63% starch, 31% polypropylene, and 6% plasticizer, and the polypropylene fraction contributes hydrophobicity. The formulation with 60% starch compound and only 10% wood flour gave the highest angle in the set at 92 degrees, above neat PLA at 80 degrees.

Do these composites need a compatibilizer?

The baseline suggests a narrower job for one than usual. Density reached a relative density above 0.95 without an additive, and interfacial hydrogen bonding was present, so the remaining problems are dispersion and the water barrier rather than a dead interface. The contact angle and negative spreading coefficient show the surface is still wettable, and the blend was phase-separated on thermal evidence.

How much weight did the composites lose in soil burial testing?

Between 4 and 7% over 6 weeks. Specimens were discs 2.5 mm thick buried at 15 cm depth in construction soil at pH 8 plus or minus 0.3 and 25 to 30 degrees Celsius, in soil where no plants had been cultivated for at least 10 years to keep the microbial population consistent. Separate samples were prepared for each of the 3, 6, and 9 week time points.

Does higher crystallinity change anything besides strength?

Several things. Greater crystallinity generally raises the glass transition temperature and improves dimensional stability, because intermolecular bonding matters more in the crystalline phase and chains must orient before the material deforms. It also increases resistance to degradation, which means the same material degrades more slowly under the burial conditions.

What were the limits of the spectroscopy evidence?

The authors were explicit about two. The carbonyl shift is indicative rather than conclusive without deconvolution, temperature-dependent measurements, or X-ray photoelectron spectroscopy. And the increased absorption between 1600 and 1650 reciprocal centimeters should not be assigned to a shifted carbonyl, because that region also carries absorbed water bending, lignin aromatic stretching, and overlapping starch signals.