When solution-cast at a wet-film thickness of
300–500 µm and dried under forced convection at
60 °C, unplasticised polyvinyl alcohol homopolymer films exhibit a dielectric constant
ε′ ≈ 8–10 at
1 kHz and a dissipation factor below
0.02, making them intrinsically attractive as the dielectric layer in all-printed capacitive sensor arrays. The high density of secondary hydroxyl groups along the chain—typically
1.9 × 10²² cm⁻³—promotes percolation-stabilising interactions with carbonaceous nanofillers, enabling a resistance-strain linearity deviation better than
±3% over a
0–50% elongation cycle when the filler network is immobilised by mild thermal crosslinking. However, these same hydroxyl sites render unmodified films acutely sensitive to ambient moisture; equilibrium water uptake at
85% RH can exceed
30 wt% for grades with a degree of hydrolysis below
98 mol%, causing a reversible conductivity drift of
0.5–1.2 decades per
10% change in relative humidity. The trade-off between environmental stability and processability defines the selection envelope for PVA grades in flexible sensor architectures, where model designations such as
PVA 0588,
PVA 1788, and
PVA 1799 encode the nominal polymerisation degree and hydrolysis level that dictate crystalline domain spacing, solubility windows, and thermomechanical load-bearing limits.
How Hydrolysis Degree and Viscosity Grade Affect Sensor Substrate Performance
The four digits in common PVA bulk-grade nomenclature map directly to processing and end-use properties: the first two represent the average degree of polymerisation scaled by a factor of roughly
100, while the last two give the saponification percentage. A
05-series grade (polymerisation degree
~500, Mw
22 000–25 000 g mol⁻¹) yields aqueous-solution viscosities of
5–6 cP at
4% concentration and
20 °C (Brookfield LV, spindle 1,
60 rpm), suitable for spray-coating thin sacrificial layers. Conversely,
17-series grades (polymerisation degree
~1700, Mw
74 000–81 000 g mol⁻¹) deliver
20–30 cP under identical conditions, producing mechanically coherent films with ultimate tensile strength above
50 MPa (tested per
ASTM D882-18, gauge length
50 mm, crosshead speed
50 mm min⁻¹). Full hydrolysis (
≥98 mol%, e.g.,
PVA 1799) drives crystallinity above
50%, elevating the Vicat softening point to
~105 °C but reducing room-temperature elongation at break to
<10% unless plasticiser is incorporated. Partial hydrolysis (
87–89 mol%, e.g.,
PVA 1788) retains residual acetate groups that disrupt crystallite perfection, keeping elongation at break between
150% and
300% when plasticised with glycerol at
15–25 wt%, yet dissolution onset temperature drops to
~45 °C, restricting continuous sensor operation to
<40 °C ambient. The table below summarises key merit indices for sensor substrate selection across three representative industrial grades.
| Parameter |
PVA 0588 |
PVA 1788 |
PVA 1799 |
| Degree of hydrolysis (mol%) |
87.0–89.0 |
87.0–89.0 |
≥98.0 |
| Viscosity, 4% aq., 20 °C (cP) |
5.0–6.0 |
20.0–30.0 |
25.0–35.0 |
| Tensile strength, unplasticised film (MPa, ASTM D882) |
25–35 |
55–70 |
80–120 |
| Elongation at break, unplasticised (%) |
5–15 |
10–30 |
5–10 |
| Swelling ratio at 90% RH, 25 °C (wt%) |
~45 |
~35 |
~20 |
| Water dissolution temperature, non-crosslinked (°C) |
~15 |
~40 |
≥70 |
For strain sensors that must survive repeated cold-water laundering, fully hydrolysed
1799 matrices crosslinked with glutaraldehyde vapour (
2.5 vol% in N₂,
30 min exposure at
40 °C) achieve gel fractions above
85% and maintain gauge factors within
±10% after
10 immersion cycles in deionised water at
25 °C.
For wearable capacitive pressure sensors where the substrate modulus must approach that of the stratum corneum (
<100 kPa), neat PVA films are an order of magnitude too stiff. Blending
PVA 1788 with poly(ethylene glycol) of Mw
400 at a
30:70 mass ratio lowers the Young’s modulus from
1.2 GPa to
15 MPa (nanoindentation,
Berkovich tip,
500 µN peak load), although the water vapour transmission rate rises to
420 g m⁻² day⁻¹ (
ASTM E96/E96M-22, wet-cup method), necessitating a thin-film encapsulation layer of parylene-C deposited by chemical vapour deposition at
0.5 µm thickness. The addition of a dynamic covalent crosslinker—borax at
1.5 wt% based on PVA dry mass—imparts rapid self-healing (recovery of
85% ultimate tensile strength within
15 s at
pH 8) without obliterating the capacitive signal output measured at
100 kHz with an LCR meter (
Keysight E4980AL). Such compositions are found in pre-formulated masterbatches supplied under technical data sheets that specify a dry-blend moisture content of
<1.0 wt% prior to hot-pressing at
130 °C and
5 MPa.
Controlling Swelling and Ionic Conductivity in Aqueous-Responsive Sensing Layers
When PVA functions as the humidity-sensing polyelectrolyte in an interdigitated electrode architecture, the diffusion coefficient of charge carriers through the swollen matrix determines the impedance-to-humidity transfer function. Uncrosslinked
PVA 1799 films doped with lithium chloride (
8 wt%) and conditioned at
90% RH exhibit an ionic conductivity of
2.3 mS cm⁻¹ (EIS,
10 mV AC amplitude,
1 Hz–1 MHz), but mechanical integrity is lost within
90 min as crystallites solvate. Citric acid crosslinking at
5 wt% with sodium hypophosphite catalyst (
1 wt%, curing
140 °C for
5 min) fixes the gel fraction at
>90% while preserving a usable impedance swing from
10⁷ Ω at
20% RH to
10⁴ Ω at
90% RH. Response time t₉₀ for a
20–80% RH step is
12 s during adsorption and
45 s during desorption, measured in a humidity generator calibrated per
IEC 60068-2-30. The product differentiation here lies in the crosslinker pack compatibility with screen-printing pastes: difunctional aldehydes rapidly increase ink viscosity through imine formation with residual acetate saponification by-products, whereas thermal-activated di- or tricarboxylic acids permit an open pot life of
>8 h at
25 °C without viscosity drift exceeding
5%. Commercially available single-part PVA sensor inks therefore frequently employ oxalic acid at
2.0 wt% as the latent crosslinker, with cure activation occurring above
125 °C in a convection oven.
Direct ink writing of PVA/silver nanowire composite tracks on thermoplastic polyurethane (TPU) substrates for articulating joint motion capture imposes narrow rheological boundaries. A printable ink formulated with
PVA 0588 (
8 wt% in DI water), silver nanowires of average length
35 µm and diameter
90 nm (
3 wt%), and propylene glycol as humectant (
10 wt%) exhibits a yield stress of
95 Pa and a shear-thinning index of
0.42 (power-law fit, shear rate
0.1–100 s⁻¹). Printing through a
200 µm conical nozzle at
15 mm s⁻¹ and
80 kPa back pressure produces filaments that dry to a cross-section of
85 ± 5 µm, delivering a linear resistance of
12 Ω cm⁻¹. Electromechanical stability under cyclic loading to
40% strain (
1 Hz,
5000 cycles) shows a gauge factor drift of
+7% when the substrate is pre-treated with a dilute PVA primer layer containing
0.2 wt% dynamic covalent boron ester bonds—an intervention that suppresses delamination cracks observed at layer-substrate interfaces in control samples after
800 cycles.
When to Select PVA Over PDMS and Thermoplastic Polyurethane for Flexible Sensor Architectures
The choice between PVA and the incumbent stretchable dielectrics—primarily platinum-catalysed addition-cure silicones (PDMS) and polyester-based thermoplastic polyurethanes (TPU)—hinges on four quantifiable performance vectors: dielectric permittivity, moisture-mediated surface regeneration, biocompatibility clearance cost, and ecological end-of-life pathway. PVA’s relative permittivity of
8–10 at
1 kHz is
2.5–3.5 × that of PDMS (
ε′ ≈ 2.7) and
1.5–2 × that of TPU (
ε′ ≈ 5–6), translating directly into higher specific capacitance per unit electrode area and reducing the need for micro-patterned high-aspect-ratio dielectric structures in
<1 pF tactile pixels. Water solubility, often cited as a limitation, enables solvent-free aqueous processing under
EU Directive 2004/42/CE obligations and, when combined with a spray-applied chitosan topcoat (
2% w/v in acetic acid, dried to
5 µm), yields a biodegradable skin-patch form factor that passes cytotoxicity threshold per
ISO 10993-5:2009 (extract dilution test, L929 fibroblasts, viability
>80%). The table below captures key comparative benchmarks among a representative PVA sensor film, a 10:1 Sylgard 184 PDMS slab, and a polyester-based TPU (Estane 58277) film of
100 µm thickness.
| Property |
PVA 1799 (plasticised, 15% glycerol) |
PDMS (Sylgard 184, 10:1) |
TPU (Estane 58277) |
| Young’s modulus (MPa) |
380–520 |
1.2–2.5 |
12–25 |
| Elongation at break (%) |
180–260 |
120–170 |
450–650 |
| Dielectric constant (1 kHz) |
8.2 |
2.7 |
5.5 |
| Water uptake at 100% RH, 24 h (%) |
38 |
<0.1 |
1.2 |
| Gauge factor (5 wt% MWCNT, 0–50% strain) |
75–110 |
8–18 |
25–45 |
| Hysteresis after 1000 cycles (%) |
12–14 |
3–5 |
18–22 |
| Biodegradation standard compliance |
ASTM D6400 (industrial composting) |
None |
None |
| ISO 10993-5 cytotoxicity |
Pass (>80% viability) |
Pass post-extraction |
Pass (>70% viability) |
The markedly higher hygroscopic expansion of PVA—linear swelling coefficient
0.04–0.06 % per
%RH above
60% RH—prohibits its use in unprotected outdoor strain gauges without vapour-deposited diffusion barriers. For dry indoor applications such as bed-occupancy matrices or smart packaging puncture detectors, however, the water-borne layering compatibility and disposal via industrial composting according to
EN 13432 provide a regulatory advantage that neither PDMS nor TPU can match without substantial additive modification. Moreover, when high filler loadings (
>6 wt% carbon black) are required for low-resistance traces, the polar PVA matrix maintains dispersion stability far longer than PDMS due to stronger polymer-filler acid–base interactions; this is evidenced by a
3× longer shelf life for screen-printing pastes stored at
5 °C without stirring.
Slot-die coating of PVA solutions for continuous sensor web manufacture becomes defect-limited at line speeds exceeding
5 m min⁻¹ if the low-shear viscosity falls below
1 200 cP. Thickening the aqueous solution with
0.08 wt% xanthan gum (food-grade, pre-hydrated) establishes a yield-pseudoplastic flow profile that eliminates ribbing instability while allowing levelling to a dried film thickness uniformity of
±2.5% across a
300 mm coating width; the added polysaccharide contributes a
+0.15 increase in loss tangent at
10 Hz, which does not mask the resistive response of the underlying MWCNT network when the mass fraction of carbon remains above
3.5 wt%. Drying is executed in a three-zone forced-air oven set to
60/80/100 °C with a residence time of
4 min, after which the web is calendered between two polished steel rolls (
80 °C, line pressure
40 N mm⁻¹) to densify the film and reduce porosity from
~8 vol% to
<1 vol%, a step critical for consistent capacitive baseline capacitance of
45 ± 1 pF cm⁻² in finished 4×4 taxel arrays. On twin-screw compounding lines used for PVA/carbon masterbatch, a co-rotating intermeshing configuration with
L/D 36 and a mild screw profile (kneading blocks limited to
15% of total length) is specified to cap melt temperature at
210 °C and prevent black speck formation from degraded PVA; the compounded strand is pelletised under a dry nitrogen blanket and immediately sealed in aluminium-lined bags with desiccant, because exposure to
>0.5 g H₂O kg⁻¹ ambient moisture during storage elevates the melt flow rate by
>20% at
190 °C and
2.16 kg (per
ISO 1133-1:2022), rendering subsequent filament extrusion for fused-filament fabrication inconsistent in diameter. Pre-drying at
80 °C for
4 h to a residual moisture of
<0.08 wt% (Karl Fischer titration) restores processability with an ovality tolerance of
±0.03 mm on a
1.75 mm filament—an operational boundary that directly traces back to the product datasheet’s specification of
≤0.3 wt% volatile content as supplied.
