How to Perform PUND Measurements — A Beginner's Lab Manual

A Step-by-Step Guide to Separating True Ferroelectric Switching from Artefacts By Dr. Rolly Verma  |  AdvanceMaterialsLab.com  |  Updated: March 2026  |  Ferroelectrics & Smart Materials

Introduction: Why PUND and Why Now?

In the previous tutorial, we established that conventional ferroelectric measurement — standard P-E hysteresis loop the standard P-E hysteresis loop — can misrepresent the true ferroelectric behaviour of a material. Artefacts arising from leakage currents, dielectric displacement, and parasitic capacitance can produce loop shapes that mimic ferroelectricity even in non-ferroelectric materials. This is a serious scientific problem, particularly in emerging thin film and polymer ferroelectrics where leakage currents are often non-negligible.

This is precisely where PUND — Positive-Up Negative-Down — enters as the definitive solution. PUND is a pulse-based measurement technique that separates the true ferroelectric switching response from all non-switching contributions in the time domain. Rather than relying on loop shape alone, PUND directly subtracts the non-switching current from the total measured current, leaving only the intrinsic domain-switching polarization.

If you are just beginning in ferroelectric research, PUND might appear complex at first encounter. This beginner's manual will walk you through the complete theory, setup, and step-by-step procedure in a clear and accessible manner. Whether you are a student performing your first characterisation experiment, a research scholar validating new materials, or simply curious about advanced material characterisation techniques, this guide is designed to make you confident and hands-on with PUND.

Key Advantages of PUND Over Conventional P-E Measurement

The advantages of PUND analysis over conventional ferroelectric measurements — such as simple P-E hysteresis loops — are significant and well-established in the literature. Let us examine each advantage carefully:

Step-by-Step Guide to PUND Analysis

The following section outlines the complete, systematic procedure for performing PUND analysis. Follow each step carefully to obtain accurate and reproducible results.

The four measurement pulses — P, U, N, and D — each serve a distinct and irreplaceable purpose:

About the Square Waveform

Each of the five PUND pulses is applied as a square waveform — a type of periodic signal that alternates abruptly between two fixed voltage levels—typically a high (positive) and a low (negative or zero)—with a constant duration at each level. It is characterized by its sharp transitions, equal time spent in high and low states (for 50% duty cycle), and flat tops and bottoms. Square waveforms are generated using function generators or digital pulse circuits, and they are widely used in electronic testing, digital clocks, and ferroelectric material analysis (like PUND).

Graphically, a square wave appears as a series of horizontal lines at the high and low voltage levels (fig. 2), connected by vertical jumps at regular intervals—forming a pattern that looks like a sequence of squares or steps. The x-axis represents time, while the y-axis shows the voltage amplitude. A square waveform is a type of non-sinusoidal periodic waveform that alternates between two fixed amplitude levels—high and low—with an equal duration spent at each level (in the case of a 50% duty cycle). It is a special case of a rectangular wave and is often used in digital electronics, timing circuits, and signal processing. The waveform alternates between two voltage levels at a fixed frequency. In a digital circuit, this translates to switching between logic 0 and logic 1.

The square waveform is the preferred choice in PUND for a very important physical reason: it provides a constant electric field during the entire pulse duration. This means the polarisation has the full pulse width to complete switching, leading to more complete domain reversal and more accurate measurement of saturation polarisation. In contrast, sinusoidal or triangular waveforms have a continuously varying field, so the electric field spends only a brief moment at the peak value — underestimating the true switching polarisation.

Case Study: PUND Measurement of P(VDF-HFP) Polymer Thin Film

To make all of the above concrete, let us walk through a complete PUND measurement example using a PVDF-HFP (polyvinylidene fluoride-co-hexafluoropropylene) thin film — a ferroelectric polymer system annealed at 85°C.

dP = P* − P̂ (Equation 1: switchable polarisation at peak field) dPr = Pr* − Pr̂ (Equation 2: switchable remnant polarisation at zero field)

Where:

• P* = switchable polarisation + non-switchable polarisation (total at peak)
• P̂ = non-switchable polarisation only (non-remnant)
• Pr* = switchable remnant + non-switchable remnant
• Pr̂ = non-switchable remnant polarisation only

Applying these equations to our PVDF-HFP case study data:

dP = P* − P̂ = 10.66 − 8.18 = 2.48 mC/m² dPr = Pr* − Pr̂ = 4.31 − 1.84 = 2.47 mC/m²

Step 8b: Interpretation of Results

Once dP and dPr are calculated, interpretation is guided by the following rules:

• Large dP and dPr: Confirms genuine ferroelectric switching — true domain reversal is occurring
• Small or near-zero dP: Suggests leakage-dominated or non-ferroelectric behaviour — P* ≈ P̂ means the material is not truly switching
• dP ≈ dPr: Healthy ferroelectric — the switchable polarisation is predominantly remnant (persists at zero field)
• dP ≫ dPr: Significant non-remnant switching — domains switch during the pulse but relax back towards zero; common in weak or fatigued ferroelectrics

Why PUND Is Scientifically Superior to P-E Loop Measurement

PUND's scientific superiority lies in its fundamental approach: it operates in the time domain, exploiting the fact that ferroelectric domain switching and non-ferroelectric responses have intrinsically different temporal characteristics.

When a ferroelectric domain switches direction under an applied pulse, it generates a sharp, transient current concentrated at the moment of domain wall motion — the polarisation reversal current. In contrast, dielectric displacement currents are proportional to dE/dt (zero during the flat top of a square pulse) and leakage currents are slow, continuous, and present throughout the entire pulse duration.

By applying an identical second pulse (U after P, D after N), PUND captures the non-switching response in isolation. Subtracting this from the first pulse response eliminates all non-domain-related currents with precision. This is fundamentally impossible in a conventional P-E loop measurement, which integrates all current contributions together without separation.

Latest Applications of PUND in Emerging Materials (2024–2026)

PUND has undergone significant development and broadening of application scope in the past two years. Understanding these emerging applications positions you at the frontier of ferroelectric research.

1. Hafnium Zirconium Oxide (HZO) — The Silicon-Compatible Ferroelectric

HZO (Hf₀.₅Zr₀.₅O₂) has emerged as arguably the most important ferroelectric material for semiconductor memory applications due to its CMOS compatibility and scalability to sub-10 nm thicknesses. PUND is now the mandatory validation technique for HZO ferroelectricity, as the material contains competing crystal phases (monoclinic, tetragonal, orthorhombic) and only the non-centrosymmetric orthorhombic phase (Pca2₁) is genuinely ferroelectric.

HZO superlattice structures have been shown to maintain high remnant polarisation values even after non-ferroelectric contributions are excluded by PUND, demonstrating excellent intrinsic ferroelectric performance. Recent studies using PUND on TiN/HZO/TiN capacitors have shown that increasing deposition temperature from 200°C to 300°C increases the 2Pr value by 25%, from 30.2 to 37.8 µC/cm², as measured by PUND pulse schemes at 1 kHz.

2. Nano-PUND and STEM-EBIC Combined Imaging

A powerful new approach called Nano-PUND improves the technique's dynamic range by injecting a current calibrated to cancel linear charging currents due to stray capacitance, addressing a key limitation of conventional PUND in nanoscale devices. When combined with scanning transmission electron microscope (STEM) electron beam-induced current (EBIC) imaging, Nano-PUND can map which specific domains are switching and at what coercive field — providing spatial resolution that conventional transport PUND cannot achieve alone.

3. 2D and Van der Waals Ferroelectrics

PUND-measured polarisation-electric field hysteresis loops have been demonstrated in five-layer hexagonal boron nitride (hBN) devices, confirming ferroelectricity in 2D van der Waals systems where conventional P-E loops would be dominated by leakage artefacts. This represents one of the most challenging applications of PUND — measuring intrinsic ferroelectricity in atomically thin films where the switchable polarisation signal is orders of magnitude smaller than in conventional ferroelectrics.

Common Mistakes and Troubleshooting

Even experienced researchers encounter problems with PUND measurements. Here are the most frequent issues and their solutions:

Conclusion

PUND analysis stands out as a powerful and scientifically rigorous technique for unambiguously distinguishing genuine ferroelectric switching from experimental artefacts. By carefully following the standardised eight-step procedure outlined in this manual, researchers can confidently assess and validate true ferroelectric behaviour in a wide range of materials — from classic PZT and BNT ceramics to emerging HZO thin films and 2D van der Waals ferroelectrics.

The technique's strength lies not merely in its ability to produce a number (dPr), but in the physical clarity of what that number represents: the polarisation that switches under an applied field, persists at zero field, and is genuinely due to ferroelectric domain reversal — nothing more and nothing less. In an era where the definition of ferroelectricity itself is under rigorous scrutiny in emerging material systems, PUND provides the quantitative foundation that the field demands.

Practice Multiple Choice Questions

Key Takeaways

---

Glossary

P*

Switching + non-switching polarisation; contains both remnant and non-remnant components. Measured at peak of the first positive (P) pulse.

Pr*

Switchable remnant + non-switchable remnant polarisation. Measured at zero volts after the P pulse.

P̂ (P-hat)

Non-switching polarisation containing only non-remnant components (leakage, dielectric displacement, capacitive effects). Measured at peak of the U pulse.

Pr̂ (Pr-hat)

Non-switchable remnant polarisation. Measured at zero volts after the U pulse.

dP

True switchable polarisation = P* − P̂. Represents genuine ferroelectric domain reversal contribution at the peak applied field.

dPr

True switchable remnant polarisation = Pr* − Pr̂. The gold-standard metric for ferroelectric characterisation in publication-quality research.

−P*, −Pr*, −P̂, −Pr̂

Mirror parameters in the negative field direction, measured from the N and D pulses respectively.

Preset pulse

The initial −V_max pulse applied before the PUND sequence to initialise all domains into the negative state. No measurement is made.

Coercive field (E_c)

The electric field magnitude at which the ferroelectric polarisation passes through zero during switching. Determines the minimum V_max needed for complete switching.

Internal bias field (E_int)

An asymmetry in the coercive field between positive and negative directions, indicating built-in electric fields from imprint, defect asymmetry, or interface charges.

---

References

1. Ferrodevices, Main Vision Manual IV: Tutorials. Vision software user manual. [Online]. Available: ferrodevices.com (accessed Nov. 2025).
2. R. Verma and S. K. Rout, "Influence of annealing temperature on the existence of polar domain in uniaxially stretched polyvinylidene-co-hexafluoropropylene for energy harvesting applications," J. Appl. Phys., vol. 128, no. 23, Art. no. 234104, Dec. 2020, doi: 10.1063/5.0022463.
3. R. Verma and S. K. Rout, "Frequency-dependent ferro–antiferro phase transition and internal bias field influenced piezoelectric response of donor and acceptor doped bismuth sodium titanate ceramics," J. Appl. Phys., vol. 126, no. 9, Art. no. 094103, Sep. 2019, doi: 10.1063/1.5111505.
4. R. Verma and S. K. Rout, "The Mystery of Dimensional Effects in Ferroelectricity," in Recent Advances in Multifunctional Perovskite Materials, P. Sharma and A. Kumar, Eds. London, U.K.: IntechOpen, 2022, doi: 10.5772/intechopen.104435.
5. H. L. Chan et al., "Nano-PUND and STEM EBIC Imaging for Ferroelectric Polarization Mapping," Microscopy and Microanalysis, vol. 30, Supplement 1, July 2024, doi: 10.1093/mam/ozae044.074.
6. Nature Communications (2025): "Enhancing ferroelectric stability: wide-range adaptive control in epitaxial HfO₂/ZrO₂ superlattices," doi: 10.1038/s41467-025-61758-2.
7. ACS Applied Materials & Interfaces (2025): "Temperature-Dependent {111}-Texture Transfer to Hf₀.₅Zr₀.₅O₂ Films and Its Impact on Ferroelectricity," doi: 10.1021/acsami.4c17978.