Separating carbon dioxide (CO₂) from nitrogen (N₂) sounds like a small technical detail, but it sits right at the heart of climate and energy challenges.

From power plant flue gas to industrial exhaust streams, CO₂ often comes mixed with a lot of N₂. If we want to capture CO₂ efficiently or upgrade gas streams for cleaner processes, we need separation technologies that are:

• Energy-efficient

• Compact and scalable

• Stable over long operating times

Membranes are one of the most attractive tools for this job. Gas is pushed through a thin selective layer, and one component passes faster than the other. No phase change, no massive energy penalty like distillation.

But designing a membrane that is both highly permeable (gas passes quickly) and highly selective (it clearly prefers one gas over another) is not trivial. That’s where metal–organic frameworks (MOFs)—and more specifically, ZIF-L membranes—come into the story.

In this blog, we’ll walk through a neat piece of membrane engineering:
the heteroepitaxial growth of vertically oriented Co/Zn-ZIF-L molecular sieve membranes that show impressive CO₂/N₂ separation performance.

We’ll keep the chemistry and materials science understandable, but still detailed enough to be useful if you work in membranes, gas separation, or MOFs.

Why MOFs and ZIF-L Are Interesting for Gas Separation

MOFs in a nutshell

Metal–organic frameworks (MOFs) are crystalline materials built from:

• Metal ions or clusters (like Zn²⁺, Co²⁺, etc.)

• Organic ligands (often nitrogen- or oxygen-containing linkers)

The result is a highly porous, tunable framework with:

• Enormous surface areas

• Well-defined pore sizes

• Adjustable chemistry (you can swap metals or ligands)

That makes them attractive for:

• Gas storage and separation

• Catalysis

• Sensing

• Drug delivery and more

ZIFs: a special MOF family

Zeolitic imidazolate frameworks (ZIFs) are a sub-class of MOFs where:

• The ligands are imidazolate-type linkers

• The metal nodes are transition metals like zinc or cobalt

They structurally resemble zeolites but offer more flexibility in chemistry and pore design. They also tend to have:

• High thermal and chemical stability

• Intrinsic microporosity

• Good resistance under harsh conditions

What is ZIF-L?

ZIF-L is a 2D ZIF with a leaf-like crystal morphology. Instead of forming typical 3D cubic crystals like ZIF-8, ZIF-L forms:

• Layered nanosheets

• With cushion-like cavities inside

These features give ZIF-L:

• Strong affinity for CO₂ (thanks to its pore structure and chemistry)

• Potentially very interesting diffusion pathways when assembled into a membrane

However, turning ZIF-L powder into a continuous, defect-free, oriented membrane on a polymer support is hard. And that’s the core problem this research tackles.

The Challenge: Growing Continuous, Oriented ZIF Membranes on Polymers

Inorganic supports like alumina or stainless steel are rigid, stable, and relatively easy to coat with ZIF films. But they’re:

• Expensive

• Less flexible

• Harder to scale into large membrane modules

Polymer substrates (such as porous PAN or polysulfone) are:

• Cheap

• Flexible

• Already widely used in membrane industries

…but they bring two major headaches when you try to grow MOF/ZIF membranes directly on them:

1. Weak adhesion between the crystalline MOF layer and the soft polymer

2. Few nucleation sites on the polymer surface, so ZIF crystals tend to grow sparsely or randomly

The result?

Discontinuous films, pinholes, random orientation, and poor separation performance.

The work discussed here solves this by taking a clever two-step approach:

1. Create a functional, inorganic–organic “template” layer on the polymer

2. Grow ZIF-L vertically and heteroepitaxially into a mixed Co/Zn structure

Let’s unpack that.

Step 1: Building a Template with Aligned Halloysite Nanotubes

To control ZIF-L growth, the researchers first built an engineered interlayer on top of a porous polymer support.

Halloysite nanotubes (HNTs) as a scaffold

Halloysite is a naturally occurring aluminosilicate clay that forms hollow nanotubes. These nanotubes:

• Have a high aspect ratio (long and thin)

• Provide large surface area

• Can be arranged into ordered layers

To make them more compatible with the polymer and ZIF chemistry, the team:

• Modified HNTs with poly(sodium 4-styrenesulfonate) (PSS)

• Dispersed them in a polyvinyl alcohol (PVA) solution

• Deposited this onto a polymer substrate (like PAN) and let the solvent evaporate slowly

During slow evaporation:

• The nanotubes align horizontally, forming a layered, ordered film

• PSS and PVA provide hydrophilic functional groups (sulfonate and hydroxyl), which:

  • Help anchor the nanotubes to the polymer

  • Provide nucleation and binding sites for ZIF growth

The result is a PSS-HNT/PVA interlayer:

• Mechanically robust

• Hydrophilic

• Rich in functional groups

• With large, ordered surface area

This interlayer becomes a template for vertically growing ZIF-L.

Step 2: Vertically Growing Zn-ZIF-L on the Template

Next, the team grows an initial layer of ZIF-L using zinc as the metal node:

• A Zn-based precursor and imidazolate linker solution is brought into contact with the PSS-HNT modified surface.

• Thanks to the functional groups (–OH, sulfonate), Zn²⁺ and ligands nucleate preferentially on the nanotube layer, not randomly in bulk solution.

What emerges is:

• A seed layer of Zn-ZIF-L nanosheets

• Oriented vertically with respect to the nanotube layer

• Continuous and intergrown enough to cover the surface without obvious pinholes

The growth mechanism is roughly:

1. ZIF-L starts spreading along the nanotube surface (horizontal direction)

2. As coverage increases, further growth proceeds out of the plane, giving vertically stacked leaf-like sheets

This Zn-ZIF-L tier acts as a crystalline, oriented base layer for the next stage.

Step 3: Heteroepitaxial Growth of Co-ZIF-L on Zn-ZIF-L

Now comes the most interesting part: turning a single-metal ZIF-L structure into a mixed Co/Zn ZIF-L membrane with improved separation.

What is heteroepitaxy here?

Heteroepitaxial growth means:

• Growing one crystalline material on top of another crystal

• With matching or related lattice structures

• So that the upper layer “inherits” orientation and ordering from the seed layer

In this work:

• Zn-ZIF-L is the seed layer

• Co-ZIF-L is grown on top

Zinc and cobalt versions of ZIF-L are isostructural:

• Same framework topology

• Similar lattice constants

• Different metal ions in the nodes (Zn²⁺ vs Co²⁺)

That makes them ideal partners for heteroepitaxial assembly.

Why not grow Co-ZIF-L directly?

When Co-ZIF-L is grown directly on the polymer (or even on HNTs) under similar conditions:

• Crystallization is too fast, leading to:

  • Random nucleation in solution

  • Poor coverage

  • Large voids between crystals

• The “seed layer” is low quality, so the final membrane is full of defects and has low selectivity

By contrast, when Co-ZIF-L grows on top of Zn-ZIF-L, the situation changes:

• The Zn-ZIF-L surface provides a crystalline template with ordered faces.

• Co-ZIF-L grows in registry with the Zn-ZIF-L crystals, instead of randomly.

• This helps fill grain boundaries and intercrystalline gaps.

The result is:

• A two-component Co/Zn-ZIF-L membrane

• Microns thick (around 4.2 μm)

• With vertically oriented ZIF-L nanosheets

• And an improved grain boundary structure

Elemental mapping confirms:

• Zn is concentrated nearer the bottom (first ZIF-L layer)

• Co is enriched nearer the top (second ZIF-L layer)

So you really do have a stacked, mixed-metal ZIF-L membrane.

Why Vertical Orientation and Grain Boundaries Matter

In a crystalline MOF membrane, gas transport happens via:

1. Intracrystalline diffusion

   • Gas molecules move through the MOF’s well-defined pores

2. Intercrystalline diffusion

   • Gas moves through grain boundaries, defects, and gaps between crystals

To get good molecular sieving, you want:

• Gas transport to be dominated by intracrystalline pores, not random defects

• Grain boundaries to be tight, narrow, and as selective as possible

Vertical orientation and careful heteroepitaxial growth help with both:

• Vertically oriented nanosheets create inter-gallery channels between sheets that can serve as fast, controlled pathways for gas molecules.

• Better alignment and dense stacking reduce non-selective voids.

• Longer Co-ZIF-L growth times allow the upper layer to seal and refine grain boundaries, improving selectivity without completely choking off permeance.

This is exactly what shows up in the gas separation data.

CO₂/N₂ Separation Performance: What Do These Membranes Achieve?

The Co/Zn-ZIF-L membranes are designed to separate CO₂ from N₂—relevant, for example, to flue gas treatment.

Two key metrics are reported:

• CO₂ permeance – how fast CO₂ passes through (in GPU)

• CO₂/N₂ selectivity – how much more easily CO₂ passes relative to N₂

For the heteroepitaxially grown Co/Zn-ZIF-L membranes:

• CO₂ permeance ≈ 244.9 GPU

• CO₂/N₂ selectivity ≈ 17.8 (for mixed gases, slightly above this ~18)

This combination is quite strong, because:

• High permeance means thinner, more productive modules and lower capital cost

• A selectivity close to 18 for CO₂/N₂ is already well above many conventional polymer membranes

For comparison:

• Single-component Co-ZIF-L grown directly showed high permeance but poor selectivity (~3) due to defects

• Vertically grown ZIF-L membranes reported earlier without the Co/Zn mixed strategy reached lower permeance and lower selectivity

Here, the two-layer strategy pays off:

1. Vertically oriented Zn-ZIF-L seeds ensure ordered growth and channels.

2. Slow, stepwise heteroepitaxial growth of Co-ZIF-L refines the grain boundaries and adjusts the pore chemistry.

As Co-ZIF-L growth time increases, you see:

• Defects progressively “healed”

• Selectivity climbing

• Permeance dropping slightly but staying high enough to be attractive

The CO₂ adsorption data support this:
mixed Co/Zn-ZIF-L membranes show higher CO₂ uptake than pure Zn-ZIF-L, consistent with the stronger interaction of Co–N environments with CO₂.

What’s Special About Co vs Zn in ZIF-L?

From a structural and chemical point of view:

• Co–N bonds are stiffer and more ionic than Zn–N bonds.

• This can slightly shrink the effective pore cavities and enhance interaction with CO₂.

FTIR signatures (metal–N stretching frequencies) shift when Co is introduced, reflecting stronger Co–N bonding and a more rigid framework. This is important because:

• A slightly tighter framework helps discriminate between CO₂ and N₂ based on size and interaction.

• Stronger binding with CO₂ (relative to N₂) helps increase selectivity without completely sacrificing diffusion.

Combined with vertical orientation and well-controlled microstructure, this gives the Co/Zn-ZIF-L membranes their impressive CO₂/N₂ performance.

Mechanical Robustness and Stability

A high-performance membrane is only useful if it survives:

• Handling

• Module assembly

• Long-term operation

The composite structure here—polymer support + PSS-HNT layer + Zn-ZIF-L + Co-ZIF-L—has some advantages:

• The aligned HNT layer and PSS/PVA adhesive network provide strong interfacial bonding between the polymer and the MOF layer.

• The ZIF-L layer itself, especially the Co/Zn mixed structure, improves tensile strength and stiffness compared to the bare polymer support.

Stress–strain measurements show:

• Co/Zn-ZIF-L coated supports perform better mechanically than supports with only HNTs or only single-metal ZIF-L.

In gas separation tests over time:

• The Co/Zn-ZIF-L membranes maintain stable CO₂ permeance and selectivity, indicating good structural stability and resistance to degradation under operating conditions.

This is critical if such membranes are to be seriously considered for industrial CO₂ capture or flue gas treatment.

Why This Approach Matters

This work isn’t just “another ZIF membrane.” It showcases several design principles that are broadly useful:

1. Template-assisted growth on polymers

   • Using aligned, functional nanotubes to bridge the gap between soft polymers and crystalline MOFs.

   • This can be extended to other MOFs and other gas separations.

2. Heteroepitaxial assembly of mixed-metal MOFs

   • Combining Zn and Co ZIF-L in a vertical, layered fashion.

   • Taking advantage of each metal’s strengths: structural control from Zn-ZIF-L, stronger binding and tighter pores from Co-ZIF-L.

3. Orientation control of 2D MOF nanosheets

   • Aligning ZIF-L vertically to reduce mass transfer resistance and open up inter-gallery channels.

   • This moves beyond the more common laterally stacked 2D MOF membranes.

4. Grain boundary engineering for selectivity

   • Treating grain boundaries not as an unavoidable defect but as something you can refine via growth strategy and layer sequencing.

Together, these ideas point toward next-generation MOF membranes that are:

• High-flux

• Highly selective

• Mechanically robust

• Compatible with low-cost polymer supports

Exactly the kind of combination we need for large-scale CO₂ capture and gas purification in a carbon-constrained world.