Download PDF

Introduction

Temperature-programmed desorption (TPD) of basic probe molecules is a widely used technique

for characterizing the acid properties of zeolites. By adsorbing a base onto the zeolite surface,

then linearly increasing the temperature under inert gas flow, the desorption of the base can be

monitored.

Quantitative analysis of the desorbed species provides information about:

• Extrinsic acidity (number of acid sites)
• Intrinsic acidity (acid strength, based on desorption temperature)

This approach allows both types of acidity to be evaluated in a single experiment (Figure 1).

Principles of TPD for Acidity Measurement

The area under the desorption peak corresponds to the quantity of acid sites, while the peak

temperature (Tₘₐₓ) reflects the strength of those sites.

Figure 1. TPD experiment: Tₘₐₓ reflects acid strength (intrinsic acidity); peak area reflects number of acid

sites (extrinsic acidity).

Common Probe Molecules

Ammonia (NH₃) is the most commonly used probe due to:

• Small kinetic diameter (0.26 nm), allowing access to virtually all acid sites
• Strong adsorption on sites of varying strength
• Thermal stability over a broad temperature range

Example: Ammonia TPD on H-Y Zeolite

Desorption patterns typically show:

• <150°C: Physically adsorbed ammonia (physisorption). This signal can be minimized by

conducting adsorption at elevated temperatures (∼100°C).

• 200–500°C: Chemisorbed ammonia on acid sites. Multiple peaks may appear, reflecting a

distribution of acid strengths.

Literature Example

Zi et al. (1) observed that increasing the Si/Al ratio in H-Y zeolites resulted in a stronger high

temperature desorption peak, indicating a higher number of acid sites.

Shakhtakhtinskaya et al. (2) correlated desorption signals between 600–900 K (327–627°C) to

Brønsted acid sites, which disappeared upon dehydroxylation.

Correlating Acidity with Catalytic Activity

TPD data can provide insights into catalytic performance.

Example

For H-Y zeolites, the highest ammonia desorption temperature correlated with the cracking

activity of n-pentane, as shown by turnover frequency (TOF) data (3).

Figure 3. Correlation between n-pentane cracking activity (TOF) and the highest ammonia desorption

temperature.

Alternative Probe Molecules

While ammonia is versatile, other probe molecules offer advantages in selectivity and sensitivity

to acid site type.

Pyridine

• Adsorbs on both Brønsted and Lewis acid sites.
• Allows differentiation using infrared (IR) spectroscopy (4, 5).
• Adsorption parameters (temperature and time) are critical to ensure complete coverage,

especially for larger pore zeolites like mordenite (6).

Other Probes

A variety of bases can be employed, chosen based on acid strength and pore accessibility.

Table 1. Common Probe Bases for Acidity TPD

Note: Weak bases are generally used to probe only the strongest acid sites.

Practical Considerations

• Adsorption Temperature: Elevated temperatures reduce physisorption artifacts.
• Adsorption Time: Sufficient to ensure pore diffusion and full surface coverage.
• Reaction Risk: For strong acid sites, probe molecules may undergo side reactions.

Selection should consider thermal and chemical stability.

Summary

Temperature-programmed desorption (TPD) of basic probe molecules is a powerful and flexible

technique for characterizing the acidity of zeolites and related materials. By selecting the

appropriate adsorbate and optimizing adsorption conditions, users can reliably quantify both the

number and strength of acid sites—critical parameters that directly influence catalytic

performance.

All AMI chemisorption analyzers are equipped to perform these TPD experiments with

precision. Whether using ammonia for total acidity measurements or larger probe molecules like

pyridine to selectively assess stronger or Brønsted versus Lewis acid sites, AMI systems offer the

flexibility and control required for high-quality acidity analysis.

With robust temperature programming, sensitive detection options, and easy-to-use software,

AMI’s chemisorption product line enables researchers and catalyst developers to accurately

measure acidity and apply these insights to optimize catalyst design, performance, and longevity.

References

1. Zi, Gog, Yi, Tog, and Yugin, Applied Catalysis, 56, 83 (1989).
2. Shakhtakhtinskaya, A.T. et al., React. Kinet. Catal. Lett., 39, 137 (1989).
3. Shertukde, P.V., Dereppe, J.M., Hall, W.K., Marcelin, G. (Unpublished). Ammonia TPD conducted in-house.

4. Ward, J.W., Journal of Catalysis, 5, 225 (1967).
5. Anderson, M.W. and Klinowski, J., Zeolites, 6, 455 (1986).
6. Karge, H.G., Z. Phys. Chem. Neue Folge, 122, 103 (1980).

Download PDF

Choice of Adsorbate

The choice of adsorbate is critical in temperature-programmed desorption (TPD) experiments. The selected gas should chemisorb selectively on the metal, avoiding sorption on the support or other catalytic components. Ideally, the adsorbate should form a stable monolayer and avoid irreversible reactions with either the metal or the support.

Table 1: suitable adsorbates

Example: CO can react with nickel to form volatile—and hazardous—nickel carbonyl (Ni(CO)₄),

making it unsuitable for certain Ni catalyst systems.

The adsorbate–metal interaction can be visualized using potential energy diagrams (Figure 1).

The initial physisorption step involves a small activation barrier (ΔE₁) and a minor energy release

(ΔH₁). Transitioning to a chemisorbed state requires overcoming a second activation barrier,

which may be small (ΔE₂) or large (ΔE₃). The heat of chemisorption (ΔH₂) is independent of this

barrier.

If the activation barrier is high, the process is classified as activated chemisorption, which

proceeds slowly and may require higher temperatures or longer adsorption times for full surface

coverage.

Case Example:

Hydrogen chemisorption on supported cobalt metal is activated. As shown in Figure 2:

• Curve A: Room temperature adsorption yielded minimal desorption.
• Curve B: Elevated temperature adsorption, followed by cooling in hydrogen, resulted in

full site coverage and a significant TPD signal.

Because theoretical guidance is limited, adsorbate selection typically relies on literature

precedent and practical experience. Table 1 provides a summary of suitable adsorbates for

common catalytic metals.

Choice of

Adsorption Conditions

Adsorption conditions must balance complete surface coverage with minimal side reactions.

Key Considerations:

• Temperature: Sufficiently high to ensure adsorption, but below levels that promote

undesirable reactions (e.g., CO disproportionation to CO₂ and carbon).

• Time: Long enough to allow for equilibrium.

Spillover Warning:

Spillover, where adsorbates migrate from metal crystallites to the support (see Figure 3), is a

kinetically slow process that can distort TPD data if adsorption times or temperatures are too

high.

Recommended Starting Conditions:

• Temperature: 100–200°C
• Time: 30–90 minutes
• Post-adsorption flush: At low temperature to remove weakly held species.

Adsorbate Stoichiometry

Interpreting TPD results requires knowledge of the adsorbate-metal stoichiometry. While direct

measurement is not possible in simple TPD experiments, stoichiometry can be estimated by:

• Comparing chemisorption surface area to BET surface area.
• Infrared (IR) spectroscopy.
• Crystallite size measurements via TEM or XRD.

General Rules:

• Hydrogen & Oxygen: Typically dissociative adsorption (stoichiometry = 0.5).
• CO: Stoichiometry varies widely (0.5 to 2), depending on metal and crystallite size.

In some cases (e.g., very small Rh or Ir crystallites), H₂/M stoichiometries of 1 have been reported,

but these are rare.

When precise stoichiometry cannot be determined, CO uptake can still be used as a relative basis

for comparing catalysts.

Advanced Stoichiometry Determination Using the AMI-300IR

For IR-active adsorbates (such as CO, NO, and selected hydrocarbons), the AMI-300IR provides a

superior method for stoichiometry determination.

By integrating in-situ IR spectroscopy with TPD and chemisorption analysis, the AMI-300IR

enables:

• Quantitative analysis: Peak areas correlate directly to adsorbed species concentrations.
• Speciation: Differentiation between linear, bridged, or multidentate adsorption geometries.

• Dynamic monitoring: Real-time observation of adsorbate behavior during adsorption/desorption cycles.

Example:

During CO adsorption, the AMI-300IR can distinguish linear CO on atop sites from bridged CO

species. This capability enhances stoichiometric precision and provides deeper insight into

adsorption mechanisms.

The AMI-300IR is particularly valuable for:

• Validating adsorption models.
• Confirming stoichiometries in complex or supported catalysts.
• Enhancing the accuracy of TPD interpretations.

References

1. "Chemisorption and Catalysis on Supported Metals", AMI Notes.
2. "Temperature-Programmed Desorption of Adsorbed Species from Catalyst Surfaces", AMI Notes

3. "Measuring Acidity in Zeolites Using TPD", AMI Notes
4. "Surface Area Measurement from Temperature-Programmed Desorption Data", AMI Notes

Download PDF

Introduction

Chemisorption—the formation of chemical bonds between gas-phase molecules and surface atoms—is the foundational step in heterogeneous catalysis. On supported metal catalysts, this process occurs on small metal crystallites anchored to high surface area oxide materials. These chemisorbed species react with additional adsorbed molecules or gas-phase reactants to generate catalytic products.

The chemisorption behavior of a catalyst directly impacts both the reaction rate and selectivity toward desired products. Understanding and quantifying chemisorption is thus essential for both catalyst design and performance optimization.

The Link Between Chemisorption and Catalysis

Optimal catalytic performance requires a balance between the strength and quantity of chemisorbed species:

• Binding Strength:
  Too strong — hinders product formation as molecules adhere too tightly.
  Too weak — reactants desorb before reacting.
  Moderate-strength chemisorption yields the highest catalytic activity, illustrated by the classic volcano curvefor reactions like ammonia synthesis.
• Number of Sites:
  The number of chemisorbed species correlates to the number of surface sites available. More sites translate to higher catalytic productivity.

 Figure 1: Volcano Curve

Measuring Chemisorption

Quantitative assessment of chemisorption requires techniques that can evaluate both the number and strength of adsorption sites:

Among these, TPD offers the most comprehensive data, capturing both quantitative and qualitative characteristics of chemisorption sites.

Temperature-Programmed Desorption (TPD): Principles and Procedure

A standard TPD experiment involves:

1. Sample Preparation:

• Catalyst reduced to yield clean metal crystallites.
• Introduction of the chemisorbing gas (typically at ambient temperature).
  2. Gas Switching & Flushing:
• Replace chemisorbing gas with inert gas.
  3. Controlled Heating:
• Linear temperature ramp.
• Desorption of chemisorbed species occurs at characteristic temperatures.
  4. Detection:
• Quantify desorbed species using calibrated detectors.
• Calculate site quantity and evaluate adsorption strength.

Example:
For H₂ chemisorption on Ni/SiO₂ (H₂:Ni = 1:2), TPD reveals both the number of available Ni sites and the strength distribution of hydrogen binding.

AMI Solutions: Automated Chemisorption Analysis

AMI Chemisorption Analyzers automate the entire process:

• Precise flow controland gas switching.
• Programmable temperature ramps.
• Quantitative detectionand data analysis.
• Fully customizable experiment parametersvia user-friendly software.

The AMI platform delivers reproducible, operator-independent TPD and chemisorption measurements, empowering researchers to optimize catalysts and advance reaction engineering.

Conclusion

Chemisorption is not just a surface phenomenon—it is the gateway to catalytic function. Through advanced, automated analysis tools like the AMI Chemisorption Series, scientists can quantify and understand the key properties driving catalyst performance.

Download PDF

In previous issues of AMI notes, we discussed the application of dynamic techniques—specifically, pulse

chemisorption and temperature-programmed desorption—for catalyst characterization. A frequently asked

question is how results from these dynamic methods compare with those obtained from static volumetric

chemisorption. This note addresses that question by comparing data collected for various catalysts using

both static and dynamic approaches.

Pulse Chemisorption

This technique is particularly well-suited for catalyst-adsorbate systems with relatively fast adsorption

kinetics, meaning the adsorption process is not activated.

For this study, we selected two supported platinum (Pt) catalysts with low metal loading:

• An ASTM standard catalyst containing 0.5 wt% Pt on alumina
• An in-house catalyst with 0.3 wt% Pt on alumina

Carbon monoxide (CO) used as the adsorbate.

While hydrogen, CO, and occasionally oxygen are commonly used adsorbates for static (volumetric)

chemisorption of Pt catalysts [2-5], hydrogen adsorption on Pt exhibits slow kinetics under dynamic

conditions, making it a less suitable choice. Conversely, CO adsorption on Pt is well-established as a rapid

process, making CO the ideal choice for pulse chemisorption experiments.

Results

Table 1 presents a comparison of CO uptake measurements for the two Pt catalysts using both volumetric

and pulse techniques. The results show excellent agreement between the methods. Notably, the pulse

method yielded slightly lower uptake values, likely because the volumetric method also captures weakly

held or "reversible" CO that the pulse technique does not detect.

Advantages of the Pulse Method

One significant advantage of the pulse chemisorption method over the volumetric approach is the time

required to complete the analysis. After the necessary pretreatment—which is similar for both techniques—

a full pulse chemisorption experiment typically takes less than 30 minutes, including system calibration.

In comparison, a standard five-point volumetric measurement can easily require six hours or more.

A second advantage is the ease with which the sensitivity of the pulse method can be enhanced. This can be

achieved by using smaller pulse loops or by diluting the adsorbate in an appropriate inert carrier gas. For

example, a 10% CO in helium mixture can replace pure CO. While this may slightly increase the duration of

the analysis, it significantly improves reproducibility.

Temperature-Programmed Desorption (TPD)

Temperature-programmed desorption (TPD) is especially useful when catalyst–adsorbate kinetics are not

favorable for pulse chemisorption measurements (see Altamira Notes No. 19, Winter 1994) [6]. Cobalt metal

catalysts are a typical example. For supported cobalt catalysts, hydrogen chemisorption proceeds slowly at

room temperature but can be accelerated by raising the sample temperature to approximately 100°C [7].

Carbon monoxide also chemisorbs slowly at room temperature and poses an additional challenge: at

elevated temperatures, CO can disproportionate via the Boudouard reaction:

2CO→C+CO2

In this study, the hydrogen uptake of two cobalt catalysts was measured using both volumetric

chemisorption and TPD. The catalysts contained 20 wt% Co supported on alumina; one catalyst also

included 0.5 wt% ruthenium as a promoter to aid reduction and improve dispersion. Figure 1 presents the

hydrogen TPD profiles of these catalysts. Both exhibited broad desorption profiles, with the Ru-containing

catalyst showing a larger desorption signal, as expected. The key parameters for obtaining reliable TPD

results were an adsorption temperature of 50°C and an adsorption time of 30 minutes.

Table 2 summarizes the hydrogen uptake measured by both TPD and a five-point volumetric method. Once

again, excellent agreement was observed between the two techniques.

As with pulse chemisorption, TPD offers significant time savings compared to traditional volumetric

methods. Furthermore, TPD provides valuable qualitative information about the strength of chemisorption

based on the temperature distribution of the desorption profile.

Summary

This study demonstrates that both dynamic (pulse chemisorption and TPD) and static (volumetric)

chemisorption techniques, when properly applied, yield consistent and reliable results across a variety of

catalysts. Dynamic methods offer clear advantages in terms of simplicity, flexibility, and significantly

reduced analysis time—often cutting hours down to minutes. Additionally, dynamic techniques provide

enhanced sensitivity and the potential for gaining qualitative insights into adsorption strength and surface

interactions, making them ideal for both research and quality control environments.

References

1. "Pulse Chemisorption", AMI Note
2. Wilson, G.R. and Hall, W.K.: J. Catal., 17, 190 (1970).
3. Satterfield. C.N.; “Heterogeneous Catalysts in Practice” , McGraw-Hill, NY(1980).

4. Lemaitre, J.L.; Menon, P.G., and Delannay, F.; “Characterization of Heterogeneous Catalyst” (F. Delannay, ed.) Marcel Dekker, N.Y. (1974).

5. Ferrauto,R.; AlChE Symposium Series, 70, 9 (1974)
6. "Conditions and Parameters for TPD Experiments: Supported Metal Catalysts"AMI Note

7. Reuel, R.C. and Bartholomew, C.H., J. Catal., 85, 78 (1984).

Download PDF

Today, advanced techniques such as TEM, Laser Raman Spectroscopy, EELS, ¹³C NMR, and temperature-programmed oxidation (TPO) are widely used to study coked catalysts. Among these, TPO has become one of the most commonly applied methods due to its simplicity and effectiveness.

This Altamira Note discusses the use of TPO combined with an innovative detection method developed by Dr. S.C. Fung and Dr. C.A. Querini at Exxon Research and Engineering Company. This approach is straightforward and enables continuous monitoring of the rate of coke oxidation.

Experimental

In this TPO method, CO₂—a gas to which flame ionization detectors (FID) are typically insensitive—is converted to CH₄, which is easily detected by an FID. The conversion occurs in the presence of a carrier gas containing oxygen.

For these experiments, an AMI Catalyst Characterization System equipped with a methanator and FID was used. Figure 1 shows the system’s flow diagram.

The methanator consisted of a small reactor filled with a ruthenium catalyst, positioned downstream of the sample U-tube. When hydrogen passed through the methanator, the Ru catalyst quantitatively hydrogenated CO₂ to CH₄. The FID continuously monitored the rate of CH₄ formation, providing a real-time measurement of the coke oxidation rate.

A GC column was unnecessary because the FID is insensitive to oxygen and water vapor in the gas stream.

For these experiments, approximately 20 mg of coked catalyst were loaded into the sample cell. A helium carrier gas containing a low concentration of oxygen flowed over the sample at a rate of 20–80 cc/min. The temperature was increased linearly from room temperature until the complete oxidation of all carbon deposits.

The methanator contained approximately 500 mg of 40 wt% Ru/zeolite 13X. A pure hydrogen stream was injected into the methanator at a flow rate of 22 cc/min. Under these conditions, CO₂ was quantitatively converted to CH₄, while the oxygen in the carrier gas was reduced to water. The combined gas stream then flowed directly into the FID.

The FID continuously monitored the methane generation rate, which was equivalent to the rate of coke oxidation.

Influence of Oxygen Concentration, Flow Rate, and Methanator Temperature

Since TPO experiments require an excess of oxygen, it was necessary to evaluate how oxygen concentration affects the hydrogenation of CO₂ to CH₄ and to establish optimal operating conditions. The effects of flow rate and methanator temperature on the hydrogenation efficiency of the ruthenium catalyst were also investigated, along with the impact of various pretreatments on the Ru catalyst’s activity.

To study the influence of oxygen concentration, pulses of 1%, 2%, and 4.26% CO₂ in helium were introduced into the methanator using a pure helium carrier. Additional CO₂ pulses were introduced using helium carriers containing 0.5%, 1%, and 3% oxygen.

As shown in Table 1, the CO₂ pulses were completely converted to CH₄ except when the oxygen concentration was increased to 3%. One possible explanation for this behavior is that water formed in the methanator (due to the oxygen present in the carrier gas) reduced the equilibrium conversion of CO₂ to CH₄.

CO₂ + 4H₂ → CH₄ + 2H₂O

Higher oxygen concentrations lead to greater water formation, which in turn affects the equilibrium conversion of CO₂ to CH₄. Table 2 shows that TPO experiments should be conducted with oxygen concentrations at or below 3% and at methanator temperatures below 430°C to avoid equilibrium limitations. However, the incomplete conversion of CO₂ to CH₄ observed in Table 1 is not attributable to equilibrium constraints.

An alternative explanation is that water inhibits the methanation activity of the ruthenium catalyst. To investigate this, experiments were conducted varying three parameters: the oxygen concentration in the carrier gas, the methanator temperature, and the carrier gas flow rate. As shown in Table 3, it was necessary to reduce the flow rate when higher oxygen concentrations were used

Additional experiments introduced water directly into the system by saturating the carrier gas at room temperature, producing a 2.6% water concentration in helium. This water level is comparable to that generated during oxidation in a 1.3% oxygen environment. These results confirmed previous findings that oxygen concentrations should ideally remain below 2%. For experiments requiring higher oxygen levels, an oxygen trap can be installed upstream of the methanator. These traps effectively remove oxygen without affecting the CO₂ concentration exiting the sample U-tube.

Temperature-Programmed Methanation Studies

Temperature-programmed methanation experiments were also performed to evaluate the stability of the ruthenium catalyst and to determine the optimal methanator temperature for converting CO₂ to CH₄.

As indicated in Table 3, CO₂ conversion increased with both rising temperature and decreasing flow rate. This behavior suggests that conversion limitations in the presence of oxygen are primarily kinetic in nature. Additionally, the catalyst’s activity improved with temperature. However, the methanator temperature should be kept as low as possible to minimize the risk of sintering the ruthenium particles, which would permanently reduce catalyst activity.

Effect of Carrier Gas Flow Rate and Catalyst Stability

Experimental results indicated that FID sensitivity increased linearly with carrier gas flow rates up to 60 cc/min. At higher flow rates, the FID response plateaued, suggesting that flow rates above this level do not further improve sensitivity.

An additional important observation was the deactivation of the ruthenium catalyst in the methanator due to sulfur poisoning. This deactivation was caused by sulfur oxides generated during the combustion of sulfur-containing coke deposits. The most effective solution was the installation of a sulfur oxide trap upstream of the methanator, which successfully removed sulfur contaminants without affecting the CO₂ concentration.

Conclusions

These experiments demonstrate that TPO coupled with methanation and FID detection is a highly effective technique for monitoring the carbon oxidation rate of coked catalysts. By optimizing experimental parameters, complete conversion of CO₂ or CO to CH₄ is achievable, even in the presence of oxygen-containing carrier gases.

This method is sensitive enough to detect carbon concentrations below 0.1% and can distinguish subtle variations in the coke distribution on catalyst surfaces.

This Altamira Note summarizes a presentation delivered by Dr. S.C. Fung at an AMI (formally: Altamira's) U.S. User's Meeting. For further details on this TPO methodology, see: S.C. Fung and C.A. Querini, "A Highly Sensitive Detection Method for Temperature-Programmed Oxidation of Coke Deposits: Methanation of CO₂ in the Presence of O₂," Journal of Catalysis, 138, p. 240 (1992).

Note: AMI is the only company to integrate a chemisorption analyzer platform with an FID detection system and methanation reactor for advanced TPO studies. This unique configuration enables precise, real-time quantification of coke oxidation rates with unparalleled sensitivity.

Download PDF

Temperature-programmed reduction (TPR) is a powerful technique for obtaining direct information on the reducibility of catalysts and catalyst precursors. It is widely used to characterize a variety of catalyst materials. In a typical TPR experiment, the sample is exposed to a flowing mixture of a reducing agent—such as hydrogen diluted in an inert gas—while the temperature is increased linearly. The consumption rate of the reducing agent is continuously monitored and correlated to the reduction behavior of the sample.

Figure 1 presents a representative TPR profile for a 10% NiO/SiO₂ catalyst using a 10% H₂/Ar gas mixture at a flow rate of 30 mL/min and a linear heating rate of 20 K/min. The resulting profile provides insights into both the ease of reducibility (indicated by the temperature at the reduction peak maximum) and the extent of reducibility (reflected by the signal area).

A comprehensive description of TPR methodology can be found in Temperature-Programmed Reduction for Solid Materials Characterization by A. Jones and B.D. McNicol (Marcel Dekker, Inc., 1986).

However, comparing TPR results across different laboratories or literature reports can be challenging. No universally accepted experimental parameters exist for TPR experiments. Variables such as heating rate, reducing gas composition, flow rate, and particle size can all significantly influence the reduction profile.

This Altamira Note explores the impact of key experimental parameters on TPR results.

Monti and Baiker [1] developed an equation that relates the peak temperature (Tm) to the linear heating rate and hydrogen concentration for a first-order reduction process, as follows:

where:
Tm is the temperature at maximum signal;
[H₂] is the average hydrogen concentration;
rT is the linear heating rate;
Ea is the activation energy of reduction;
R is the gas constant; and
A is a pre-exponential factor.

This equation predicts a decrease in Tm with increasing hydrogen concentration and with decreasing heating rate. It also predicts that the observed temperature maximum is independent of flow rate—a prediction that is not supported by experimental observations.

The primary value of this equation lies in its ability to compare data obtained under different conditions. For example, Gentry and coworkers [2], in a study of CuO, determined Ea to be approximately 67 kJ/mol. Using a flow rate of 20 mL/min, a heating rate of 6.5 K/min, and an H₂ partial pressure of 0.1, they observed Tm = 280ºC. Applying their results to equation (1), it becomes possible to predict Tm for other experimental conditions. Figure 2 illustrates how the predicted Tm for CuO would vary with different hydrogen concentrations and linear heating rates according to equation (1).

The effects of flow rate are more complex and less easily predicted. Intuitively, one would expect Tm to decrease with increasing flow rate, and this trend is confirmed in the literature. Monti and Baiker [1] observed a decrease in Tm of 15°C for supported NiO when the total flow rate was increased from 30 mL/min to 60 mL/min. Similarly, in TPR studies of CuO using 5% H₂ and a heating rate of 20 K/min, increasing the flow rate from 30 mL/min to 80 mL/min resulted in a Tm decrease of 15°C. As a general guideline, a doubling of the flow rate typically results in a Tm decrease of approximately 10–20°C.

Because TPR is a bulk process, not all particles are exposed to the reducing gas simultaneously, making Tm dependent on particle size. However, predicting this dependence is complicated by the specific reduction mechanism involved. Lemaitre [3] examined this dependence across different reduction mechanisms, highlighting two that are especially relevant for catalysis:

• Phase-boundary-controlled reduction, typical of bulk oxides, and
• Nucleation-controlled reduction, typical of supported metals.

Interestingly, the predicted relationship between Tm and particle size varies with the reduction mechanism. For bulk oxides, Tm increases with particle size, whereas for supported metals, Tm decreases as particle size increases.

These various factors—and their combined effects on the TPR profile—are summarized in Figure 3. All should be carefully considered when comparing data from different laboratories or experimental setups.

References

• A.M. Monti and A. Baiker, J. Catal.83, 323 (1983).
• J. Gentry, N.S. Hurst, and A. Jones, J. Chem. Soc., Faraday Trans. I75, 1688 (1979).
• L. Lemaitre, in Characterization of Heterogeneous Catalysts, F. Delannay (Ed.), Marcel Dekker, Inc.

Figure 1. Temperature-programmed reduction (TPR) profile of a 10% NiO/SiO₂ catalyst (10% H₂/Ar, 30 mL/min, heating rate 20 K/min).

Figure 2. Effect of hydrogen concentration and heating rate on predicted Tm for CuO

Figure 3. Relationship between experimental parameters and the observed Tm in TPR experiments.

Download PDF

Background Introduction

Catalysts are at the heart of modern chemical manufacturing, with more than half of all chemical products depending on catalytic processes. Improving catalyst performance requires a deeper understanding of the physicochemical phenomena that occur at the catalyst–reactant interface, including diffusion, adsorption/desorption, surface reactions, and structural changes at the surface [1].  To investigate these processes, researchers study both the reaction mechanism and key kinetic parameters, such as surface coverage and reaction rates. One particularly powerful technique is SSITKA (Steady-State Isotopic Transient Kinetic Analysis), first introduced in the 1970s by Happel, Bennett, and Biloen.

SSITKA enables the measurement of kinetic parameters under true steady-state conditions. By rapidly switching from a reactant gas to its isotopically labeled analog—while maintaining constant flow, pressure, temperature, and catalyst surface state—SSITKA captures real-time information on surface intermediates, site coverages, and turnover rates without disrupting the reaction equilibrium. This application note reviews key literature examples of SSITKA in catalysis and highlights why AMI developed the AMI-200 then 300SSITKA—a fully integrated system designed to meet the growing demand for reliable, precise kinetic measurements.

Definition of SSITKA

SSITKA  is a technique that rapidly switches from a reactant gas to its isotopically labeled counterpart under steady-state conditions. Using mass spectrometry, the system monitors transient response curves as unlabeled reactants and products decrease and labeled species increase. This enables the quantitative study of heterogeneous catalytic mechanisms and surface kinetics. The “steady state” ensures constant flow rate, pressure, temperature, surface coverage, and reactant/product concentrations throughout the switch—minimizing interference from isotopic effects [2]. Since most industrial catalytic processes operate under steady-state conditions, SSITKA provides highly relevant insights into reaction pathways, site coverages, and apparent activation energies. While early versions relied on radioactive isotopes, modern SSITKA systems now use stable isotopes such as ¹³C, ¹⁸O, ¹⁵N, and D₂. The isotopic switch is typically achieved with a fast-acting four-way valve, alternating between labeled and unlabeled feeds. Analysis of the resulting transients reveals key kinetic parameters and mechanistic details.

Applications of SSITKA in Catalysis

SSITKA enables direct measurement of key kinetic parameters, including mean residence time, surface intermediate concentrations, and intrinsic rate constants. These values provide quantitative insight into active site density, reaction pathways, and the influence of supports, promoters, alloying, and deactivation mechanisms.

SSITKA in Catalytic Mechanism Studies

Ethylene, a key building block in organic synthesis, can be selectively oxidized to either acetaldehyde (using PdCl/CuCl catalysts) or ethylene oxide (EO) over silver-based catalysts. Despite decades of study, the identity of the active oxygen species (e.g., Agₓ–O vs. Ag–O–O–Ag) and the dominant reaction mechanism—whether Langmuir–Hinshelwood (L-H), Eley–Rideal (E-R), or Mars–van Krevelen (MvK)—remains under debate. To address this, Professor Israel E. Wachs and his team at Lehigh University combined in situ Raman spectroscopy with SSITKA to investigate ethylene oxidation over Ag/α-Al₂O₃ catalysts and clarify both the nature of the active sites and the operating mechanism [3].

Figure 1: SSITKA Experiment for C₂= + ¹⁶O₂ → ¹⁸O₂

Figure 1 shows mass spectra from SSITKA experiments using AMI-200. After switching from C₂= + ¹⁶O₂ → ¹⁸O₂ (Figure 1b, c), EO and CO₂ signals rapidly decay to zero within ~7 minutes, confirming ethylene epoxidation follows the L-H mechanism (requiring adsorbed ethylene and Ag₄-O₂ species). Post-switch, C₂H₄¹⁶O and C¹⁶O₂ signals gradually decrease, while C¹⁶O¹⁸O rises, indicating MvK (lattice oxygen) participation in CO₂ formation. SSITKA demonstrates that ethylene epoxidation predominantly follows L-H, while complete oxidation involves both L-H and MvK mechanisms.

SSITKA in Fischer-Tropsch (F-T) Synthesis

The conversion of synthesis gas (CO + H₂) into clean fuels and value-added chemicals is a key pathway for the efficient utilization of coal. However, the complexity of the product distribution and low selectivity of the reaction pose challenges to its industrial application. In recent years, cobalt-based catalysts have gained attention for producing high-quality diesel fuels, owing to their relatively slow deactivation and favorable carbon chain growth characteristics. While catalyst additives are known to significantly influence performance, their effects on reaction kinetics remain underexplored.  To address this, Professor Yang Jia’s team at the Norwegian University of Science and Technology, in collaboration with Professor Xiaoli Yang’s group at Qingdao University, modified Co-based catalysts with Rh, Ir, Sb, and Ga to evaluate their effects on Fischer–Tropsch synthesis [4]. SSITKA was used to investigate the catalysts’ intrinsic activity and surface adsorption behavior, providing deeper insight into the role of these additives in reaction kinetics.

Figure 2: (a) CO conversion rate and CH₄ formation rate on CoM/Al₂O₃ catalyst under SSITKA condition. Normalized transient curves of (b) CoA, (c) CoRhA, (d) CoIrA, (e) CoSbA, and (f) CoGaA.

SSITKA experiments involved switching from ¹²CO/H₂/Ar to ¹³CO/H₂/Ar. Normalized transient curves (Figure 2) were analyzed to derive parameters in Table 1:

Table 1: SSITKA Parameters for Co-Based

Through steady-state SSITKA parameter analysis, it is found that: The amount of N_CO surface intermediates of the reactants of CoRh/Al₂O₃ and CoIr/Al₂O₃ was (236 and 234 μmol gcat^-1), respectively, which was 2 times higher than that of CoA (109.6 μmol gcat^-1). The reduction of CoSb/Al₂O₃ and CoGa/Al₂O₃ catalysts is 53 and 39 μmol gcat^-1, respectively. These data indicate that the precious metals Ir and Rh can promote the catalyst and increase more active sites, so that more CO can be adsorbed. The quantity of intermediates on the surface of the product N_CHx has the same trend. However, non-precious metal Sb and Ga auxiliaries showed the opposite trend. At the same time, the residence time of the four auxiliaries was analyzed, and it was found that the residence time of the four auxiliaries had little difference, and the precious metal had no obvious effect on the intrinsic activity of the catalyst, while the non-precious metal reduced the active site and the intrinsic active site.

Figure 3: Relation between CO reaction rate of CoM/Al₂O₃ and concentration of CO and CHx intermediates

Figure 3 compares the CO reaction rate, surface intermediate concentrations, and intrinsic rate constants. The reaction rate of CO was found to be independent of NCO concentration, suggesting that CO conversion is not directly governed by NCO coverage, but rather by the surface coverage of CO and H₂. A linear relationship with NCHₓ concentration was observed, indicating that CHₓ intermediates play a key role in determining the reaction rate of CO.  Figure 4 illustrates the proposed carbon chain growth pathway. The addition of Ir and Rh was found to inhibit carbon chain growth, while Sb and Ga promoted it. These additives influence the surface concentration of CHₓ species during the reaction, which in turn affects both the CO reaction rate and the carbon chain growth rate constant, ultimately altering the product distribution.  Further analysis suggests that these effects stem from electronic modifications introduced by the additives. Changes in the electronic properties of the active metal lead to reduced adsorption strength, thereby shifting surface reactivity and selectivity.

Figure 4: Reaction Pathways for CHₓ Intermediate Conversion to CH₄ and C₂+

In summary, this paper by Yang J. et al. provides kinetic insight into how different additives influence the performance of Co-based catalysts in Fischer–Tropsch synthesis. Their work identifies the key kinetic factors underlying enhanced catalytic activity and offers a foundation for further catalyst optimization.

Steady-State Isotopic Transient Kinetic Analysis (SSITKA) is a powerful technique for quantifying surface intermediates and extracting kinetic parameters under true reaction conditions. With advancements in instrumentation and analytical methods, SSITKA is increasingly combined with complementary approaches such as in situ spectroscopy, kinetic modeling, and DFT calculations to provide a more comprehensive understanding of catalytic mechanisms. Recognizing the growing need for integrated, precise, and user-friendly tools, the AMI-300SSITKA system—a dedicated platform designed to perform reliable SSITKA experiments with high temporal resolution, stable gas switching, and seamless coupling to mass spectrometry. In future articles, we will explore the underlying principles of SSITKA and demonstrate how the AMI-300SSITKA supports advanced catalyst characterization and kinetic analysis.

References

[1] Recent Approaches in Mechanistic and Kinetic Studies of Catalytic, Cristian Ledesma, Jia Yang, De Chen, and Anders Holmen, ACS Catal. 2014, 4, 4527−4547

[2] Li, C. Y., & Shen, S. K. (1999). Steady-state isotopic transient kinetic analysis. Progress in Chemistry, 11(2), 49–59.

[3] Tiancheng Pu, Adhika Setiawan, Revealing the Nature of Active Oxygen Species and Reaction Mechanism of Ethylene Epoxidation by Supported Ag/-Al₂O₃ Catalysts: ACS Catal. 2024, 14, 406−417.

[4] Xiaoli Yang, Jia Yang, Anders Holmen, Kinetic insights into the effect of promoters on Co/Al₂O₃ for Fischer-Tropsch synthesis, Chemical Engineering Journal 445 (2022) 136655.

Download PDF

Building on our previous overview of SSITKA (Steady-State Isotopic Transient Kinetic Analysis), this article delves into the core principles and computational methods behind the technique. In addition, it explores how SSITKA can be integrated with in situ infrared spectroscopy to provide deeper insight into surface reactions and active site behavior.

1. Principle Explanation

1.1 SSITKA Experimental Device

The schematic diagram of the SSITKA experimental setup is shown in Figure 1 [1]. It consists of three main components: a gas flow control system, a reactor, and a mass spectrometer. The gas flow system must support steady-state transient operation, enabling rapid and stable switching between isotopic feeds. It is also essential that both the pre- and post-switch conditions are well-defined and reproducible. The mass spectrometer must be capable of fast response to accurately capture transient signals.

Figure 1 Schematic Diagram of SSITKA Experimental Device

Most SSITKA experiments today rely on microreactor-based systems that are either manual or semi-automated, often leading to operator-induced variability. The AMI 300TKA system addresses this challenge by enabling fully integrated SSITKA experiments through dedicated gas circuit design and coupled mass spectrometry, as shown in the software interface in Figure 2. Transient switching is achieved using a four-way valve, which alternates between two feed streams: Aux Gases and Blend Gases. These streams introduce either the unlabeled reactant (¹²CO) or the isotopically labeled reactant (¹³CO). Upon valve switching, the system seamlessly transitions from ¹²CO to ¹³CO under steady-state conditions.

Figure 2 Software Interface of AMI 300TKA

SSITKA experiments can be executed automatically following the program shown in Figure 3. The fully automated process eliminates the need for manual intervention, significantly reducing the risk of human error and improving the accuracy and reproducibility of test results. The procedure is both practical and user-friendly, ensuring reliable operation even for complex transient kinetic studies.

Figure 3 Kinetic Parameter Solving of SSITKA Program Setting of SSITKA on AMI 300TKA

1.2 Kinetic Parameters

1.2.1 General Parameters

1. The surface residence time (τₚ) of surface intermediates and the surface coverage (Nₚ) are determined by monitoring signal intensity via mass spectrometry, as illustrated in Figure 4(a). Here, P represents the unlabeled product, P* is the isotopically labeled counterpart, and I denotes an inert tracer. At time t = 0, the four-way valve switches, introducing the labeled reactant. As a result, P and I display a decay in signal intensity, while P* shows a corresponding increase. By normalizing the decay profile of P, the transient response curve shown in Figure 4(b) is obtained.

Figure 4 (a) Mass Spectrum (b) Transient Response Curve

Without making kinetic assumptions or defining surface reaction mechanisms, two key parameters can be directly extracted from the transient response curve: the surface residence time (or surface lifetime) τₚ of the intermediate species that form product P, and the surface coverage Nₚ of those intermediates. The expression for calculating Nₚ is given below [2]:

During the experiment, the reactants were rapidly switched while maintaining a constant reaction rate. The reaction rate of the unlabeled product was determined using the following equation [3], where rₚ(t) represents the steady-state reaction rate of the unlabeled product, and rₚ(t) denotes the reaction rate of the isotopically labeled product.

rₚ(t) is the steady-state reaction rate of the unlabeled product, and rₚ(t) is the reaction rate of the isotopically labeled product. As shown in Figure 4(b), normalizing the decay of the unlabeled product yields the transient response curve. The expression for Fₚ(t) is given as follows:

By rearranging and calculating the above three formulas, the surface residence time τ_P of the intermediate species can be obtained.

Integrating the transient response curve yields the surface residence time τ_P, as shown in Figure 5.

Figure 5 Surface Residence Time τ_P

The surface coverage (θ) of intermediates, the reaction rate constant (k), and the turnover frequency (TOF) are calculated using chemisorption techniques to determine the total number of exposed metal atoms (Nₐ) on the catalyst surface.

Assuming that the chemical reaction on the surface is pseudo-first-order, the rate equation of the pseudo-first-order reaction can be expressed as:

Thus, the pseudo-first-order reaction rate constant is obtained:

TOF (turnover frequency) represents the number of catalytic reactions occurring per unit time per active site. It reflects the catalyst’s instantaneous efficiency based on the number of surface active sites. The formula is as follows:

Here, θ represents the surface coverage of intermediate species. The number and distribution of active sites on the catalyst surface can be analyzed using deconvolution techniques, enabling further investigation into the kinetics and mechanism of the catalytic reaction.

1.2.2 Modeling of SSITKA

A heterogeneous catalytic reaction often proceeds through one or more surface-bound intermediate steps. The general parameters introduced earlier—τₚ (the total surface residence time of intermediate species) and Nₚ (the total quantity of surface intermediates leading to product P)—represent the overall behavior of all intermediates involved in forming product P. To distinguish the contribution of individual intermediates, Shannon [4] and Chen et al. [5], building on the work of Biloen et al. [6], proposed a surface reaction mechanism model, summarized in Table 1. This model categorizes reactions into reversible and irreversible types, and further into cases involving single, sequential (series), parallel, or more complex arrangements of intermediate species. Based on these classifications, transient response models were derived using material balance principles. These models allow for the extraction of dynamic parameters such as the quantity and residence time of each intermediate species involved in producing product P.

Table 1 [7]: Mechanism Model, Transient Response, and Kinetic Parameters Obtained from SSITKA

1.2.3 Inferring Catalyst Surface Reaction Mechanisms

Kobayashi et al. [8] demonstrated experimentally that the shape of the transient response curve can provide insights into the underlying reaction mechanism. Building on this, Shannon et al. presented representative response profiles for different irreversible reaction pathways—specifically, cases involving a single intermediate, two intermediates in series, and two intermediates in parallel—as illustrated in Figure 6.

Figure 6 Schematic Diagram of Transient Responses for Different Surface Reaction Mechanisms

As shown in Figure 6, the transient response curves exhibit characteristic S-shaped profiles. Curve (b), which shows the slowest decay, corresponds to two sequential (serial) reaction intermediates—reflecting the extended time required for consecutive reactions at two sites. Curve (a), displaying a single exponential decay, represents a single intermediate species. Curve (c), with a multi-exponential decay pattern, corresponds to two parallel reaction intermediates, where the simultaneous operation of different reaction pathways leads to faster, overlapping decay components. By analyzing these distinct curve shapes, researchers can readily differentiate between kinetic models and infer the nature of surface reactions occurring on the catalyst.

1.3 Factors Influencing SSITKA Experiments

1.3.1 Chromatographic effects and re-adsorption:

In SSITKA experiments, both reactants and products can adsorb not only on the catalyst surface but also on the reactor walls and connecting pipelines, introducing a chromatographic effect. This effect can distort transient responses but may be minimized by reducing the distance from the four-way valve to the detector, increasing gas flow rates, and thermally insulating the transfer lines. The AMI-300TKA system addresses these challenges through the use of 1/16-inch tubing to reduce dead volume, thermal insulation of the valve box, and precise gas flow control via mass flow controllers—measures that collectively enhance measurement accuracy.

Additionally, product re-adsorption can significantly influence transient response data. When re-adsorption occurs at active sites, it reduces catalytic activity and lowers reaction rates. If it occurs at non-reactive sites, the activity remains unaffected, but the measured surface residence time of the product becomes artificially extended—combining the actual intermediate lifetime with the re-adsorption time. To correct for this, inert tracers and empirical correction formulas are commonly applied.

Where τ_inert is the surface residence time during inert gas response, x is empirically taken as 0.5.

1.3.2 Isotope Effect

SSITKA experiments are typically conducted under the assumption of steady-state conditions, with isotope effects considered negligible—that is, the kinetic behavior of isotopically labeled and unlabeled species is assumed to be identical. However, special attention is required when working with hydrogen and its isotopes (e.g., H/D), as the significant differences in mass and bond energy can lead to pronounced kinetic and thermodynamic effects. During H/D isotope exchange, changes in reaction rates and surface intermediates may disrupt steady-state conditions, potentially compromising experimental reliability. As a result, H isotope SSITKA experiments must be approached with caution. Despite these challenges, such experiments are valuable for probing surface reactivity, particularly in identifying bond cleavage events associated with adsorption, desorption, or reaction of specific molecular species.

2. The Use of SSITKA in Conjunction with Spectroscopy

SSITKA is a powerful technique for determining the abundance and kinetic parameters of surface intermediates. However, its primary limitation is the inability to directly characterize the chemical structure of these intermediates or observe their surface reactions in real time. In contrast, in situ infrared (FTIR) spectroscopy enables direct observation of adsorbed species under reaction conditions [9]. By integrating SSITKA with FTIR, it becomes possible to accurately identify surface intermediates—including their chemical structure and surface coverage—and to distinguish reactive adsorbed species from non-reactive ones [10]. Figure 7 shows the AMI-300SSITKA system developed by AMI, which combines SSITKA with in situ infrared spectroscopy for advanced surface characterization.

Figure 7 AMI-300TKA In-Situ Characterization

Since its development in the 1970s, SSITKA has been widely applied to investigate the mechanisms of numerous important industrial catalytic reactions. Today, the integration of SSITKA with spectroscopic techniques enables direct observation of surface intermediates and provides deeper insight into reaction mechanisms. Furthermore, combining SSITKA with complementary methods such as kinetic modeling and density functional theory (DFT) enhances our understanding of reaction pathways. The AMI-300SSITKA is the only commercially available SSITKA system that also serves as a fully featured chemisorption analyzer, offering unmatched versatility in a single platform. By leveraging the strengths of SSITKA alongside these advanced techniques, researchers can obtain a comprehensive picture of catalytic processes under true reaction conditions, facilitating the elucidation of complex chemical mechanisms.

References

[1] Li Chun-yi Shen Shi-hole transient mechanical analysis of steady-state isotopes [J] Advances in Chemistry, 1999,11(2):49-59.

[2] Recent Approaches in Mechanistic and Kinetic Studies of Catalytic, Cristian Ledesma, Jia Yang, De Chen, and Anders Holmen, ACS Catal. 2014, 4, 4527−4547.

[3] Anders Holmen, Jia Yang, and De Chen Springer Handbook of Advanced Catalyst Characterization, Springer Handbooks, Part VII Transient and Thermal Methods,2023,41,935 966.

[4] Shannon, S. L.; Goodwin, J. G. Chem. Rev. 1995, 95, 677−695.

[5] Berger, R. J.; Kapteijn, F.; Moulijn, J. A.; Marin, G. B.; De Wilde, J.; Olea, M.; Chen, D.; Holmen, A.; Lietti, L.; Tronconi, E.; Schuurman, Y. Appl. Catal., A 2008, 342, 3−28.

[6] Kondratenko, E. V. Catal. Today, 2010, 157, 16−23.

[7] Pansare, S.; Sirijaruphan, A.; Goodwin, J. G. In Isotopes in Heterogeneous Catalysis; Hutchings, G. J., Ed.; World Scientific Publishing Co.: London, 2006; Catalytic Science Series, Vol. 4, pp 183−206.

[8] Kobayashi H，Kobayashi M, Catal. Rev-Sci. Eng. 1974, 10, 139.

[9] Yokomizo G H, Bell A T, J. Catal., 1989, 119, 467—482.

[10] Efstathiou A M, Chafik T, Bianchi D et al. J. Catal., 1994, 148, 224—239.

Download PDF

Abstract                                                                        

Hydrogen's high gravimetric energy density (120 MJ/kg) positions it as a critical energy carrier for the transition to clean energy systems. However, its extremely low density at ambient conditions significantly limits its volumetric energy density, complicating storage and transportation. Among various hydrogen storage strategies, solid-state hydrogen storage has emerged as the most promising due to its safety, efficiency, and volumetric advantages. This application note presents a comprehensive analysis of hydrogen storage materials—including Mg-based hydrides, rare earth alloys, carbon-supported systems, and MOFs—evaluated using AMI advanced sorption instrumentation.

Introduction

The volumetric energy density of hydrogen is limited by its low density at ambient conditions—0.0824 kg/m³ compared to 1.184 kg/m³ for air. Methane and gasoline have volumetric energy densities of approximately 0.04 MJ/L and 32 MJ/L, respectively. Hydrogen’s flammability, diffusivity, and explosion risks further challenge its storage [1].

Three primary hydrogen storage methods are commonly used:

• Compressed gas
• Cryogenic liquid
• Solid-state materials

Solid-state hydrogen storage stands out for safety, energy density, and moderate operating conditions, relying on physical adsorption or reversible chemical bonding. The development of high-performance materials is now central to advancing this technology.

Material Classes and Storage Mechanisms

1. Magnesium-Based Hydrogen Storage Materials

Magnesium hydride (MgH₂) offers a theoretical hydrogen capacity of ~7.6 wt% and excellent stability under ambient conditions. However, its high desorption enthalpy (ΔH = 76 kJ/mol) and poor kinetics limit practical use. Agglomeration during cycling also reduces reversibility.

Approaches to Improve MgH₂ Performance:

• Nanoconfinement: Chen et al. [2] encapsulated Mg-V nanoparticles in a 1 nm carbon shell, achieving 6.6 wt% H₂ storage and excellent reversibility (5.2 wt%) at 200–300°C with reduced activation energy.
• Ultrasound-assisted synthesis: Zhang and Liu [2] produced 4–5 nm MgH₂ particles with reversible capacity up to 6.7 wt% at 30°C and 99% retention after 50 cycles.

2. Alloying with Transition and Rare Earth Metals

Alloying Mg with elements like Ni, La, Ce, or Pr forms metastable phases, reducing reaction temperatures and improving kinetics.

• MgNi/Graphene composites (Samantaraya et al. [3]) achieved 5.4 wt% H₂ at 3 MPa due to high dispersion and surface area.
• Mg₃RE alloys (Ouyang et al. [4]) enhanced dehydrogenation rates due to amorphous structure formation and stable cycle behavior.

Figure: Adsorption Kinetics at Different Temperatures & Cyclic Adsorption Performance

3. Carbon-Supported Magnesium Systems

Carbon materials improve dispersion, prevent agglomeration, and enhance hydrogen kinetics via electron transfer and surface defects.

• Mortazavi et al. [5] increased hydrogen storage from 0.7 to 1.5 wt% with CNTs.

4. LaNi₅ and Its Derivatives

LaNi₅ offers:

• High hydrogen capacity (~1:1 H/M ratio)
• Rapid kinetics
• Mild operation conditions (2–10 atm, room temperature)
• Stable multi-cycle performance

Zhu et al. [7] and Liu et al. [8] showed LaNi₅₋ₓCoₓ alloys maintain structure and capacity over 1000 cycles. Substituting La with Pr, Ce, or Gd improves equilibrium pressure and kinetics. Neto et al. [9] demonstrated improved absorption kinetics in PEI-LaNi₅ composite films at 40°C/20 bar.

Figure: Effects of Different Annealing Conditions on Hydrogen Absorption Kinetics

5. Metal-Organic Frameworks (MOFs)

MOFs combine high porosity, surface area, and tunable pore chemistry for physisorption-based H₂ storage.

• Musyoka et al. [9] enhanced Zr-MOF with rGO, increasing capacity from 1.4 to 1.8 wt%.

Experimental Evaluation Using AMI

1. Volumetric Sorption via RuboSorp MPA

Using the RuboSorp MPA, AMI’s high-pressure volumetric gas sorption analyzer, LaNi₅ was tested at room temperature under pressures up to 3 MPa. Results showed rapid hydrogen uptake at low pressures (to 1.35 wt%), saturating at six hydrogen atoms per unit cell—consistent with theoretical expectations. An observable hysteresis loop indicated structural changes in LaNi₅ during hydrogen cycling.

Figure: Pressure vs. hydrogen uptake curve for LaNi₅ using RuboSorp MPA

2. Gravimetric Sorption via RuboSorp MSB

Using the RuboSorp MSB, a high-precision magnetic suspension balance system, real-time weight changes during hydrogen adsorption were recorded. The MSB provides higher accuracy than traditional volumetric systems and enables visualization of subtle structural changes through its unique high-resolution volume acquisition capabilities.

     Figure: Gravimetric hydrogen uptake curve recorded by RuboSorp MSB

3. Thermal Desorption via AMI-300 and AMI-400TPx

Hydrogen desorption kinetics of commercial MgH₂ were evaluated using the AMI-300 and AMI-400TPx. TCD-based thermal desorption testing revealed that faster heating rates increased the hydrogen evolution temperature (421–435°C) and slightly reduced total hydrogen release (7.12 → 6.98 wt%). These results illustrate the importance of thermal uniformity and diffusion efficiency during dynamic desorption.

Figure: Desorption peaks of MgH₂ at varying heating rates

Conclusion

Solid-state hydrogen storage materials—especially Mg-based hydrides, LaNi₅ alloys, and MOFs—offer a viable route to safe, efficient hydrogen storage. Key factors such as thermodynamic stability, kinetic barriers, nanostructuring, and composite design strongly affect performance. Using AMI’s RuboSorp MPA, RuboSorp MSB, and the AMI-300 and AMI-400TPx systems enables precise evaluation of these effects. These instruments provide a comprehensive toolkit to support the development and commercialization of next-generation hydrogen storage materials.

References

[1] Preuster, P., Alexander, A., Wasserscheid, P. Hydrogen storage technologies for future energy systems. Annu Rev Chem Biomol Eng., 8 (2017), 445–475.
[2] Zhang, X., Liu, Y., Ren, Z., et al. Realizing 6.7wt% reversible storage of hydrogen at ambient temperature with non-confined ultrafine magnesium hydrides. Energy & Environmental Science, 14(4): 2302–2313, 2021.
[3] Samantaraya, S. S., Anees, P., Parambath, V. B., Ramaprabhu, S. Graphene Supported MgNi Alloy Nanocomposite as a Room Temperature Hydrogen Storage Material – Experiments and Theoretical Insights. Acta Materialia, 215: 117040, 2021.
[4] Ouyang, L. Z., Qin, F. X., Zhu, M. The hydrogen storage behavior of Mg₃La and Mg₃LaNi₀₋₁. Scripta Materialia, 55(12): 1075–1078, 2006.
[5] Reyhani, A., Mortazavi, S. Z., Mirershadi, S., Golikand, A. N., Moshfegh, A. Z. H2 adsorption mechanism in Mg modified multi-walled carbon nanotubes for hydrogen storage. [J]International Journal of Hydrogen Energy, 2012, 37(5): 1919-1926.

[6] Zhu, Z., Zhu, S., Lu, H., et al. Stability of LaNi5-x Cox alloys cycled in hydrogen—Part 1. Evolution in gaseous hydrogen storage performance. [J] International Journal of Hydrogen Energy, 2019, 44(29): 15159-15172.

[7] Liu, J., Zhu, S., Zheng, Z., et al. Long-term hydrogen absorption/desorption properties and structural changes of LaNi4 Co alloy with double desorption plateaus. [J] Journal of Alloys and Compounds, 2019, 778: 681-690.

[8] Neto, G. R. de A., Beatrice, C. A. G., Leiva, D. R., Pessan, L. A. Polyetherimide-LaNi5 composite films for hydrogen storage applications. [J] International Journal of Hydrogen Energy, 2021, 46(23).

[9] Musyoka, N. M., Ren, J., Langmi, H. W., et al. Synthesis of rGO/Zr-MOF composite for hydrogen storage application. [J]Journal of Alloys and Compounds, 2017, 724: 450-455.

[10] U.S. Department of Energy. Hydrogen Storage Factsheet. Office of Energy Efficiency and Renewable Energy (EERE), 2020.

[11] Jain, I. P., Lal, C., Jain, A. Hydrogen storage in Mg: A most promising material. International Journal of Hydrogen Energy, 2010, 35(10), 5133–5144.

[12] Sandrock, G. A panoramic overview of hydrogen storage alloys from a gas reaction point of view. Journal of Alloys and Compounds, 1999, 293–295, 877–888.

[13] Li, Y., Yang, R. T. Hydrogen storage in metal–organic frameworks by bridged hydrogen spillover. Journal of the American Chemical Society, 2006, 128(25), 8136–8137.

[14] Thommes, M. et al. Physisorption of gases, with special reference to the evaluation of surface area and pore size distribution. Pure and Applied Chemistry, 2015, 87(9–10), 1051–1069.

Download PDF

1.   Introduction

Supported metal catalysts feature catalytically active metals dispersed across porous carriers such as alumina, activated carbon, or silica. These metals are typically present in a microcrystalline form, maximizing surface area and enhancing reactivity. However, in practice, only the surface-exposed metal atoms participate in catalytic reactions—atoms buried within the bulk structure remain inactive.

As a result, metal dispersion—the proportion of surface atoms relative to the total metal content—plays a pivotal role in catalytic performance. This is commonly quantified by IUPAC as:

Dispersion (%) = (Number of surface metal atoms / Total number of metal atoms) × 100

Highly dispersed catalysts offer enhanced activity, selectivity, and resistance to deactivation phenomena such as carbon deposition and sintering. Since many catalysts utilize precious metals, maximizing dispersion not only improves efficiency but also reduces material costs. Thus, accurately measuring dispersion is essential for both technical optimization and economic viability.

A variety of techniques are available to evaluate metal dispersion, broadly categorized into physical and chemical methods [1]. Physical techniques such as X-ray Diffraction (XRD), X-ray Photoelectron Spectroscopy (XPS), and Transmission Electron Microscopy (TEM) estimate dispersion indirectly by assessing crystallite size or surface composition. However, these approaches often require complex modeling and may struggle with heterogeneous or amorphous samples.

In contrast, chemical adsorption methods—such as pulse chemisorption and static chemisorption—offer a more direct measurement by quantifying the amount of probe gas that binds to active metal sites.

These techniques are especially valuable for characterizing the reactive surface area most relevant to catalytic behavior [2].

Chemisorption can also provide insights into crystallite size, active surface area, and the relative contributions of reversible and irreversible adsorption. Despite its power, the static method has some limitations: high-vacuum requirements, longer analysis times for multi-point isotherms, and potential errors from effects like hydrogen spillover [3] or strong metal–support interactions that block access to reactive sites [4].

Figure 1: Representation of metal sites on a support

2.   Methods

Dynamic chemisorption—commonly known as pulse chemisorption—is a widely used technique for measuring the surface-active metal sites in supported catalysts. In this method, reactive gas molecules selectively adsorb onto exposed metal atoms, without interacting with the carrier support.

The experiment is performed under isothermal conditions, typically at ambient temperature and atmospheric pressure. A calibrated sample loop injects fixed volumes of reactive gas into a flowing carrier gas stream. As the gas mixture passes over the catalyst bed, the reactive species adsorb onto available metal sites—often through associative adsorption—while unadsorbed gas continues downstream to a detector, such as a thermal conductivity detector (TCD).

Successive gas pulses are introduced until the catalyst surface becomes saturated and no further adsorption occurs. This saturation behavior, reflected in the detector signal, allows precise quantification of the adsorbed gas and thus enables accurate calculation of metal dispersion and active surface area.

Figure 2: Pulse Chemisorption Instrumentation

Figure 3: “Missing Peaks” representation of the TCD signal

3.   Calculations

In pulse chemisorption, the adsorption quantity is the key parameter for quantitative analysis. It represents the amount of reactive gas adsorbed per unit mass of catalyst, typically expressed in µmol/g. This value is conceptually aligned with physical adsorption but derived through chemical interaction between the adsorbate and active metal sites.

During the experiment, a fixed volume of reactive gas is repeatedly pulsed into a flowing carrier gas stream through a calibrated sample loop. As the gas passes over the catalyst bed, it interacts with exposed metal sites, while unadsorbed gas is carried to a thermal conductivity detector (TCD), generating a series of pulse peaks (Figure 3).

As the surface nears saturation, gas uptake decreases and the detector signal stabilizes. The final peak, corresponding to complete saturation, serves as the baseline for calculating gas uptake in earlier pulses.

Quantification of adsorption is based on comparing the area of each unsaturated pulse to the average area of saturated peaks. Two primary calculations are used:

Quantitative Correction Value (Cv):

Cv = (V_loop × C_gas) / (ΣA_sat / n_sat)

Where:

• V_loop = Volume of the sample loop
• C_gas = Concentration of the adsorptive gas
• ΣA_sat = Sum of saturation peak areas
• n_sat = Number of peaks used for saturation averaging

Sample Adsorption Amount (Uptake):

Uptake = Cv × Σ(A_i - A_sat-avg)

Where:

• A_i = Area of each unsaturated pulse
• A_sat-avg = Average saturation peak area (= ΣA_sat / n_sat)

The calculated adsorption quantity forms the basis for further analysis of catalyst structure, including:

• Metal dispersion (%)
• Crystallite size (nm)

Required known values:

• Metal loading (wt%)
• Relative atomic mass (g/mol)
• Stoichiometric factor (reaction-specific)

The stoichiometric factor reflects the number of metal atoms associated with each adsorbed gas molecule and depends on the adsorption mechanism:

• H₂ adsorption (dissociative): SF = 2
• CO adsorption:

- Linear: SF = 1

- Bridging: SF = 0.5

- Multi-type (on oxides): SF = 1–n

Metal dispersion indicates the percentage of metal atoms located on the surface:

Dispersion (%) = [Adsorption (µmol/g) × Relative atomic mass (g/mol)] / [Metal loading (%) × Stoichiometric factor × 100]

Where:

• Adsorption (µmol/g): Calculated from pulse chemisorption
• Relative atomic mass: e.g., Pt = 195.08
• Metal loading (%): From catalyst specification
• SF: From adsorption mechanism

Crystallite size can be estimated using geometric models. Two common models are:

Hemispherical Model:

Particle diameter (Å) = 6 × 10⁶ / [Density (g/cm³) × Max SSA (m²/g) × Dispersion (%)]

Cubic Model:

Cube edge length (Å) = 5 × 10⁶ / [Density (g/cm³) × Max SSA (m²/g) × Dispersion (%)]

Where:

• Density (g/cm³): e.g., Pt = 21.45
• Max specific surface area (SSA): From chemisorption data
• Dispersion (%): From formula above
• 1 Å = 0.1 nm

Note: These formulas are valid for single-metal catalysts. For bimetallic or alloy systems, peak separation via Temperature-Programmed Desorption (TPD) is recommended for accurate analysis.

Figure 4: Software Calculation Interface

4.   Experiment

The metal dispersion degree of a 1 wt% Pt/CeO₂ catalyst was measured using the AMI-300 chemisorption analyzer, known for its high performance and precision in pulse chemisorption experiments.

• Sample mass:0816 g
• Instrument used: AMI-300
• Adsorptive gas: H₂
• Method: Pulse chemisorption
• Detection: Thermal Conductivity Detector (TCD)

The sample underwent a pre-treatment process prior to measurement. The conditions are outlined below:

The pulse chemisorption experiment was conducted under the following operating parameters:

• Gas flow rate: 30 cm³/min
• Pulse volume (quantitative loop): 57 μL
• Test temperature: 50 °C

Following the pre-treatment, a series of gas pulses were introduced to the catalyst sample. The resulting TCD response curve reflects the consumption of hydrogen gas over successive pulses until adsorption saturation was reached.

Figure 6: TCD for Pulse Chemisorption

Based on the TCD signal and experimental conditions, the hydrogen pulse chemisorption analysis of the 1 wt% Pt/CeO₂ sample yielded the following results:

• Adsorption capacity: 4.742 µmol/g
• Metal dispersion degree: 18.5%
• Metal surface area: 43.477 m²/g
• Estimated crystallite size:

• Spherical model diameter:4338 nm
• Cubic model edge length:3615 nm

These results indicate a moderately dispersed Pt phase on the CeO₂ support, with nanoscale crystallites and a high accessible metal surface area of 43.5 m²/g. A dispersion value of 18.5% is typical for platinum catalysts prepared by conventional impregnation methods and subjected to high-temperature calcination, where dispersion often ranges between 10% and 30%. These characteristics suggest the catalyst is well-suited for applications requiring accessible Pt active sites, such as hydrogenation or oxidation reactions.

1.   Discussion

Accurate quantification of metal dispersion by pulse chemisorption depends on several experimental variables. The following factors can significantly impact data quality and should be carefully considered to ensure reproducible and reliable results.

1. Selection of Adsorption Gas and Measurement Method

Some noble metal-supported catalysts (e.g., Pt, Pd, Rh) exhibit the hydrogen spillover effect when using H₂ as the adsorptive gas. This can lead to overestimated dispersion values, occasionally exceeding 100%.

Cause:

Hydrogen dissociates on the metal surface, forming atomic hydrogen that migrates onto the support material (typically a metal oxide). The detector then incorrectly attributes this additional uptake to the metal.

Recommended Solutions:

• Use CO as the probe gas to avoid spillover.
• If H₂ is required, cross-check results with complementary methods such as TPD or TEM.

2. Quantitative Loop Size and Gas Volume

If the first pulse peak is similar in area to later pulses, the sample may be saturated on the first injection—leading to poor resolution of adsorption behavior.

Cause:

The sample loop volume is too large relative to the adsorption capacity of the catalyst.

Recommended Solution:

• Use a smaller-volume loop to better capture the progressive uptake profile.
• The AMI-300 chemisorption analyzer offers interchangeable quantitative rings to match loop volume to sample capacity.

3. Gas Concentration Optimization

A flat adsorption curve may indicate that the gas concentration is too high, resulting in saturation within a single pulse.

Cause:

High adsorbate concentration delivers more reactive gas than the catalyst can gradually adsorb.

Recommended Solution:

• Lower the concentration of the adsorptive gas to allow a more gradual uptake.
• The AMI-300 system features four wide-range, high-precision MFCs (5–100 mL/min) that enable accurate gas mixing—even down to 0.0025% concentration for trace-level analysis.

4. Incomplete Saturation After Multiple Pulses

In some cases, saturation may not be reached even after many gas pulses.

Cause:

The adsorbate volume per pulse is insufficient for the catalyst’s capacity.

Recommended Solutions:

• Increase the pulse volume by selecting a larger loop.
• Raise the adsorbate concentration to improve the delivered dose per injection.
• The AMI-300’s flexible gas mixing and modular loop system support easy adjustments.

5. Temperature Effects on Adsorption Accuracy

Temperature has a major influence on adsorption behavior and data accuracy.

Potential Issues at Elevated Temperatures:

• Hydrogen spillover
• Unwanted side reactions between gas and support
• Thermal decomposition or dissociation of the adsorbate

Recommended Practices:

• Perform tests at or near room temperature, which is standard for most metal–gas systems.
• For some sensitive measurements, low-temperature adsorption improves accuracy and minimizes spillover.

Example Conditions:

• For Pt with H₂ or CO: Test at room temperature or 195 K
• The AMI-300’s integrated cooling module enables testing as low as 143 K (–130 °C) for enhanced control and resolution.

Figure 7. Pulse loops available for the AMI-300 chemisorption system.

6.   Conclusions

The AMI-300 chemisorption analyzer provides precise, reliable measurement of surface metal dispersion and crystallite size in supported metal catalysts. By enabling control over key parameters—such as gas type, concentration, pulse volume, and temperature—the system supports detailed investigations into:

• Surface chemistry and metal–support interactions
• Catalyst activity and efficiency
• Reaction mechanisms and intermediates
• Deactivation behavior and regeneration strategies

With its simple operation and high repeatability, pulse chemisorption using the AMI-300 is an indispensable tool for researchers and engineers across a wide range of catalytic and materials science applications.

2.   References

[1] Whyte T E.Catal Rev, 1973, 8:  117-145

[2] Yang Chunyan, Yang Weiyi, Ling Fengxiang, Fan Feng. Determination of Surface Metal Dispersion of Supported Metal Catalysts [J]. Chemical Industry and Engineering Progress, 2010, 29(8): 1468-1501.

[3] Liu Weiqiao, et al. Practical Research Methods for Solid Catalysts [M]. Beijing: China Petrochemical Press, 2000: 38-39, 44, 230-232.

[4] Chen Songying, et al. Adsorption and Catalysis [M]. Zhengzhou: Henan Science and Technology Press, 2001: 124-125

Download PDF

Research Background

Heterogeneous catalytic processes are extremely complex surface physicochemical processes. The main participants in these processes are catalysts and reactant molecules, primarily involving the cyclic repetition of elementary steps, including diffusion, chemical adsorption, surface reaction, desorption, and reverse diffusion. The most critical steps are adsorption and surface reaction.

Therefore, to elucidate the intrinsic role of a catalyst in a catalytic process and the interaction mechanism between reactant molecules and the catalyst, it is necessary to investigate the catalyst's intrinsic structure (e.g., specific surface area and pore structure), adsorption properties (e.g., structure of adsorption centers, energy state distribution, adsorption states of molecules on adsorption centers), and catalytic properties (e.g., nature of active catalytic sites, metal dispersion). Further studies should also include mass transfer processes, reaction mechanisms, long-term stability, and pilot-scale evaluation to comprehensively assess catalyst effectiveness.

Solutions

2.1 Pore Structure

Heterogeneous catalytic reactions occur on the surface of solid catalysts. To maximize reaction activity per unit volume or weight, most catalysts are designed with porous structures to increase surface area. The porous structure and pore size distribution directly influence diffusion and mass transfer, which in turn affect catalytic activity and selectivity.

Zeolite catalysts are a class of microporous crystalline materials with uniform pore structures and extremely high surface areas. Figure 1 demonstrates nitrogen adsorption-desorption curves. In the low-pressure region, nitrogen adsorption sharply increases due to micropore filling. The HK pore size distribution reveals the most probable diameter is 0.57 nm, and the BET surface area is calculated as 675 m²/g.

FIGURE 1: Zeolite N₂ Adsorption-Desorption Isotherms (A) Linear Scale, (B) Logarithmic Scale, (C) HK Micropore Size Distribution

ctivated alumina is another widely used catalyst support. It shows excellent surface acidity and thermal stability. Figure 2 shows adsorption curves of two alumina samples with surface areas of 192.32 m²/g and 210.81 m²/g. BJH analysis indicates pore size peaks at 3 nm and 21 nm.

Nickel-loaded cerium dioxide (CeO₂) catalysts are notable for their redox cycling and oxygen storage. Figure 3 shows how increasing Ni loading reduces surface area (from 15.25 to 7.59 m²/g) and pore volume, due to Ni occupying surface sites.

FIGURE 3: (a) Ni@CeO₂ N₂ Adsorption-Desorption Isotherms, (b) BJH Pore Size Distribution

2.2 Active Centers

The intrinsic active sites of catalysts are key to their reactivity. These are best characterized dynamically under operating conditions. AMI systems apply temperature-programmed techniques (TPx series) for this purpose.

Temperature-Programmed Reduction (TPR)

TPR measures reducibility and interactions of active metal oxides with supports. Figure 4(a) shows a metal-supported alumina catalyst with a single strong reduction peak at 234°C and hydrogen consumption of 9680 µmol/g. Figure 4(b) shows three distinct peaks for Mn₂O₃ → Mn₃O₄ → MnO → Mn transformations. Figure 4(c) compares different Ni loadings on CeO₂, showing increasing reduction temperatures and decreasing hydrogen consumption as Ni loading increases.

FIGURE 4: H₂-TPR (a) Alumina-supported Catalyst, (b) Mn₂O₃-based Catalyst, (c) Ni@CeO₂ Catalyst Temperature-Programmed Oxidation (TPO)

TPO is used to evaluate coke deposition and regeneration conditions. Figure 5 shows TPO data for a Cr₂O₃ catalyst after reaction, with three peaks at 500°C, 578°C, and 631°C. The high-temperature coke species dominate, indicating a need for high-temperature regeneration.

Conclusion

Comprehensive catalyst characterization requires more than surface-level insight. From understanding pore architecture to quantifying redox behavior and coke formation, each technique provides a piece of the puzzle.

The integration of N₂ physisorption, TPR/TPO/TPSR, and pulsed chemisorption is essential for evaluating both performance and durability under realistic reaction conditions.

AMI’s chemisorption and physisorption instrument platforms enable researchers to conduct these analyses with precision and flexibility, combining automated gas handling, programmable thermal profiles, and real-time detection in a unified workflow.

Whether developing next-generation catalysts or optimizing industrial formulations, AMI provides the tools needed to accelerate R&D and ensure reliable, reproducible results.

With proven solutions for academia, government labs, and industry, AMI continues to support catalyst development from lab-scale discovery to pilot-scale deployment.

Download PDF

1. Introduction

Electronic specialty gases are essential foundational materials in modern electronics manufacturing—often referred to as the "blood" or "food" of the industry. These high-purity gases are critical to the production of semiconductors, display panels, LEDs, and photovoltaics. In 2021, the demand from these sectors accounted for 43%, 21%, 13%, and 6% of total consumption, respectively [1].

The purity requirements for electronic specialty gases are stringent, typically at the 5N (99.999%) level, with some applications demanding 6N (99.9999%) or even higher. Gas purity and quality directly influence device yield and performance, making both synthesis and purification technologies central to specialty gas production. Common synthesis methods include electrolysis, chemical reactions, and combined electrochemical approaches, while purification techniques span adsorption, distillation, absorption, and membrane separation [1].

Semiconductor manufacturing, comprising thousands of highly specialized steps, relies on over 100 types of specialty gases—44 of which are commonly used. These include compounds such as trifluorine nitride, tetrafluorocarbon, and sulfur hexafluoride, typically supplied in liquid or pressurized gas form.

Each major semiconductor process uses specialty gases for critical roles:

• Cleaning: Removes contaminants from wafers and surfaces using gases such as SF₆ and CF₄.
• Coating (CVD/ALD): Deposits films through gas-phase reactions using precursors like tungsten hexafluoride, silane, ammonia, and nitrous oxide.
• Lithography: Involves plasma formation from mixed gases (e.g., Ar/F/Ne, Kr/Ne) to generate stable, high-precision light sources for photomask patterning.
• Etching: Selectively removes material using fluorocarbons such as CH₃F, CH₂F₂, CF₃H, and halogen gases like Cl₂ and HBr to define intricate microstructures.
• Doping: Introduces conductivity to semiconductors using gases like arsine, phosphine, diborane, and trifluoroborane [2].

These processes demand precise control and validation of gas delivery, purity, and usage conditions—making gas analysis, gravimetric sorption, and breakthrough testing indispensable tools in quality assurance and process development.

The adsorption method leverages the principle that porous materials exhibit selective gas uptake depending on the molecular properties of the gas and the characteristics of the material. For a given porous adsorbent, different gases exhibit different adsorption capacities. When a gas mixture is passed through an adsorbent bed, gases with stronger affinities to the surface will be preferentially adsorbed, while those with weaker interactions will exit the system—achieving a separation effect. The adsorbed gas can subsequently be desorbed by thermal regeneration or gas purging, allowing for recovery of purified components [1].

One prominent application of this technique is the separation of xenon (Xe) and krypton (Kr)—a challenging and high-value target for gas purification. Xe and Kr are critical to industries such as semiconductors, medical imaging, aerospace, and lighting.

Wang et al. developed a negatively charged coordination ultramicroporous material, NbOFFIVE-2-Cu-i (ZU-62), that demonstrated a breakthrough in Xe/Kr separation performance [3]. ZU-62 features a finely tunable pore size and flexible framework, resulting in a unique inverse size sieving effect—a rare case where the larger atom (Xe) is selectively adsorbed over the smaller atom (Kr). As shown in FIGURE 1, static adsorption isotherms confirm high Xe uptake and preferential exclusion of Kr due to the material’s precise pore chemistry and geometry.

To validate the material's real-world separation efficiency, dynamic breakthrough experiments were performed at 273 K (FIGURE 2). Kr eluted early, while Xe showed delayed breakthrough—demonstrating strong selective adsorption of Xe. The system achieved production of >99.9% pure krypton from the effluent stream and record Xe capture of 206 mL/g (2.88 mmol/g), aligning well with isotherm data and confirming the industrial relevance of the material [3].

Figure 1 Adsorption and desorption isotherms of pure components of Xe and Kr by ZU-62 at 273 K and 298 K

Figure 2 (A) Pore penetration experiment of Xe/Kr (20/80) mixture at 273 K and 1 bar conditions; (B) Desorption signals of Xe and Kr during regeneration process at 298 K with a flow rate of 3.5 mL/min-1 of N₂, the green curve represents the real-time cumulative purity of Xe.

2. Experiment

Sample Selection and Objective

Two metal-organic framework (MOF) materials were selected for adsorption and gas separation studies:

• MOF-1: Targeted for SF₆/N₂ adsorption and separation
• MOF-2: Targeted for Xe/Kr adsorption and separation

Both dynamic breakthrough experiments and static isotherm measurements were performed to evaluate the adsorption capacity and selectivity of each material under relevant conditions.

Dynamic Breakthrough Testing

Competitive adsorption experiments were performed using the BTSorb-100 Breakthrough Curve and Mass Transfer Analyzer from AMI. A 1 mL stainless steel adsorption column (inner diameter: 0.45 cm; bed height: 5 cm) was used for all tests.

• SF₆/N₂ separation (MOF-1):

• Temperature: 298 K
• Pressure: 2 bar

• Total flow: 30 mL/min
  • Composition: 13.5 mL/min N₂, 1.5 mL/min SF₆, 15 mL/min He (carrier)

• Xe/Kr separation (MOF-2):

• Temperature: 25°C
• Pressure: 1 bar
• Total flow: 3 mL/min
  • Composition: 0.6 mL/min Xe, 2.4 mL/min Kr

Breakthrough curves were recorded and used to assess gas retention times, separation resolution, and dynamic capacity for each system.

Static Adsorption Testing

Static adsorption isotherms were measured using the AMI Sync 400 Surface Area and Pore Size Analyzer. For both MOFs:

• Sample mass: 0.12 g
• Pretreatment: Vacuum degassing at 120°C for 12 hours
• Measurement conditions:

• SF₆/N₂ isotherms at 273 K (MOF-1)
• Xe/Kr isotherms at 298 K (MOF-2)

The resulting data enabled comparison between static adsorption capacity and dynamic performance, offering insight into material structure-performance relationships.

3.Results and Discussion

Perfluorinated electronic specialty gases—including NF₃, CF₄, and SF₆—play a critical role in the fabrication of silicon-based semiconductors due to the strong chemical reactivity between fluorine and silicon. However, the conversion efficiency of F-gases in plasma processes is often below 60%, leaving unreacted F-gases and byproducts such as N₂, NOₓ, HF, and H₂O in the exhaust stream. As environmental regulations tighten and gas recovery becomes increasingly important, adsorbent-based purification has emerged as a key area of research and development.

To assess adsorption-based purification of SF₆, MOF-1 was tested for its separation capability in an SF₆/N₂ gas mixture.

Static Adsorption Behavior

As shown in FIGURE 3(a), static adsorption isotherms indicate that MOF-1 exhibits a significantly higher adsorption capacity for SF₆ compared to N₂. This disparity suggests the potential for selective adsorption and separation of SF₆ from nitrogen under controlled conditions.

Dynamic Breakthrough Performance

To evaluate real-world separation behavior, a dynamic breakthrough experiment was conducted using the BTSorb-100 system. A 10:90 SF₆/N₂ gas mixture was introduced at 298 K, 2 bar, and a total flow rate of 30 mL/min.

As shown in FIGURE 3(b), the breakthrough curve illustrates clear separation behavior:

• N₂ began to elute at approximately 50 seconds
• SF₆ broke through at around 250 seconds
• The resulting separation window was approximately 200 seconds, representing the effective retention time of SF₆ on MOF-1 under these conditions

Calculated dynamic adsorption capacities were:

• N₂: 0.905 mmol/g
• SF₆: 0.857 mmol/g

These results were consistent with the static isotherm values, where MOF-1 adsorbed 1.01 mmol/g of N₂ at 95 kPa and 0.896 mmol/g of SF₆ at 9.28 kPa.

The close agreement between static and dynamic data confirms the reliability and practical relevance of the measured separation behavior. Taken together, these findings demonstrate that MOF-1 offers viable adsorption and separation performance for SF₆/N₂ systems, and highlights its potential use in semiconductor exhaust gas purification processes.

Figure 3 (a) Adsorption isotherms of SF₆ and N₂ on MOF-1 at 298K; (b) Competitive adsorption breakthrough curve of MOF-1 at a temperature of 298K, a total pressure of 2 bar, and a total flow rate of 30 mL/min (He/N₂/SF₆ = 50/45/5).

Xenon (Xe) and krypton (Kr) are high-value noble gases, often referred to as "golden gases" due to their scarcity and wide-ranging industrial applications. Their unique physical and chemical properties make them essential in fields such as semiconductor manufacturing, lighting, aerospace, medical imaging, and anesthesia [4]. As a result, the efficient separation and purification of Xe from Kr remains a high-priority challenge in specialty gas processing.

In this study, MOF-2 was evaluated for its ability to adsorb and separate Xe/Kr mixtures.

Static Adsorption Isotherms

As shown in FIGURE 4(a), adsorption isotherms measured at 273 K demonstrate that MOF-2 exhibits a significantly higher adsorption capacity for Xe compared to Kr. When the temperature was increased to 298 K (FIGURE 4(b)), the adsorption capacities of both gases decreased—consistent with exothermic physisorption behavior—but Xe remained more strongly adsorbed than Kr, suggesting favorable selectivity across a range of operating conditions.

These static results confirmed MOF-2’s potential for selective Xe adsorption from a Xe/Kr mixture.

Dynamic Breakthrough Testing

To evaluate separation performance under flow conditions, a competitive dynamic breakthrough experiment was conducted using a 20/80 (v/v) Xe/Kr binary gas mixture. The test was carried out at 298 K, 1 bar, and a total flow rate of 3 mL/min.

As shown in FIGURE 4(c):

• Kr broke through the adsorption column in under 500 seconds
• Xe did not appear in the effluent stream until ~2500 seconds
• The breakthrough time difference exceeded 2000 seconds, indicating a long and effective separation window

Calculated adsorption capacities were:

• Xe: 1.5489 mmol/g (dynamic), 1.6749 mmol/g (static at 20 kPa)
• Kr: 0.9542 mmol/g (dynamic), 1.025 mmol/g (static at 80 kPa)

The close agreement between static and dynamic data, along with the large differential in breakthrough times, confirms that MOF-2 offers strong selectivity and capacity for Xe over Kr. These results demonstrate MOF-2’s practical utility for Xe purification from Kr-containing gas mixtures, with potential applicability in industrial gas recovery and semiconductor process optimization.

Figure 4 (a) Adsorption isotherms of Xe and Kr by MOF-4 at 273K; (b) Adsorption isotherms of Xe and Kr by MOF-4 at 298K; (c) Dynamic competitive adsorption curves of Xe/Kr by MOF-4 under normal temperature and pressure, total flow rate of 3 mL/min (Xe/Kr 20/80)

4. Conclusion

The results of this study confirm that advanced MOF materials offer strong potential for the selective adsorption and separation of high-value electronic specialty gases, including SF₆/N₂ and Xe/Kr mixtures. The close correlation between static isotherm data and dynamic

breakthrough results underscores the importance of combining both measurement approaches for accurate performance evaluation.

Using Sync 400 surface area and pore size analyzer and the BTSorb-100 breakthrough system, researchers were able to characterize adsorption behavior with high resolution and repeatability under realistic process conditions. These tools provided critical insights into both capacity and selectivity, enabling a comprehensive understanding of material performance.

As demand continues to grow for high-purity gases in semiconductor, medical, and aerospace industries, precision gas separation technologies will remain a key innovation area. AMI’s application-driven instrumentation supports the development and scale-up of advanced materials by delivering robust, accessible, and high-performance analytical solutions.

5. References

[1] He, H., Liu, Y., Zhang, J., et al. Research Progress in Synthesis and Purification Technologies of Electronic Specialty Gases. Low Temperature and Specialty Gases, 2023, 41(01): 1–5+46.

[2] China's Electronic Specialty Gases: From Import Substitution to Global Supply — In-depth Report on the Electronic Specialty Gas Industry.

[3] Wang, Q., Ke, T., Yang, L., Zhang, Z., Cui, X., Bao, Z., Ren, Q., Yang, Q., Xing, H. Angewandte Chemie, 2020, 132(9): 3451–3456.

[4] Zhao, Z. Research on Enhanced Xe/Kr Selective Adsorption Separation by Regulating Adsorbent Pore Environment. Nanchang University, 2023. DOI: 10.27232/d.cnki.gnchu.2023.003407.