The TGA-MS Combination

Introduction

The combination of a thermobalance with a mass spectrometer (MS) measures the change in mass of the sample under investigation and at the same time yields qualitative information on the evolved decomposition or vaporization products. The quality of the mass spectra depends on a number of different experimental parameters. The most important of these are the sample mass, the gas flow, the accelerating voltage of the secondary electron multiplier (SEM) and the position of the tip of the connecting capillary in the furnace of the thermobalance. The following article discusses how these four parameters influence the mass spectrometric measurement results.

The experiments were performed with a TGA/SDTA851e (large furnace) coupled to an Inficon Thermostar QMS300 mass spectrometer (mass range 1–300). The decomposition of calcium oxalate monohydrate (CaC₂O₄·H₂O) was chosen as the reaction model. CaC₂O₄·H₂O loses its water of crystallization above about 120 °C. The anhydrous oxalate then decomposes in two separate reaction steps as follows:

CaC₂O₄ ⇒ CaCO₃ + CO (I)

CaCO₃ ⇔ CaO + CO₂ (II)

Associated with the above dehydration and two decomposition reactions are stoichiometric mass losses of 12.3%, 19.2% and 30.1% respectively. One mole each of H₂O, CO and CO₂ is evolved per mole of CaC₂O₄·H₂O in the different reaction steps.

Figure 1 summarizes the results of the TG-MS measurements on calcium oxalate monohydrate. The mass spectrometer detected the loss of H₂O, CO and CO₂ corresponding to the three separate mass loss steps. Furthermore, the SDTA signal shows that the three reaction steps are endothermic. 

Influence of Sample Mass

The area of the fragment ion peak occurring during a mass loss step is directly proportional to the sample mass or the height of the mass loss step. This is illustrated in Figure 2 using the CO₂ peak as an example. The diagram shows the MS signal for the m/z 44 ion (CO₂) as a function of time for different sample masses. The samples were heated at 15 K/min with a purge gas flow rate of 50 mL/min of argon. The electron multiplier voltage (SEM) was set to 900 Volt. The peak area plotted against sample mass is to a good approximation a straight line. The slope of the curve corresponds to the sensitivity of the MS for CO₂. Under the conditions used (argon 50 mL/min, SEM voltage 900 V), this gives a value of 1600 nCb/(mg CO₂). The MS signal for m/z 44 is thus calibrated (for these conditions). This allows quantitative information on CO₂ to be obtained for an unknown sample from the m/z 44 signal (assuming of course that this signal is in fact due only to CO₂). This possibility is of particular interest if different gases are evolved at the same time during a mass loss step, or if the mass loss step is so small that it cannot be properly evaluated with the thermobalance but MS peaks corresponding to it are clearly measurable.

Influence of the SEM Accelerating Voltage on Mass Spectrometer Sensitivity

In the Channeltron detector of the MS, the ions selected by the mass filter produce electrons that are multiplied in the SEM (through the formation of secondary electrons). The probability of detection for a single ion is thereby greatly increased compared with a Faraday cup detector. The amplification that can be achieved with the Channeltron detector depends on the SEM accelerating voltage, which can be set with the mass spectrometer control software. 

Conclusions

TGA-MS measurements open up a wide range of new application possibilities both for mass spectrometry and for TGA compared with when the techniques are used alone. To obtain the best results, the effects of a number of different experimental parameters have to be taken into account.

The mass spectrometer signal intensity depends on the sample mass, the accelerating voltage of the secondary electron multiplier (SEM), the gas flow in the furnace of the thermobalance and the position of the capillary tip in the furnace. The best sensitivity is obtained with low gas flow rates (to protect the balance from decomposition products a flow of at least 20 mL/min protective gas should be used) and large sample masses. In addition, the SEM voltage of the MS should be optimized.

This, however, requires care because the Channeltron detector can be physically damaged at high ion currents. In comparison, the position of the capillary tip in the furnace has only a relatively small effect on the MS results. 

_{The TGA-MS Combination | Thermal Analysis Application No. UC 275 | Application published in METTLER TOLEDO Thermal Analysis UserCom 27}