Home Technical note

T1 FA-MS: Flowing afterglow mass spectrometry

The chemical element hydrogen, H, is a mixture of two stable isotopes: hydrogen, 1H, and deuterium, 2H or D. Both of these isotopes are naturally present in water; in one million terrestrial water molecules, there is typically 999,700 1H2O molecules and 300 HDO (or 1H2HO) molecules. The probability of finding a single D2O molecule in this water is very low. The deuterium abundance is conventionally determined in liquid water, urine, saliva and blood samples by on-line reduction or off-line equilibration to hydrogen, followed by Isotope Ratio Mass Spectronetry (IRMS).

FA-MS is a new analytical approach that offers a route to on-line, real-time deuterium abundance measurements in water vapour in breath and above aqueous liquids, including urine and serum. This method involves the production and flow of thermalised hydrated hydronium cluster ions in inert helium or argon carrier gas along a flow tube following the introduction of a humid air sample. These ions react in multiple collisions with water molecules, their isotopic compositions reach equilibrium and the relative magnitudes of their isotopomers are measured by a quadrupole mass spectrometer located downstream.

FA-MS Instrumentation

In an FA-MS instrument, a weak microwave discharge is created in helium or argon carrier gas flowing through a narrow glass tube connected to a stainless steel flow tube. This forms a "flowing afterglow plasma" in the steel flow tube. The gas phase ion chemistry initiated by He+ or Ar+ ions reacting with trace amounts of H2O molecules results in the formation of H3O+ ions in the carrier gas. (Note that here H tacitly assumes the presence of both isotopes 1H and 2H). A sample of air/water vapour mixture to be analyzed is introduced at a known flow rate into the carrier gas and its composite water molecules react with the H3O+ ions to form the H3O+(H2O)0,1,2,3 cluster ions and their analogous 2H, 17O and 18O isotopic variants. The mixture of ions are sampled from the flowing swarm via a pinhole orifice located at the downstream end of the flow tube and they are mass analyzed by a differentially pumped quadrupole mass spectrometer (pressure less than 10-4 Torr) with a single channel multiplier ion counting detector.

A typical mass spectrum will show clusters of peaks at m/z 19,20,21; 37,38,39; 55,56,57; and 73,74,75. The deuterium content of a water vapour sample introduced into the helium carrier gas can be determined from such spectra if the 17O and 18O content of the ions is known. To properly understand how the analysis is achieved, we need to distinguish between the isotopic composition of the following three “phases”: the liquid water sample (designated by the subscript “liq”), the water vapour transferred from an aqueous sample headspace into the helium carrier gas (designated by the subscript “vap”) and the H3O+(H2O)0,1,2,3 ions and their isotopomers that comprise the ion swarm created in the carrier gas (designated by the subscript “ion”).

Partition of HDO Between Liquid Water and Its Vapour

In water containing a low abundance of deuterium almost all the deuterium is contained in 1HDO molecules. Therefore, in order to determine the deuterium isotope abundance ratio in a liquid water sample, R2liq =D/(1H + D), by analysing its vapour, the partition of 1HDO between the liquid and vapour phases needs to be addressed. A difference arises because 1HDO has a lower saturated vapour pressure than 1H2O at sub-boiling temperatures. The ratio K2 = R2vap/R2liq of the deuterium abundance in the headspace vapour, R2vap, to that in liquid, R2liq, has been accurately determined for a range of temperatures. Also, the isotope abundance ratios of 17O in the water vapour, R17vap = 17O/(16O+17O+18O), and that of 18O, R18vap =18 O/(16O+17O+18O), are proportional to their corresponding abundances in the liquid, R17liq and R18liq, and their partition coefficients K17 = R17vap/R17liq and K18 = R18vap/R18liq that are accurately known.

Ion Molecule Reactions; Equilibrium Between Ions and Water Vapour Molecules in the Gas Phase

The deuterium abundance is determined from an understanding of the ion chemistry that establishes the distributions between the isotopomers of the cluster ions. The following three-body association reactions occur to form these cluster ions:

1,2H3O+ + 1,2H2O + M -->1,2H5O2+ + M (1a)
1,2H5O2+ + 1,2H2O + M --> 1,2H7O3+ + M (1b)
1,2H7O3+ + 1,2H2O + M --> 1,2H9O4+ + M 1(c)

Here, M is a carrier gas atom, either He or Ar. As result of this sequence, 1,2H3O+(1,2H2O)3 ions become the dominant ionic species in the carrier gas at sufficiently large water molecule concentrations, since further association to produce H11O5+ ions is inhibited due to the lower stability of this larger cluster ion. Thus, 1,2H3O+(1,2H2O)3 cluster ions reach isotopic equilibrium with water vapour molecules. The following forward and backward reactions establish the equilibrium:

1H9O4+ + 1HDO <--> 1H8DO4+ + 1H2O (2)

These reactions proceed via isotope exchange and ligand switching reactions, involving efficient mixing of H and D atoms within the intermediate reaction complex ion (H10DO5+ )* The enthalpy change, delta H, in reaction (2) is close to zero, the translational and rotational entropy changes are also relatively small and thus the entropy change, delta S, is entirely described by statistical factors. Therefore, when equilibrium is established in reactions (2), the deuterium abundance ratio in the H3O+(H2O)3 cluster ion swarm, R2ion, will be equal to that in the water vapour, R2vap. Note that isotope exchange and ligand switching reactions analogous to reaction (2) also ensure that the abundances of the 17O and 18O isotopes is the same in the swarm of ions as it is in the sample water vapour molecules. This chemistry has been experimentally validated under both SIFT-MS and FA-MS conditions using standard deuterium enriched water. The speed of approach to equilibrium in these reactive systems depends on the number density of water molecules in the carrier gas. Thus, for FA-MS deuterium analyses the number density of water molecules in the carrier gas should be more than 1013 cm-3, in order to establish equilibrium in about a millisecond.

Ion Signals and Isotopomer Overlap

The final important issue to be addressed is the relationship of R2vap to the observed ion signals. The equilibrium distribution of the various isotopomer ions (and hence their count rates at the detector), can be expressed using the binomial distribution. Thus, it is seen that the ratio of ion signals counted at mass to charge ratios, m/z, 74 and 73, i.e. I(74) and I(73) is:

I(74) [H8D16O4+][H917O16O3+]9R2vap4R17vap

I(73)[H916O4+][H916O4+]1 - R2vap1 - R17vap

Note that the deuterium abundance in these cluster ions is amplified by a factor of about 9 relative to water vapour molecules and that the abundance of 17O is amplified about four times.

Very large count rates at m/z of 73, I(73) can be seen, typically several millions per second. Such large count rates cannot be counted sufficiently accurately by a conventional counting system. Instead, I(74) and I(75) are used to determine the deuterium content of the water vapour sample (e.g. in breath). The ion at m/z =75 (18O isotopomer of the H9O4+ ) will have a signal level given by:

I(75) [H916O318O+] 4R18vap

I(73) [H916O4+] 1 - R18vap

Note that it has been shown that when R2 < 10-3 the contribution of doubly deuterated H7D2O4+ ions to I(75) is negligible (less than 3×10-5 representing less than 0.15 % of a typical R18 value of 0.02). Thus, from an accurate measurement of the ion count rate ratio, Q = I(74)/I(75), R2vap, is determined by combining Eqs. (3) and (4) and considering all R2vap, R17vap, R18vap are much less than unity. Then:

R2vap = 4/9 (QR18vap - R17vap) (5)

The Q value obtained for normal water is typically 0.35. Thus, the actual measurements involve ion count rates at m/z of 74 and 75 that are not very different (in this case by a factor of 3), which is inherently more accurate than attempting to measure count rates that are vastly different. When the signals of both m/z 74 and 75 ions are within the range 10,000 to 30,000 counts per second, optimum accuracy and precision can be achieved. In practice, it is also important to ensure that the abundance sensitivity of the analytical mass spectrometer (separation between masses) is better than 10-5 and mass discrimination between the m/z 74 and 75 ions is less than 0.1%

In the commercially available instrumentation (Instrument Science, UK) all numerical analysis using equations (5) and the appropriate K2 partition coefficient are performed on-line by data acquisition software, thus providing an instantaneous readout of the Q, R2vap and R2liq values.

Accuracy and Precision

The accuracy of the measured R2vap also depends on the accuracy of the adopted values of R17vap and R18vap. Unless the accurate local water values of the last two parameters are available, the generic natural abundances of R17liq =0.000379 and R18liq =0.002006 can be used. When breath is analysed for deuterium, values of K2 corresponding to the alveolar interface temperatures must be used, which can range from 34 to 37 ºC. This spread in temperature results in a variation of K2 that is less than 0.3%.

The measurement precision is dominated by the Poisson distribution of the total numbers of ions counted within the sampling time interval. The standard error in R2 calculated using equation (5) is thus:

delta R1 = 4/9 R18(N(74)+N(75)) (6)

In this case, N(74) and N(75) represent the total numbers of ions counted at m/z = 74 and 75 not the ion count rates as used previously. Equation (6) is evaluated on-line providing an immediate estimate of the measurement precision. Validation of the SIFT-MS and FA-MS methods for deuterium analyses, carried out using standard mixtures, demonstrates that both accuracy and precision (reproducibility) are typically 1 % for headspace sampling when using the experimental procedure described above. A precision better than 1 % can be achieved for breath deuterium analysis when the average value for three consecutive exhalations is taken.

Direct Sampling of Breath

Air or breath is sampled directly into the flow tube via a heated, calibrated capillary tube coupled to the flow tube via heated stainless steel tubing. Heating these lines minimises the condensation of water, which could otherwise lead to “memory” effects.

The tip of the capillary extends into a stainless steel coupling positioned perpendicularly to the capillary axis. Exhaled breath is introduced into the coupling via a standard disposable cardboard mouthpiece (about 15 mm diameter). This arrangement is patient friendly, offering a suitable resistance to the flow of breath such that a steady exhalation can be made over a few seconds. The exhaled breath totally displaces the ambient air from the entrance to the sampling capillary and so a sample of breath enters the capillary and immediately expands into the coupling tubing and enters the flow tube/carrier gas (pressure of about 1 Torr). The entrance to the capillary is again exposed to the ambient air upon oral inhalation.

By suitable choice of the dimensions of the calibrated capillary, the flow rate of the sampled air/breath is sufficient to establish a concentration of H2O molecules in the carrier gas that converts the majority of the ions in the ion swarm into the H3O+(H2O)3, i.e. H9O4+ ions, at m/z 73 and its isotopomers at 74 and 75. The carrier gas is flowing rapidly, the flow time through the flow tube being typically 3 milliseconds, and so the net response time of a FA-MS instrument is about 20 ms. A typical breath exhalation is about 5 s, so time profiles of the individual ion signals can be defined.

Sampling Liquid Headspace

FA-MS can be used very effectively to accurately determine deuterium (HDO) in the headspace above water and aqueous media such as urine, dialysate or serum. In a typical analysis, about 10 mL of fluid is placed in 200 mL glass bottle sealed with a septum-sealed. The bottle is then placed in a temperature controlled water bath and the headspace is allowed to develop for about 10 min. A sampling needle connected directly to the input line of the FA-MS instrument (Fig. 1) then punctures the septum and headspace vapour is drawn into the carrier gas by the existing pressure differential (initially from atmosphere down to the flow tube pressure). Again, the sampling lines are held at about 100 ºC to inhibit the condensation of water vapour and other condensable vapours.


Spanel P, Smith D. Flowing afterglow mass spectrometry (FA-MS) for the determination of the deuterium abundance in breath water vapour and aqueous liquid headspace. In: Amann A, Smith D, eds. Breath Analysis for Clinical Diagnosis and Therapeutic Monitoring. Singapore: World Scientific, 2005. Home