Determination of Nitrogen Content in Oil Using the Gamma-Activation Method

R. P. Meshcheryakov, B. M. Yakovlev, G. I. Tronov

The determination of nitrogen, as well as oxygen, in oil is of definite interest both for enterprises of the oil-refining industry and for isotope geology, on whose analytical results the contribution of these two elements may have a significant influence. In addition, there are grounds to suppose that both of these elements enter into the composition of oils directly from the moment of their formation. In this case, proceeding from the nitrogen and oxygen content, it would be possible to obtain information on the genesis of the oil deposit.

The nitrogen content in oils from various deposits is usually small and amounts to several tenths of a percent. Nitrogen in oil can be determined by standard methods of analytical chemistry, the time required to carry out one analysis being several hours, while the accuracy of the determination does not exceed $\pm 10\%$. We have found no references in the literature to reliable direct methods for determining oxygen in oil, but its content is usually judged from difference data. With the exception of hydrogen and carbon, which constitute the basis, the content of other elements in oil usually does not exceed on the order of $10^{-4} \div 10^{-5}\%$. This circumstance creates favorable conditions for determining the content of nitrogen and oxygen by activation methods, since, with the proper choice of the type of radiation ($\gamma$-quanta, neutrons, charged particles) and the radiation energy, the contribution of interfering activities can be neglected.

Experiments on the determination of nitrogen and oxygen in oil were carried out by us using the gamma-activation method of analysis. As the source of high-energy bremsstrahlung, a high-current betatron of the Research Institute of Nuclear Physics at TPI was used. In this betatron, at a maximum radiation energy of 25 MeV, up to $\sim 10^{12}$ electrons are accelerated per pulse. The accelerator installation is powered from an industrial-frequency current source.

In the experiments carried out by us, the content of nitrogen, as well as oxygen, was analyzed both in the oil itself and in its fractions. Before analysis, samples of crude oil were purified of mechanical impurities and emulsified water by centrifugation. The cassettes into which the oil was placed were made of polyethylene. They had the shape of flat cylinders with dimensions: diameter 50 mm, height 15 mm. To avoid evaporation of light oil fractions, the wall thickness

the cassettes was taken to be 1 mm. After filling with oil, the cassettes were welded shut to prevent leakage.

Irradiation of the samples in the determination of nitrogen was carried out at a bremsstrahlung energy of 18.5 MeV for one half-life. Since the intensity of the accelerator bremsstrahlung may vary with time, the internal-standard method could have given us insufficiently correct results. Therefore, monitoring of the bremsstrahlung dose received by the sample during a specified irradiation time was performed by the induced-activity method; copper disks 0.3 mm thick, whose diameter was equal to the diameter of the cassettes, were used as monitors. During irradiation of the sample, the copper disks were placed in front of and behind the cassette, in direct proximity to it. After the end of irradiation and the counting of the nitrogen activity (by the radioactive isotope nitrogen \(N^{13}\), formed as a result of the \((\gamma, n)\) reaction at bremsstrahlung energies above the threshold for this element) in the oil on a special measuring setup, the total induced activity in both copper disks was measured. Thus, after introducing corrections for the time elapsed after the end of irradiation and for the contribution of long-lived components, the activity of the copper isotope \(Cu^{62}\) determined the radiation dose received by the oil sample. As shown by control experiments and by the calculations we carried out, the use of this type of radiation monitor ensured reproducibility of the results within 2% when the position of the irradiated cassette was varied over wide limits relative to the axis of the accelerator bremsstrahlung beam.

The half-lives of the radioactive isotopes of copper \(Cu^{62}\) and nitrogen \(N^{13}\) practically coincide and are equal to \(\sim 10\) min (this value is a weighted average from the results of measurements performed by many authors) with an accuracy of up to several percent, which practically eliminates the influence of fluctuations in the radiation intensity on the results of the experiments.

Unfortunately, the ratio of activities in the energy range of bremsstrahlung required for the measurements is a nonlinear function of energy, and a special circuit [1] was used to stabilize the latter. The positron activity of the radioactive isotopes formed, \(Cu^{62}\) and \(N^{13}\), was measured using two NaI(Tl) crystals, with pulses selected by amplitude and by means of a coincidence circuit. The statistical error of the measurements did not exceed 2.5%.

The distribution of nitrogen content over the oil fractions is characterized on average by the following data:

a) oil fractions with a distillation temperature up to 100°C contain approximately 0.15% nitrogen;

b) oil fractions with a distillation temperature from 100 to 300°C contain approximately 0.08–0.10% nitrogen (with a gradual increase in nitrogen content);

c) bitumens—0.4–1.0% nitrogen.

Since the cross section of the \((\gamma, n)\) reaction on carbon near the threshold has a discrete character and the yield of radioactive nuclei up to the first break in the curve is small, the maximum bremsstrahlung energy in the determination of oxygen in oil samples was taken to be 19 MeV. The irradiation time of the samples was chosen to be 2 min; the activity of the irradiated samples was recorded during the following 8 min in time intervals of 2 min, with interpretation of the obtained data.

Because of the contribution of interfering activities, the accuracy of determining oxygen in oil samples did not exceed 5–6%.

With a further increase in the energy of the bremsstrahlung radiation, the activity of oil samples is determined predominantly by carbon, which makes it possible to determine the content of this element as well, with an accuracy of up to several relative percent within 10–15 minutes.

REFERENCES

1. R. P. Meshcheryakov, B. M. Yakovlev. Stability of the energy of betatron bremsstrahlung radiation, in: Electronic Accelerators (Proceedings of the Fifth Conference on Electronic Accelerators). Moscow, Gosatomizdat, 1966, p. 177.