Water adsorption dynamics in active carbon probed by terahertz spectroscopy

Abstract

It is vital to characterize the adsorption dynamics in oil–gas reservoirs and pollution control industry. Terahertz (THz) spectroscopy was used to study the adsorption of the water molecules in active carbon. The absorbance at selected frequencies and the first principal component scores over the whole THz range were related to the corresponding time lengths. The collective tendency expressly tracked the dynamics of water adsorbed in active carbon pores. Therefore, THz technique can be used as a promising tool to monitor the adsorption issues in petroleum and environment fields.

---

The adsorption phenomenon plays a significant role in many fields such as oil–gas reservoirs and wastewater treatment.^1,2 Shale gas, which is an unconventional natural gas hidden in the strata or mudstone layers in a free or an adsorbed state, is becoming an important force in the world market. The research regarding oil–gas adsorption in tight reservoirs has important practical significance.^3,4 Some studies are related to the adsorption as well as absorption of organic molecules or ions from water or other solutions along with the interaction between the adsorbed substances and porous materials.^5–12 The discussion regarding the dynamic adsorption process will be a key issue. Active carbon is used to adsorb water molecules in this study to simulate the adsorption dynamics in current oil–gas reservoirs and the environmental pollution industry. Scanning electron microscopy (SEM) and atomic force microscopy (AFM) are appropriate ways to describe the surface and structure of the holes; however, the dynamic process cannot be clearly observed by SEM and AFM. Most importantly, active carbon with liquid water was not appropriate during SEM measurements because a vacuum environment is necessary when measuring, and liquid water is not allowed because it may volatilize and harm the SEM setup. Moreover, the vacuum environment will greatly affect the adsorption dynamics because of the very large water concentration gradient between sample and vacuum environment. In addition, the vacuum, an extreme condition, is not normally found in the actual petroleum industry, thus the experiment is not appropriate to simulate the true adsorption dynamics. Finally, the water molecules cannot be clearly observed because of its nanometer-scope size, which cannot be clearly distinguished using SEM. In addition, although AFM has a very large resolution of atom, its scanning velocity is very small and the time length very long. AFM is also not a suitable online method to monitor the adsorption dynamics because the adsorption length is less than twenty minutes.¹³

Terahertz time-domain spectroscopy (THz-TDS), a technique to bridge the gap between the microwave and the infrared spectroscopy, has been rapidly developed over the last few decades.^14–16 This method can provide ample information on the intermolecular and intramolecular vibration modes and also simultaneously give the amplitude and phase information. The hydrogen bond collective network formed by water molecules changes on a picosecond (ps) timescale, thus causing the THz spectrum to be sensitive to fluctuations in the dipole moment of the water.^17,18 THz technique is an appropriate method for process monitoring because of its online properties and simple measurement conditions when detecting, which are necessary for the rapid identification of adsorption dynamics in the actual industry. Moreover, THz provides an indirect method for the characterization of adsorption dynamics depending on the absorption effect in THz frequency range. In this study, the THz measurements of water drops adhered on active carbon at various time frames have been discussed. The research focuses on the process observation that the water molecules gradually move into active carbon. First, the absorption spectra of the sample were obtained over the range of 0.1–1.45 THz. Second, principal component analysis (PCA) was adopted to build a relationship between the THz adsorption and the timeframes. Finally, the THz absorbance at 0.5, 0.8, 1.0, 1.1, 1.2, 1.3, and 1.4 THz were extracted along with the first principal component scores over the entire range and were associated with the corresponding timeframes. The results showed that the THz technique identified the different stages, especially the adsorption process of water molecules into the pores of active carbon.

Fig. 1 shows the THz field amplitude as a function of time after the transmission of the THz pulse through the sample at different timeframes. The hydrogen bond network, which is a special intermolecular or intramolecular interaction and a type of strong molecular link, was formed with the mode of O–H⋯O in the water molecules. The ceaselessly forming and breaking of hydrogen bonds on the ps timescale, which is connected to the reorientation dynamics of the water molecules, were detected due to the sensitivity of THz-TDS. In this research, an air environment was selected, which is the most common condition. A relatively normal adsorption condition is better to simulate the adsorption phenomenon in actual petroleum and environment industries. After employing the fast Fourier transform to THz-TDS in Fig. 1, the THz frequency-domain spectra (THz-FDS) was obtained. The absorbance (A) spectra of the sample was then calculated using −log(Amp_Sam./Amp_Ref.), where Amp_Sam. and Amp_Ref. are the THz-FDS amplitudes of the sample and reference (air), respectively. The absorbance spectra of the samples only reflect the absorption effect of active carbon with or without water in the THz range. Fig. 2 illustrates the frequency dependent absorbance as the water adhered to active carbon at various timeframes over the range of 0.1–1.45 THz and there is no characteristic peak because liquid was used in this experiment, which is consistent with the statement of liquid water absorption in the THz range in a previous report.¹⁹ The results indicate the changes in the THz optical constants with the increasing adhering time.

Fig. 1 THz-TDS of the sample with different timeframes from 1.0 to 40.0 minutes.

Fig. 2 The frequency dependence of absorbance. The x-axis, y-axis and z-axis indicate the frequency range (0.1–1.45 THz), the adhering time of water drop on the active carbons, and the absorbance, respectively.

The dependence of the adsorption process on the adhering time was evaluated in detail at randomly selected frequencies. The corresponding results are displayed in Fig. 3, which show the absorbance as a function of the adhered time after the transmission of the THz pulses. The absorbance basically remained unchanged in the range of 0–10.5 minutes; then, decreased in the 10.5–29.0 minutes range and remained invariant in the 29.0–40 minutes range. The non-linear dynamics on the basis of collective tendency was brought into correspondence with each other at 0.5, 0.8, 1.0, 1.1, 1.2, 1.3 and 1.4 THz, indicating a special rule, in which a certain response existed for the water molecules that adhered onto and into the active carbon in the THz range.

Fig. 3 The time dependence of absorbance at selected frequencies.

The observed changes in the THz absorbance were very significant during the adhering process. As mentioned above, the whole process can be divided into three stages: 1–10.5, 10.5–29.0 and 29.0–40 minutes, in accordance with the adhered time, as shown in Fig. 3. In the first stage, the water drop adhered onto the superficies of active carbon and diffused to the surrounding area according to the concentration gradient of the water molecules. This diffusion process was divided into two parts: the surface motion and the depth diffusion. The two processes were carried out simultaneously and their competing actions made the THz absorbance change only slightly. At this stage, the quantity and intensity of hydrogen bonds do not change because water molecules are not adsorbed. However, during the second stage, the THz absorbance gradually decreased with the adherence time of the water drop. This stage was homologous with the adsorption process water molecules. The active carbon, which has a very large specific surface area, adsorbed the water molecules into the voids. In the adsorption stage, with the increasing molecules adsorbed into active carbon, the water molecules were scattered in different holes, especially in inner ducts; thus, the intramolecular vibration changed and the THz response was weaker and weaker. Consequently, the THz absorbance spectra can trace the motion as the water molecules were gradually adsorbed into the voids of active carbon. In the third stage, when the adsorption ends, the intramolecular vibrations remained unchanged; thus, the absorbance values of the samples at different timeframes remained unchanged, indicating that the adsorption process ended and the water molecules were adsorbed in a stable state in the pores of the carbon. To highlight the difference between the non-adsorbed and adsorbed samples, a contrast of the amplitude and phase of THz frequency-domain spectra (THz-FDS), which were calculated by fast Fourier transform of THz-TDS, and optical parameters including refractive index (n) and absorbance (A) of the active carbon with water drop adsorbed at 40 minute and without water is illustrated in Fig. 4. Although several peaks are found in Fig. 4(a), they result from the vapor in air, which is selected as a measurement condition. The absorbance spectra in Fig. 4(d) do not have any characteristic peaks because of the calculation of −log(Amp_Sam./Amp_Ref.) and is consistent with the spectra shown in Fig. 2. The significant differences that exist between the samples within the range of 0.1–1.45 THz indicate that the water molecules were adsorbed into active carbon rather than volatilized into the air.

Fig. 4 THz-FDS amplitude (a), phase (b) and optical parameters including refractive index n (c) and absorbance A (d) for active carbon with water at 40 minutes and without water.

In this study, a PCR method was used, which was able to narrow multiple variables to a few principal components (PCs) with dimension reduction technology, to analyze the motion process with the input of the THz absorbance spectra of the samples in the 0.1–1.45 THz range; however, none of the spectral pretreatments were used.²⁰ As shown in Fig. 5, the x-axis and y-axis indicate the first and the second PC scores, of which the contribution rates were 97.5% and 2.0%, respectively. Therefore, the first two PCs, particularly PC1, represent the majority of the sample information. The sample occupies different positions in the coordinate system at different adhering times. In regards to the adhered time, the sample had similar PC1 scores and different PC2 scores at 1.0–10.5 minute, different PC1 and PC2 scores at 10.5–29.0 minute, and the same PC1 and PC2 scores at 29.0–40.0 minute. The adsorption process with different adhered timeframes reflects the different PC scores; consequently, it can be classified as having three stages.

Fig. 5 PC1 versus PC2 for the absorbance data over the range of 0.1–1.45 THz.

Because PC1 presented most of the information on the original variables of the system due to its high contribution rate (97.5%), the PC1 scores were extracted and associated with the corresponding time, as shown in Fig. 6. An analogous trend was obtained and its time intervals of transition points were similar to that shown in Fig. 3. Each time interval was evident. Therefore, based on the regular curves from Fig. 3 and 6, a conclusion was drawn that the THz technique can be used as an effective and promising tool to track the adsorption process of the fluid adsorbed into the porous structure.

Fig. 6 The time dependence of PC1 scores in the PCA system.

To test and verify the repeatability and accuracy of the abovementioned conclusion, another similar experiment was performed where three drops were used and the PCA was also used to calculate the PC scores. Only the PC1 scores were extracted due to its high contribution rate (96.33%) and associated with the corresponding timeframes, as shown in Fig. 7. Although the time intervals of the transition points appeared different, the three stages were obviously distinguished and the second stage was the adsorption process over the time range of 11 to 42 minutes. The longer adsorption time resulted from the increase in the quantity of water molecules. These results showed that the adsorption process was evident and accurately expressed by the THz technique.

Fig. 7 Dependence of PC1 scores on the corresponding time of three-drop-water experiment.

Conclusions

In summary, the adsorption process of water into active carbon was monitored using THz-TDS. The absorbance at any frequency reflected the water adsorption dynamics into active carbon and the adsorption process was then validated by PCA calculation with the absorbance over the entire THz range. Therefore, THz-TDS represents a promising technique to monitor the adsorption dynamics; thus, THz technique might be greatly popularized in petroleum and environment industry in the future.

Acknowledgements

This work is supported by the National Key Basic Research Program of China (grant no. 2014CB744302), the Specially Funded Program on National Key Scientific Instruments and Equipment Development (grant no. 2012YQ140005), NSFC (grant no. 61405259), and the Beijing National Science Foundation (grant no. 4122064).

Notes and references

1. J. J. Shen and K. Chen, Journal of Unconventional Oil and Gas Resources, 2014, 5, 1.
2. F. Zietzschmann, J. Altmann, A. S. Ruhl, U. Dünnbier, I. Dommisch, A. Sperlich, F. Meinel and M. Jekel, Water Res., 2014, 56, 48.
3. F. Grancesco and P. Zeniewski, Energy, 2013, 57, 44.
4. A. Vengosh, N. Warner, R. Jackson and T. Darrah, Procedia Earth Planet. Sci., 2013, 7, 863.
5. C. Moreno-Castilla, Carbon, 2004, 42, 83.
6. C. Namasivayam and D. Kavitha, Dyes Pigm., 2002, 54, 47.
7. M. I. Bautista-Toledo, J. Rivera-Utrilla, R. Ocampo-Pérez, F. Carrasco Marín and M. Sánchez-Polo, Carbon, 2014, 73, 338.
8. W. T. Tsai, C. Y. Chang, M. C. Lin, S. F. Chien, H. F. Sun and M. F. Hsieh, Chemosphere, 2001, 45, 51.
9. S. Senthilkumaar, P. R. Varadarajan, K. Porkodi and C. V. Subbhuraam, J. Colloid Interface Sci., 2005, 284, 78.
10. A. J. Groszek, Langmuir, 1999, 15, 5956.
11. Y. Ozcelik and A. Ozguven, Construct. Build. Mater., 2014, 63, 257.
12. J. Zhang and D. Grischkowsky, Opt. Lett., 2004, 29, 1031.
13. M. Sujka and J. Jamroz, LWT–Food Sci. Technol., 2009, 42, 1219.
14. N. Horiuchi and X. C. Zhang, Nat. Photonics, 2010, 4, 662.
15. R. M. Bao, S. X. Wu, K. Zhao, L. J. Zheng and C. H. Xu, Sci. China: Phys., Mech. Astron., 2013, 56, 1603.
16. I. Lundholm, W. Y. Wahlgren, F. Piccirilli, P. D. Pietro, A. Duelli, O. Berntsson, S. Lupi, A. Perucchi and G. Katona, RSC Adv., 2014, 4, 25502.
17. E. Castro-Camus, M. Palomar and A. A. Covarrubias, Sci. Rep., 2013, 3 DOI:10.1038/srep02910.
18. D. H. Choi, H. Son, S. Jung, J. Park, W. Y. Park, O. S. Kwon and G. S. Park, J. Chem. Phys., 2012, 137, 175101.
19. J. Xu, K. W. Plaxco and S. J. Allen, J. Chem. Phys., 2006, 124, 036101.
20. A. D. Burnett, W. Fan, P. C. Upadhya, J. E. Cunningham, M. D. Hargreaves, T. Munshi, H. G. M. Edwards, E. H. Linfield and A. G. Davies, Analyst, 2009, 134, 1658.

---

Footnotes

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c4ra14730h

‡ These authors contributed equally.

---