- Email: [email protected]
Thermal release of nicotine and its salts adsorbed on silica gel
Accepted Manuscript Title: Thermal release of nicotine and its salts adsorbed on silica gel Authors: Qing Hua, Wenjie Lu, Saijing Zheng, Yichun Zhang, Wei Zhang, Da Wu, Yi Shen PII: DOI: Reference:
S0040-6031(17)30211-3 http://dx.doi.org/10.1016/j.tca.2017.08.013 TCA 77814
To appear in:
Thermochimica Acta
Received date: Revised date: Accepted date:
3-6-2017 2-8-2017 27-8-2017
Please cite this article as: Qing Hua, Wenjie Lu, Saijing Zheng, Yichun Zhang, Wei Zhang, Da Wu, Yi Shen, Thermal release of nicotine and its salts adsorbed on silica gel, Thermochimica Actahttp://dx.doi.org/10.1016/j.tca.2017.08.013 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Thermal release of nicotine and its salts adsorbed on silica gel Qing Hua, Wenjie Lu, Saijing Zheng, Yichun Zhang, Wei Zhang, Da Wu, Yi Shen*
Shanghai Tobacco Group Co., LTD, Shanghai 200082, China. *Corresponding author: Yi Shen Tel: +86 21 61668311; Fax: +86 21 61668366. E-mail address: [email protected] Graphical Abstract
1
Highlights
Nicotine and its salts adsorbed on silica gel are employed to simulate the nicotine release behaviors of RTPs.
TG and DRIFT are used to determine the interaction between amine groups of nicotine and silica gel.
Three adsorption modes of nicotine/its salts on the surface of silica gel have been established.
The pyrolysis temperatures of nicotine salts adsorbed on silica gel (and contained in RTPs) are determined.
ABSTRACT
Nicotine and its salts adsorbed on silica gel are employed to simulate the nicotine release behaviors of reconstituted tobacco particles (RTPs). Thermogravimetric analysis (TGA) was applied to provide an overview of the thermal behaviors of the nicotine, its salts and the RTPs. Diffuse reflectance infrared spectroscopy (DRIFTS) measurements of pyrolysis processes were performed on a FTIR spectrometer equipped with an in-situ DRIFTS reaction cell and a MCT/A detector. Firstly, the amine groups of nicotine and its salts are mainly adsorbed on the isolated hydroxyl groups of silica gel. Secondly, three adsorption modes of nicotine have been established. Finally, smoking machine was applied to study the release behavior of nicotine from the reconstituted tobacco particles (RTPs). The interaction between nicotine and the adsorbent is identified. Our results show interesting observations on 2
thermal release of nicotine from the different substrates, proving the adsorption and thermal release of nicotine and it salts from RTPs, which have a guiding significance for the development of heat-not-burn cigarettes.
KEYWORDS: Nicotine salt; DRIFT; TGA; Pyrolysis; Interaction
1. Introduction
Nicotine, an addictive component of tobacco smoke, is of interest in its evolution during the smoking process [1-4]. In tobacco leaves, nicotine is mainly in the form of nicotine salts [5, 6], most of which are nicotine malate and nicotine citrate. As we all known, cigarette smoke of conventional cigarettes is a highly complex aerosol system, involving over 6,000 identified chemicals in a dynamic and reactive mixture [6, 7]. These chemicals are generated by incomplete combustion of tobacco [8]. Tar generated in the burning process contains most toxicants from the tobacco smoke. Nowadays, an effective method to reduce the generation of toxicants from tobacco is to develop heat-not-burn cigarettes [9-11], which generates less tar than conventional cigarettes in the smoking process. There are two forms of heat-not-burn cigarettes. One form of heat-not-burn cigarettes looks like a cigarette. It involves a lit carbon tip that heats incoming air, which in turn heats tobacco-based substrates, forming an
3
aerosol containing mainly water, glycerol, nicotine and volatile tobacco components [12-14]. Another type of heat-not-burn cigarettes has a heating device that is powered by a battery, which works with a specially designed tobacco rod [15]. The heating of the tobacco includes heating from the inside of the tobacco rod, or from the external surface of the tobacco rod, or a combination of both. In both of these two forms of heat-not-burn cigarettes, the tobacco heating temperatures are typically below 300°C [15]. The temperature is high enough for tobacco to release nicotine, but tar generates still much less than conventional cigarettes. The related studies have indicated that the aerosol composition in heat-not-burn cigarette smoke contains less toxicant than that in conventional cigarette smoke [9-11]. The mechanism of nicotine release during smoking of conventional cigarettes is still not fully elucidated. Although the aerosol composition in heat-not-burn cigarette smoke is simpler than that in conventional cigarette smoke, it is difficult to study nicotine release behavior of heat-not-burn cigarettes directly. In order to simplify the experimental sample, silica gel is selected as the adsorbent in our research for the following reasons. Firstly, silica gel has great adsorption capability due to its big Brunner−Emmet−Teller (BET) area, which can adsorb enough nicotine/nicotine salts to show clear pictures of TGA and DRIFTS spectra [16]. Another reason for silica gel selected as the adsorbent in our research is that there are numerous hydroxyl groups present on the surface of silica gel. And the surface of tobacco substrate is polyhydroxy [17-21]. Therefore, silica gel with nicotine adsorbed on and its salts may truly reflect the release behaviors of nicotine from the tobacco substrate. 4
In this work, thermogravimetric analyses (TGA) was used to provide the thermal behaviors of nicotine and its salts. And diffuse reflectance infrared spectroscopy (DRIFTS) is applied to study the interaction between probes and adsorbent [22-24]. We employed TGA and DRIFTS to study the release behaviors of nicotine and its salts adsorbed on silica gel at selected temperature. TGA draws a continuous mass loss process, and first derivative of the averaged weight loss (DTG) indicates different pyrolysis stages. DRIFTS spectroscopy is capable of carrying out in situ characterization of the interaction between nicotine (its salts) and adsorbent. Combining the TGA and DRIFTS results, desorption/pyrolysis mechanisms of nicotine/nicotine salts can be investigated. Finally, the nicotine release behavior of reconstituted tobacco particles (RTPs) is examined, which completely fulfills our TGA and DRIFTS results. In a word, our results show interesting observations on thermal release of nicotine from the different substrates, proving the adsorption and thermal release of nicotine and it salts from RTPs, which have a guiding significance for the development of heat-not-burn cigarettes.
2. Material and methods 2.1 Material The nicotine salts were purchased from Huabao International Holdings Limited. The nicotine salts were dissolved in water and the details were in the table S1 (Supplementary data). The silica gel was purchased from Chemical Division of Silicycle. The reconstituted tobacco particles (noted as RTPs) was purchased from 5
Shanghai Tobacco Group Co., LTD. The RTPs were prepared by 100% reconstituted tobacco sheet by a grinder, and the size of those was between 1-2mm. The total content of nicotine in RTPs was 0.91%, and the moisture content was 16%. The RTPs and all chemicals were used as received.
Nicotine/silica gel samples were denoted followed the typical procedures: 40 μL nicotine, 100 μL anhydrous ethanol and 200 mg silica gel were co-added into a gray GC bottle and adequately mixed by stirring. Finally, the samples were dried in vacuum at room temperature (RT) for 2 hours. Nicotine salts/silica gel samples were denoted followed the typical procedures: 80 μL nicotine salts aqueous solution, 100 μL anhydrous ethanol and 200 mg silica gel were co-added into a gray GC bottle and adequately mixed by stirring. Finally, the samples were dried in vacuum at RT for 2 hours. 2.2 Thermogravimetric analyses (TGA) TGA (PerkinElmer, STA 8000) was performed on the nicotine, its salts and the reconstituted tobacco particles to provide an overview of their thermal behavior. For this experiment, 20~25 mg sample was loaded into a Pt crucible. For each TGA experiment, the temperature ranged from 35°C to 500°C in N2 (sample purge: 20 mL/min; balance purge: 40 mL/min) under 20 °C/min heating rate, and held for 5 min at 500°C. First derivative of the averaged weight loss (DTG) was also obtained. 2.3 Diffuse reflectance infrared spectroscopy (DRIFTS) DRIFTS measurements of pyrolysis processes were performed on a Nicolet 5700 FTIR spectrometer equipped with an in situ DRIFTS reaction cell (Harrick Scientific 6
Products, INC) and a MCT/A detector. He (flow rate: 20 ml/min) was employed as the carrying gas for the pyrolysis. Silica gel was purged in the He stream (flow rate: 20 ml/min) at 30 °C for 1 h until there were no changes of the DRIFTS spectrum. And the corresponding DRIFTS spectrum was taken as the background spectrum. Nicotine salts adsorbed on silica were pretreated in the He stream (flow rate: 20 ml/min) for 0.5 h to remove the adsorbed water on the surface of silica gel. The pyrolysis process was in situ studied by the DRIFTS measurements performed with 64 scans and a resolution of 4 cm-1. The corresponding DRIFTS spectrum was taken when the sample was heated in the He stream (flow rate: 20 ml/min) to desirable temperatures under 10 °C/min heating rate and held at the desirable temperature for 3 min. 2.4 Nicotine release study of reconstituted tobacco particles (RTPs) The reconstituted tobacco particles (RTPs) and Cambridge filter pads (CFPs, Borgwaldt, Germany) used for our study were conditioned at 22±1 °C and a humidity of 60±2% for at least 48 h. 65 mg RTPs were filled into a U-sharp quartz tube. The heating equipment was a GC oven (Agilent 6890). The RTPs were puffed in a puff volume of 35 mL with 2 s/puff duration in every 15 s by using a 20-port Borgwaldt RM200 smoking machine (Borgwaldt, Germany) according to ISO 4387 at selected temperatures. The smoke particulate matter of the RTPs in one tube was collected in one CFP with a diameter of 44 mm in totally 20 puffs. The schematic diagram was shown in Fig. S1 (Supplementary data).
7
3. Results and discussion In this work, silica gel was employed as the adsorbent to adsorb nicotine and its salts to study their pyrolysis behaviors. Silica gel is a common adsorbent, which can adsorb enough nicotine/nicotine salts to show clear TGA and DRIFTS spectra [16]. Firstly, we studied the thermal stability of silica gel (as shown in Fig. 1A). The TG curve of silica gel indicates that the weight loss of silica gel is about 2.5% when the temperature is raised up to 150℃. When the temperature is raised up to 500℃, the weight loss of silica gel is about 3.5%. In order to further study the mechanism of the weight loss of silica gel, a series of DRIFTS spectra in situ for the thermal stability of silica gel in He acquired at selected temperatures are shown in Fig. 1B. Compared with the DTG curve (Fig. 1A), it is clearly shown that the water adsorbed on the surface of silica gel firstly desorbed before 150℃. The IR band of silica gel at 3740 cm−1 is slowly enhanced, proving the increase in the number of isolated surface hydroxyl groups on the surface of silica gel [21, 25]. Silica gel has a considerable adsorption capacity due to its large surface area (~ 500 m2/g) [16]. Therefore, silica gel is selected as an excellent adsorbent to study the pyrolysis of nicotine and its salts. According to Fig. 2A, the mass losses of nicotine/silica gel (noted as Nic/SG), nicotine malate/silica gel (noted as Nic-mal/SG) and nicotine citrate/silica gel (noted as Nic-cit/SG) were about 15%. The mass loss under 100°C was contributed by adsorbed water. A detailed pyrolysis process of nicotine and its salts adsorbed on silica gel has been studied by DTG (Fig. 2B). The nicotine undergoes a wide mass-loss process from 100°C to 332°C and reaches the 8
fastest desorption rate at 225°C [26]. In this process, the mass-loss is attributed to desorption of nicotine from the adsorbent (silica gel). The mass-loss processes of nicotine malate and nicotine citrate were divided into two parts. First part is the pyrolysis of nicotine salts, and the second part is desorption of the rest nicotine adsorbed on the adsorbent. The second parts of DTG curves of nicotine malate and nicotine citrate were analogous. Both of these two parts are the desorption processes of the pyrolyzed nicotine adsorbed on the silica gel. The first parts of DTG curves of nicotine malate and nicotine citrate were different because of the difference of the stabilities of nicotine malate and nicotine citrate adsorbed on silica gel. The difference is due to that the stabilities of nicotine malate and nicotine citrate adsorbed on silica gel are not the same. The nicotine malate adsorbed on silica gel begins to pyrolyze from 153°C and reaches the fastest pyrolysis rate at 203°C. In contrast, the nicotine citrate adsorbed on silica gel begins to pyrolyze from 128°C and reaches the fastest pyrolysis rate at 183°C. The first parts of DTG curves of nicotine malate and nicotine citrate indicate that nicotine malate is more stable than nicotine citrate when they are adsorbed on silica gel. Similarly, in order to further study the mechanism of the pyrolysis processes of nicotine and its salts adsorbed on silica gel [27]. A DRIFTS study of nicotine, nicotine malate and nicotine citrate was carried out. Fig. 3 is the DRIFTS spectra in situ for the desorption process of nicotine adsorbed on silica gel in He acquired at selected temperature. As the temperature is raised, the negative peak at 3738 cm-1 disappears slowly. The broad peak at 2300 cm-1 reduces slowly, and the peaks between 3200 cm-1 9
to 2500 cm-1 reflect the same trend with it. The negative peak of 3738 cm-1 belongs to isolated surface hydroxyl groups on the surface of silica gel [18,19]. Compared with the standard IR spectrum of nicotine (Fig. S2 in Supplementary data), the peaks between 3200 cm-1 to 2500 cm-1 come from nicotine. The broad peak of 2300 cm-1 comes from the interactions between the hydroxyl groups of silica gel and the amine groups of nicotine [21,24,25]. Since the background of the DRIFTS spectra is dry silica gel, it has strong IR band at 3738 cm-1. The DRIFTS spectra of silica gel with different purged time are shown in Fig. S3 (Supplementary data). Therefore, there are few isolated surface hydroxyl groups on the sample of Nic/SG. When the adsorbed nicotine is being desorbed with the temperature being raised, the amount of isolated hydroxyl groups on the surface of silica gel gradually is increasing. This indicates that nicotine occupies the isolated hydroxyl groups on the surface of silica gel. From Fig. 3, we can see that nicotine adsorbed on silica gel begins to desorb at 120℃. When the temperature reaches 270℃, almost all nicotine adsorbed on the surface of silica gel has desorbed. The desorption process obtained from DRIFTS is in agreement with that from TG (Fig. 2). Interestingly, a negative IR band at 3250 cm-1 appears when most of the adsorbed nicotine desorbs at 240℃. The IR band at 3250 cm-1 belongs to molecular water [20]. This part of moisture on the surface of silica gel desorbs until most of adsorbed nicotine desorbed. The result confirms that this part of moisture plays a supporting role to enhance the interaction between nicotine and silica gel. Fig. 4A is the DRIFTS spectra in situ for the pyrolysis process of nicotine malate adsorbed on silica gel in He acquired at selected temperatures. The 10
negative IR band at 3738 cm-1 disappears slowly with the temperature being raised. The broad IR bands at 2235 cm-1 and 2555 cm-1 reduce slowly, and the peaks between 3200 cm-1 to 2500 cm-1 show the same trend with them. The IR bands at 1728 cm-1 and 1578 cm-1 have the same decreasing trend until the temperature reaches 270℃. When the temperature is raised up to 300℃, the IR band at 1728 cm-1 is still noticed but with burrs. The negative peak at 3738 cm-1 belongs to isolated surface hydroxyl groups on the surface of silica gel. The peaks between 3200 cm-1 to 2500 cm-1 come from nicotine. The broad peak of 2235 cm-1 comes from the interactions between the hydroxyl groups of silica gel and the amine groups of nicotine. And the broad peak of 2555 cm -1 comes from the interactions between the hydroxyl groups of silica gel and the carbonyl groups of malic acid. According to the standard IR spectra of nicotine and malic acid (Fig. S2 in Supplementary data), the IR band at 1728 cm-1 is attributable to the carbonyl groups of malic acid, and the IR band at 1578 cm -1 is attributable to the amine groups of nicotine. From Fig. 4A, we can see that nicotine malate adsorbed on silica gel begins to pyrolize at 150℃. When the temperature reaches 270℃, almost all nicotine malate adsorbed on the surface of silica gel has pyrolized. The pyrolysis process obtained from DRIFTS solidly agrees with that from TG (Fig. 2). Interestingly, a negative IR band at 3250 cm-1 appears when most of the adsorbed nicotine malate pyrolized at 240℃. The IR band at 3250 cm-1 belongs to molecular water. This part of moisture on the surface of silica gel desorbs until most of adsorbed nicotine malate pyrolized. The result confirms that this part of 11
moisture plays a supporting role in enhancing the interaction between nicotine malate and silica gel. The pyrolysis process of nicotine citrate adsorbed on silica gel is same as that of nicotine malate adsorbed on silica gel (Fig. 4B). However, the IR band of carbonyl group at 1728 cm-1 is different between nicotine malate/SG and nicotine citrate/SG. The IR bands of carbonyl group at 1728 cm-1 are different for nicotine malate/SG and nicotine citrate/SG. When the temperature is raised up to 300℃, the IR band at 1728 cm-1 of nicotine malate/SG is still noticed but with burrs. At this temperature, the IR band at 1728 cm-1 of nicotine citrate/SG is very small. These results indicate that the IR band at 1728 cm-1 with burrs is formed via the carbonization of nicotine malate. According to TG, nicotine malate is more stable than nicotine citrate when they are adsorbed on silica gel. In another word, the interaction between nicotine malate and silica gel is stronger than that between nicotine citrate and silica gel. Nicotine citrate pyrolizes and desorbs from the silica gel at 300℃, while nicotine malate will pyrolize but not desorb from the silica gel in time. Therefore, a small part of nicotine malate will be carbonized, and the IR band at 1728 cm-1 forms. By further studying the pyrolysis DRIFTS results of nicotine and its salts adsorbed on the surface of silica gel, two things should be mentioned: 1) The amine groups of nicotine and its salts interact with isolated surface hydroxyl groups of silica gel; 2) When most of the adsorbed nicotine/nicotine salts get pyrolized and desorbed, there would be molecular water desorbed from silica gel. Based on the above, the adsorption mode of nicotine and its salts on the surface of 12
silica gel can be established. Molecular nicotine has two N atoms, one is in the pyridine ring, and another is in the 1-methylpyrrolidine ring. Herein, we summarize three adsorption modes of nicotine on the surface of silica gel. 1) Both two N atoms of nicotine interact with two hydroxyl groups nearby, respectively (Fig. 5 upper part, A); 2) The N atom of the pyridine ring interacts with an isolated hydroxyl group (Fig. 5 upper part, B); 3) The N atom of the 1-methylpyrrolidine ring interacts with an isolated hydroxyl group (Fig. 5 upper part, C). Since 1-methylpyrrolidine is more basic than pyridine, and the silica gel shows weak acidic character, the intensity of the interaction of the three adsorption modes is ranked in the order: A>C>B. There are three types of adsorbed nicotine molecules desorb from the surface of silica gel, which can well explain the DTG curve of nicotine with a broad mass loss process in Fig. 2. When the temperature is raised up to 270℃, there will be a negative peak of molecular water present. This phenomenon can also be explained by our adsorption modes of nicotine on the surface of silica gel. When all the adsorbed nicotine desorbs from the surface of silica gel at 270℃, the occupied isolated hydroxyl groups will be released. Two isolated hydroxyl groups will lose a water molecule (Fig. 5 lower part) [28]. The moisture, coming from the occupied isolated hydroxyl groups on the surface of silica gel, plays a significant role in enhancing the interaction between nicotine/nicotine salts and silica gel. The nicotine precursor in our RTPs sample is mainly composed of nonprotonated nicotine, nicotine malate and nicotine citrate. We have carefully studied the desorption 13
and pyrolysis behaviors of nonprotonated nicotine, nicotine malate and nicotine citrate adsorbed on silica gel with TG and DRIFTS above. Since RTPs is a complex system, it is hard to study the interactions between nicotine precursor and RTPs by DRIFTS. As shown in Fig. S4 (Supplementary data), we can only catch the information including the reduction of adsorbed water (IR band at 3411 cm-1) and the generation of carbonyl (IR band at 1716 cm-1) with the temperature rising [26,29-31]. TG and DTG curves corresponding to the pyrolysis of RTPs in N2 are shown in Fig. 6. Within the studied temperature range, we can see that the remaining mass of RTPs is 25% at 500℃. From 200℃to 350℃, the mass loss rate is fastest. Here, we fit the DTG curve from 200℃to 350℃(Fig. 6), and two peaks of pyrolysis processes can be got. From the fitting results, the two steps of pyrolysis processes begin at 200℃and 250℃, respectively. In order to distinguish the two steps, we have studied the colour variety of RTPs pyrolysized at 210℃and 260℃, respectively [27,32]. The colours of RTPs pyrolized at 210℃ and 260℃ are shown in Fig S4 (Supplementary data), respectively. The RTPs at RT is bright yellow. When the RTPs are pyrolized at 210℃ for 5 minutes, the colour of them becomes dark yellow. The colour of the RTPs becomes black when they are pyrolized at 260℃for 5 minutes. According to the above study, the first pyrolysis process is a dehydration process beginning at 200℃, and the fastest mass loss rate is at 266℃. The second pyrolysis process is the carbonization of RTPs beginning at 250℃, and the fastest mass loss rate is at 305℃. In the DRIFTS results of nicotine and its salts, nicotine releases completely until the temperature was raised to 270℃. According to the results, we draw a conclusion that nicotine releases 14
at the first pyrolysis process. Next, we used a smoking machine to study release behavior of nicotine from the RTPs at selected temperatures. Cambridge filter pads (CFPs) were used to trap nicotine in the smoking process. The nicotine release of unit dry reconstituted tobacco particles (RTPs without moisture) at selected temperature is shown in Fig. 7. From 180℃to 270℃, nicotine release is gradually increasing. After 270℃, the nicotine release is constant, which is stable at around 6 mg/g. Nicotine releases before the second pyrolysis process, which means that nicotine releases before the carbonization of RTPs. By studying the simulated smoking state of RTPs, we find that the release behavior of nicotine corresponds well with our TG and DRIFTS results above.
4. Conclusions In conclusion, we employ nicotine, nicotine malate and nicotine citrate adsorbed on silica gel as the nicotine precursor to simulate the nicotine release process of RTPs. TG and DRIFTS are applied to study desorption and pyrolysis process of nicotine, nicotine malate and nicotine citrate, respectively. Finally, we used a smoking machine to study release behaviors of nicotine from the RTPs. Comparing the results between simulated study of nicotine and its salts adsorbed on silica gel and true nicotine release from RTPs, we can make the following conclusions: 1) Nicotine and its salts interacted with isolated hydroxyl groups of RTPs; 2) The main formof nicotine is nicotine salts in RTPs, and nicotine salts pyrolize from 200℃to 270℃. 15
We use a simple method to explain a complex question. The adsorption and thermal release of nicotine (its salts) from the RTPs have been indicated. Our results have a guiding significance for the development of heat-not-burn cigarettes.
ACKNOWLEDGEMENT The authors acknowledge the financial support from China National Tobacco Corporation, Shanghai Tobacco Group Co., LTD and Shanghai New Tobacco Research Institute.
.
REFERENCES [1] M. K. Akalin, S. Karagoz, Bioresources, 6 (2011) 1520-1531. [2] M. K. Akalin, S. Karagoz, Bioresources, 6 (2011) 1773-1805. [3] J. G. Lee, C. G. Lee, J. J. Kwag, M. S. Rhee, A. J. Buglass, G. H. Lee, J. Agric. Food. Chem., 55 (2007) 1097-1102. [4] C. L. Rainey, P. A. Conder, J. V. Goodpaster, J. Agric. Food. Chem., 59 (2011) 2745-2751.
16
[5] D. J. Eatough, C. L. Benner, H. Tang, V. Landon, G. Richards, F. M. Caka, J. Crawford, E. A. Lewis, L. D. Hansen, N. L. Eatough, Environ. Int., 15 (1989) 19-28. [6] M. S. Chen, L. Xu, H. L. Tian, C. Y. Yin, Z. L. Zhou, Y. Li, J. G. Ma, F. Zhong, J. Therm. Anal. Calorim., 118 (2014) 1747-1753. [7] M. Borgerding, H. Klus, Exp. Toxicol. Pathol., 57 (2005) 43-73. [8] C. F. Woodward, A. Eisner, P. G. Haines, J. Am. Chem. Soc., 66 (1944) 911-914. [9]
M. F. Borgerding, J. A. Bodnar, H. L. Chung, P. P. Mangan, C. C.
Morrison, C. H. Risner, J. C. Rogers, D. F. Simmons, M. S. Uhrig, F. N. Wendelboe, D. E. Wingate, L. S. Winkler, Food Chem. Toxicol., 36 (1998) 169-182. [10]
B. R. Bombick, H. Murli, J. T. Avalos, D. W. Bombick, W. T. Morgan,
K. P. Putnam, D. J. Doolittle, Food Chem. Toxicol., 36 (1998) 183-190. [11]
D. W. Bombick, P. H. Ayres, K. Putnam, B. R. Bombick, D. J. Doolittle,
Food Chem. Toxicol., 36 (1998) 191-197. [12]
S. C. Lee, K. L. Stokkeland, J. B. Cole, S. A. Bangh, Clin. Toxicol., 52
(2014) 777-777. [13]
D. J. Maron, S. P. Fortmann, Am. J. Public Health, 77 (1987) 546-549.
[14]
O. Rom, A. Pecorelli, G. Valacchi, A. Z. Reznick, Ann. Ny. Acad. Sci.,
1340 (2015) 65-74.
17
[15]
M. Forster, C. Liu, M. G. Duke, K. G. McAdam, C. J. Proctor, Chem.
Cent. J., 9 (2015). [16]
V. Valtchev, L. Tosheva, Chem. Rev. 113 (2013) 6734-6760.
[17]
W. C. Wilfong, B. W. Kail, C. W. Jones, C. Pacheco, M. L. Gray, ACS
Appl. Mater. Inter., 8 (2016) 12780-12791. [18]
J. Sirita, S. Phanichphant, F. C. Meunier, Anal. Chem., 79 (2007)
3912-3918. [19]
F. Rascon, R. Wischert, C. Coperet, Chem. Sci., 2 (2011) 1449-1456.
[20]
Y. Wang, B. A. Morrow, Langmuir, 12 (1996) 4153-4157.
[21]
N. Lazarevic, B. Adnadjevic, J. Jovanovic, Thermochim. Acta, 589 (2014)
100-106. [22]
Q. Hua, T. Cao, X.-K. Gu, J. Lu, Z. Jiang, X. Pan, L. Luo, W.-X. Li, W.
Huang, Angew. Chem. Int. Ed., 53 (2014) 4856-4861. [23]
U. Tumuluri, M. Isenberg, C. S. Tan, S. S. C. Chuang, Langmuir, 30
(2014) 7405-7413. [24]
L. Luo, Q. Hua, Z. Jiang, W. Huang, Rev. Sci. Instrum., 85 (2014)
064103. [25]
W. C. Wilfong, S. S. C. Chuang, Indeng. Chem. Res., 53 (2014)
4224-4231. [26]
A. Marcilla, M. I. Beltran, A. Gomez-Siurana, D. Berenguer, I.
Martinez-Castellanos, J. Anal. Appl. Pyrolysis, 119 (2016) 162-172.
18
[27]
A. Marcilla, M. I. Beltran, A. Gomez-Siurana, I. Martinez-Castellanos, D.
Berenguer, V. Pastor, A. N. Garcia, J. Anal. Appl. Pyrolysis, 112 (2015) 48-55. [28]
O. Sneh, M. A. Cameron, S. M. George, Surf. Sci., 364 (1996) 61-78.
[29]
C. R. Cardoso, C. H. Ataide, J. Anal. Appl. Pyrolysis, 99 (2013) 49-57.
[30]
O. Senneca, R. Chirone, P. Salatino, L. Nappi, J. Anal. Appl. Pyrolysis,
79 (2007) 227-233. [31]
O. Senneca, S. Ciaravolo, A. Nunziata, J. Anal. Appl. Pyrolysis, 79 (2007)
234-243. [32]
T. Ahamad, S. M. Alshehri, J. Hazard. Mater., 199 (2012) 200-208.
Figure captions: Fig. 1. A: TG (weight loss) and DTG (derivative of weight loss) curves corresponding to the thermal stability of silica gel in N2. B: DRIFTS spectra at selected temperature for the thermal stability of silica gel in He. Fig. 2. TG (A) and DTG (B) curves corresponding to the pyrolysis of nicotine and its salts adsorbed on silica gel in N2. Fig. 3. DRIFTS spectra at selected temperature for the desorption process of nicotine adsorbed on silica gel in He. 19
Fig. 4. DRIFTS spectra at selected temperature for the pyrolysis process of nicotine malate(A) and nicotine citrate(B) adsorbed on silica gel in He.
Fig. 5. Schematic diagram of the adsorption (upper) and desorption (lower) of nicotine on the surface of silica gel. Fig. 6. TG (weight loss) and DTG (derivative of weight loss) curves corresponding to the pyrolysis of RTPs in N2. Fig. 7. Nicotine release of unit dry RTPs at selected temperature.
Fig. 1. A: TG (weight loss) and DTG (derivative of weight loss) curves corresponding to the thermal stability of silica gel in N2. B: DRIFTS spectra at selected temperature for the thermal stability of silica gel in He.
20
Fig. 2. TG (A) and DTG (B) curves corresponding to the pyrolysis of nicotine and its salts adsorbed on silica gel in N2.
Fig. 3. DRIFTS spectra at selected temperature for the desorption process of nicotine adsorbed on silica gel in He. 21
Fig. 4. DRIFTS spectra at selected temperature for the pyrolysis process of nicotine malate(A) and nicotine citrate(B) adsorbed on silica gel in He.
Fig. 5. Schematic diagram of the adsorption (upper) and desorption (lower) of nicotine on the surface of silica gel.
22
Fig. 6. TG (weight loss) and DTG (derivative of weight loss) curves corresponding to the pyrolysis of RTPs in N2.
Fig. 7. Nicotine release of unit dry RTPs at selected temperature.
23
S0040-6031(17)30211-3 http://dx.doi.org/10.1016/j.tca.2017.08.013 TCA 77814
To appear in:
Thermochimica Acta
Received date: Revised date: Accepted date:
3-6-2017 2-8-2017 27-8-2017
Please cite this article as: Qing Hua, Wenjie Lu, Saijing Zheng, Yichun Zhang, Wei Zhang, Da Wu, Yi Shen, Thermal release of nicotine and its salts adsorbed on silica gel, Thermochimica Actahttp://dx.doi.org/10.1016/j.tca.2017.08.013 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Thermal release of nicotine and its salts adsorbed on silica gel Qing Hua, Wenjie Lu, Saijing Zheng, Yichun Zhang, Wei Zhang, Da Wu, Yi Shen*
Shanghai Tobacco Group Co., LTD, Shanghai 200082, China. *Corresponding author: Yi Shen Tel: +86 21 61668311; Fax: +86 21 61668366. E-mail address: [email protected] Graphical Abstract
1
Highlights
Nicotine and its salts adsorbed on silica gel are employed to simulate the nicotine release behaviors of RTPs.
TG and DRIFT are used to determine the interaction between amine groups of nicotine and silica gel.
Three adsorption modes of nicotine/its salts on the surface of silica gel have been established.
The pyrolysis temperatures of nicotine salts adsorbed on silica gel (and contained in RTPs) are determined.
ABSTRACT
Nicotine and its salts adsorbed on silica gel are employed to simulate the nicotine release behaviors of reconstituted tobacco particles (RTPs). Thermogravimetric analysis (TGA) was applied to provide an overview of the thermal behaviors of the nicotine, its salts and the RTPs. Diffuse reflectance infrared spectroscopy (DRIFTS) measurements of pyrolysis processes were performed on a FTIR spectrometer equipped with an in-situ DRIFTS reaction cell and a MCT/A detector. Firstly, the amine groups of nicotine and its salts are mainly adsorbed on the isolated hydroxyl groups of silica gel. Secondly, three adsorption modes of nicotine have been established. Finally, smoking machine was applied to study the release behavior of nicotine from the reconstituted tobacco particles (RTPs). The interaction between nicotine and the adsorbent is identified. Our results show interesting observations on 2
thermal release of nicotine from the different substrates, proving the adsorption and thermal release of nicotine and it salts from RTPs, which have a guiding significance for the development of heat-not-burn cigarettes.
KEYWORDS: Nicotine salt; DRIFT; TGA; Pyrolysis; Interaction
1. Introduction
Nicotine, an addictive component of tobacco smoke, is of interest in its evolution during the smoking process [1-4]. In tobacco leaves, nicotine is mainly in the form of nicotine salts [5, 6], most of which are nicotine malate and nicotine citrate. As we all known, cigarette smoke of conventional cigarettes is a highly complex aerosol system, involving over 6,000 identified chemicals in a dynamic and reactive mixture [6, 7]. These chemicals are generated by incomplete combustion of tobacco [8]. Tar generated in the burning process contains most toxicants from the tobacco smoke. Nowadays, an effective method to reduce the generation of toxicants from tobacco is to develop heat-not-burn cigarettes [9-11], which generates less tar than conventional cigarettes in the smoking process. There are two forms of heat-not-burn cigarettes. One form of heat-not-burn cigarettes looks like a cigarette. It involves a lit carbon tip that heats incoming air, which in turn heats tobacco-based substrates, forming an
3
aerosol containing mainly water, glycerol, nicotine and volatile tobacco components [12-14]. Another type of heat-not-burn cigarettes has a heating device that is powered by a battery, which works with a specially designed tobacco rod [15]. The heating of the tobacco includes heating from the inside of the tobacco rod, or from the external surface of the tobacco rod, or a combination of both. In both of these two forms of heat-not-burn cigarettes, the tobacco heating temperatures are typically below 300°C [15]. The temperature is high enough for tobacco to release nicotine, but tar generates still much less than conventional cigarettes. The related studies have indicated that the aerosol composition in heat-not-burn cigarette smoke contains less toxicant than that in conventional cigarette smoke [9-11]. The mechanism of nicotine release during smoking of conventional cigarettes is still not fully elucidated. Although the aerosol composition in heat-not-burn cigarette smoke is simpler than that in conventional cigarette smoke, it is difficult to study nicotine release behavior of heat-not-burn cigarettes directly. In order to simplify the experimental sample, silica gel is selected as the adsorbent in our research for the following reasons. Firstly, silica gel has great adsorption capability due to its big Brunner−Emmet−Teller (BET) area, which can adsorb enough nicotine/nicotine salts to show clear pictures of TGA and DRIFTS spectra [16]. Another reason for silica gel selected as the adsorbent in our research is that there are numerous hydroxyl groups present on the surface of silica gel. And the surface of tobacco substrate is polyhydroxy [17-21]. Therefore, silica gel with nicotine adsorbed on and its salts may truly reflect the release behaviors of nicotine from the tobacco substrate. 4
In this work, thermogravimetric analyses (TGA) was used to provide the thermal behaviors of nicotine and its salts. And diffuse reflectance infrared spectroscopy (DRIFTS) is applied to study the interaction between probes and adsorbent [22-24]. We employed TGA and DRIFTS to study the release behaviors of nicotine and its salts adsorbed on silica gel at selected temperature. TGA draws a continuous mass loss process, and first derivative of the averaged weight loss (DTG) indicates different pyrolysis stages. DRIFTS spectroscopy is capable of carrying out in situ characterization of the interaction between nicotine (its salts) and adsorbent. Combining the TGA and DRIFTS results, desorption/pyrolysis mechanisms of nicotine/nicotine salts can be investigated. Finally, the nicotine release behavior of reconstituted tobacco particles (RTPs) is examined, which completely fulfills our TGA and DRIFTS results. In a word, our results show interesting observations on thermal release of nicotine from the different substrates, proving the adsorption and thermal release of nicotine and it salts from RTPs, which have a guiding significance for the development of heat-not-burn cigarettes.
2. Material and methods 2.1 Material The nicotine salts were purchased from Huabao International Holdings Limited. The nicotine salts were dissolved in water and the details were in the table S1 (Supplementary data). The silica gel was purchased from Chemical Division of Silicycle. The reconstituted tobacco particles (noted as RTPs) was purchased from 5
Shanghai Tobacco Group Co., LTD. The RTPs were prepared by 100% reconstituted tobacco sheet by a grinder, and the size of those was between 1-2mm. The total content of nicotine in RTPs was 0.91%, and the moisture content was 16%. The RTPs and all chemicals were used as received.
Nicotine/silica gel samples were denoted followed the typical procedures: 40 μL nicotine, 100 μL anhydrous ethanol and 200 mg silica gel were co-added into a gray GC bottle and adequately mixed by stirring. Finally, the samples were dried in vacuum at room temperature (RT) for 2 hours. Nicotine salts/silica gel samples were denoted followed the typical procedures: 80 μL nicotine salts aqueous solution, 100 μL anhydrous ethanol and 200 mg silica gel were co-added into a gray GC bottle and adequately mixed by stirring. Finally, the samples were dried in vacuum at RT for 2 hours. 2.2 Thermogravimetric analyses (TGA) TGA (PerkinElmer, STA 8000) was performed on the nicotine, its salts and the reconstituted tobacco particles to provide an overview of their thermal behavior. For this experiment, 20~25 mg sample was loaded into a Pt crucible. For each TGA experiment, the temperature ranged from 35°C to 500°C in N2 (sample purge: 20 mL/min; balance purge: 40 mL/min) under 20 °C/min heating rate, and held for 5 min at 500°C. First derivative of the averaged weight loss (DTG) was also obtained. 2.3 Diffuse reflectance infrared spectroscopy (DRIFTS) DRIFTS measurements of pyrolysis processes were performed on a Nicolet 5700 FTIR spectrometer equipped with an in situ DRIFTS reaction cell (Harrick Scientific 6
Products, INC) and a MCT/A detector. He (flow rate: 20 ml/min) was employed as the carrying gas for the pyrolysis. Silica gel was purged in the He stream (flow rate: 20 ml/min) at 30 °C for 1 h until there were no changes of the DRIFTS spectrum. And the corresponding DRIFTS spectrum was taken as the background spectrum. Nicotine salts adsorbed on silica were pretreated in the He stream (flow rate: 20 ml/min) for 0.5 h to remove the adsorbed water on the surface of silica gel. The pyrolysis process was in situ studied by the DRIFTS measurements performed with 64 scans and a resolution of 4 cm-1. The corresponding DRIFTS spectrum was taken when the sample was heated in the He stream (flow rate: 20 ml/min) to desirable temperatures under 10 °C/min heating rate and held at the desirable temperature for 3 min. 2.4 Nicotine release study of reconstituted tobacco particles (RTPs) The reconstituted tobacco particles (RTPs) and Cambridge filter pads (CFPs, Borgwaldt, Germany) used for our study were conditioned at 22±1 °C and a humidity of 60±2% for at least 48 h. 65 mg RTPs were filled into a U-sharp quartz tube. The heating equipment was a GC oven (Agilent 6890). The RTPs were puffed in a puff volume of 35 mL with 2 s/puff duration in every 15 s by using a 20-port Borgwaldt RM200 smoking machine (Borgwaldt, Germany) according to ISO 4387 at selected temperatures. The smoke particulate matter of the RTPs in one tube was collected in one CFP with a diameter of 44 mm in totally 20 puffs. The schematic diagram was shown in Fig. S1 (Supplementary data).
7
3. Results and discussion In this work, silica gel was employed as the adsorbent to adsorb nicotine and its salts to study their pyrolysis behaviors. Silica gel is a common adsorbent, which can adsorb enough nicotine/nicotine salts to show clear TGA and DRIFTS spectra [16]. Firstly, we studied the thermal stability of silica gel (as shown in Fig. 1A). The TG curve of silica gel indicates that the weight loss of silica gel is about 2.5% when the temperature is raised up to 150℃. When the temperature is raised up to 500℃, the weight loss of silica gel is about 3.5%. In order to further study the mechanism of the weight loss of silica gel, a series of DRIFTS spectra in situ for the thermal stability of silica gel in He acquired at selected temperatures are shown in Fig. 1B. Compared with the DTG curve (Fig. 1A), it is clearly shown that the water adsorbed on the surface of silica gel firstly desorbed before 150℃. The IR band of silica gel at 3740 cm−1 is slowly enhanced, proving the increase in the number of isolated surface hydroxyl groups on the surface of silica gel [21, 25]. Silica gel has a considerable adsorption capacity due to its large surface area (~ 500 m2/g) [16]. Therefore, silica gel is selected as an excellent adsorbent to study the pyrolysis of nicotine and its salts. According to Fig. 2A, the mass losses of nicotine/silica gel (noted as Nic/SG), nicotine malate/silica gel (noted as Nic-mal/SG) and nicotine citrate/silica gel (noted as Nic-cit/SG) were about 15%. The mass loss under 100°C was contributed by adsorbed water. A detailed pyrolysis process of nicotine and its salts adsorbed on silica gel has been studied by DTG (Fig. 2B). The nicotine undergoes a wide mass-loss process from 100°C to 332°C and reaches the 8
fastest desorption rate at 225°C [26]. In this process, the mass-loss is attributed to desorption of nicotine from the adsorbent (silica gel). The mass-loss processes of nicotine malate and nicotine citrate were divided into two parts. First part is the pyrolysis of nicotine salts, and the second part is desorption of the rest nicotine adsorbed on the adsorbent. The second parts of DTG curves of nicotine malate and nicotine citrate were analogous. Both of these two parts are the desorption processes of the pyrolyzed nicotine adsorbed on the silica gel. The first parts of DTG curves of nicotine malate and nicotine citrate were different because of the difference of the stabilities of nicotine malate and nicotine citrate adsorbed on silica gel. The difference is due to that the stabilities of nicotine malate and nicotine citrate adsorbed on silica gel are not the same. The nicotine malate adsorbed on silica gel begins to pyrolyze from 153°C and reaches the fastest pyrolysis rate at 203°C. In contrast, the nicotine citrate adsorbed on silica gel begins to pyrolyze from 128°C and reaches the fastest pyrolysis rate at 183°C. The first parts of DTG curves of nicotine malate and nicotine citrate indicate that nicotine malate is more stable than nicotine citrate when they are adsorbed on silica gel. Similarly, in order to further study the mechanism of the pyrolysis processes of nicotine and its salts adsorbed on silica gel [27]. A DRIFTS study of nicotine, nicotine malate and nicotine citrate was carried out. Fig. 3 is the DRIFTS spectra in situ for the desorption process of nicotine adsorbed on silica gel in He acquired at selected temperature. As the temperature is raised, the negative peak at 3738 cm-1 disappears slowly. The broad peak at 2300 cm-1 reduces slowly, and the peaks between 3200 cm-1 9
to 2500 cm-1 reflect the same trend with it. The negative peak of 3738 cm-1 belongs to isolated surface hydroxyl groups on the surface of silica gel [18,19]. Compared with the standard IR spectrum of nicotine (Fig. S2 in Supplementary data), the peaks between 3200 cm-1 to 2500 cm-1 come from nicotine. The broad peak of 2300 cm-1 comes from the interactions between the hydroxyl groups of silica gel and the amine groups of nicotine [21,24,25]. Since the background of the DRIFTS spectra is dry silica gel, it has strong IR band at 3738 cm-1. The DRIFTS spectra of silica gel with different purged time are shown in Fig. S3 (Supplementary data). Therefore, there are few isolated surface hydroxyl groups on the sample of Nic/SG. When the adsorbed nicotine is being desorbed with the temperature being raised, the amount of isolated hydroxyl groups on the surface of silica gel gradually is increasing. This indicates that nicotine occupies the isolated hydroxyl groups on the surface of silica gel. From Fig. 3, we can see that nicotine adsorbed on silica gel begins to desorb at 120℃. When the temperature reaches 270℃, almost all nicotine adsorbed on the surface of silica gel has desorbed. The desorption process obtained from DRIFTS is in agreement with that from TG (Fig. 2). Interestingly, a negative IR band at 3250 cm-1 appears when most of the adsorbed nicotine desorbs at 240℃. The IR band at 3250 cm-1 belongs to molecular water [20]. This part of moisture on the surface of silica gel desorbs until most of adsorbed nicotine desorbed. The result confirms that this part of moisture plays a supporting role to enhance the interaction between nicotine and silica gel. Fig. 4A is the DRIFTS spectra in situ for the pyrolysis process of nicotine malate adsorbed on silica gel in He acquired at selected temperatures. The 10
negative IR band at 3738 cm-1 disappears slowly with the temperature being raised. The broad IR bands at 2235 cm-1 and 2555 cm-1 reduce slowly, and the peaks between 3200 cm-1 to 2500 cm-1 show the same trend with them. The IR bands at 1728 cm-1 and 1578 cm-1 have the same decreasing trend until the temperature reaches 270℃. When the temperature is raised up to 300℃, the IR band at 1728 cm-1 is still noticed but with burrs. The negative peak at 3738 cm-1 belongs to isolated surface hydroxyl groups on the surface of silica gel. The peaks between 3200 cm-1 to 2500 cm-1 come from nicotine. The broad peak of 2235 cm-1 comes from the interactions between the hydroxyl groups of silica gel and the amine groups of nicotine. And the broad peak of 2555 cm -1 comes from the interactions between the hydroxyl groups of silica gel and the carbonyl groups of malic acid. According to the standard IR spectra of nicotine and malic acid (Fig. S2 in Supplementary data), the IR band at 1728 cm-1 is attributable to the carbonyl groups of malic acid, and the IR band at 1578 cm -1 is attributable to the amine groups of nicotine. From Fig. 4A, we can see that nicotine malate adsorbed on silica gel begins to pyrolize at 150℃. When the temperature reaches 270℃, almost all nicotine malate adsorbed on the surface of silica gel has pyrolized. The pyrolysis process obtained from DRIFTS solidly agrees with that from TG (Fig. 2). Interestingly, a negative IR band at 3250 cm-1 appears when most of the adsorbed nicotine malate pyrolized at 240℃. The IR band at 3250 cm-1 belongs to molecular water. This part of moisture on the surface of silica gel desorbs until most of adsorbed nicotine malate pyrolized. The result confirms that this part of 11
moisture plays a supporting role in enhancing the interaction between nicotine malate and silica gel. The pyrolysis process of nicotine citrate adsorbed on silica gel is same as that of nicotine malate adsorbed on silica gel (Fig. 4B). However, the IR band of carbonyl group at 1728 cm-1 is different between nicotine malate/SG and nicotine citrate/SG. The IR bands of carbonyl group at 1728 cm-1 are different for nicotine malate/SG and nicotine citrate/SG. When the temperature is raised up to 300℃, the IR band at 1728 cm-1 of nicotine malate/SG is still noticed but with burrs. At this temperature, the IR band at 1728 cm-1 of nicotine citrate/SG is very small. These results indicate that the IR band at 1728 cm-1 with burrs is formed via the carbonization of nicotine malate. According to TG, nicotine malate is more stable than nicotine citrate when they are adsorbed on silica gel. In another word, the interaction between nicotine malate and silica gel is stronger than that between nicotine citrate and silica gel. Nicotine citrate pyrolizes and desorbs from the silica gel at 300℃, while nicotine malate will pyrolize but not desorb from the silica gel in time. Therefore, a small part of nicotine malate will be carbonized, and the IR band at 1728 cm-1 forms. By further studying the pyrolysis DRIFTS results of nicotine and its salts adsorbed on the surface of silica gel, two things should be mentioned: 1) The amine groups of nicotine and its salts interact with isolated surface hydroxyl groups of silica gel; 2) When most of the adsorbed nicotine/nicotine salts get pyrolized and desorbed, there would be molecular water desorbed from silica gel. Based on the above, the adsorption mode of nicotine and its salts on the surface of 12
silica gel can be established. Molecular nicotine has two N atoms, one is in the pyridine ring, and another is in the 1-methylpyrrolidine ring. Herein, we summarize three adsorption modes of nicotine on the surface of silica gel. 1) Both two N atoms of nicotine interact with two hydroxyl groups nearby, respectively (Fig. 5 upper part, A); 2) The N atom of the pyridine ring interacts with an isolated hydroxyl group (Fig. 5 upper part, B); 3) The N atom of the 1-methylpyrrolidine ring interacts with an isolated hydroxyl group (Fig. 5 upper part, C). Since 1-methylpyrrolidine is more basic than pyridine, and the silica gel shows weak acidic character, the intensity of the interaction of the three adsorption modes is ranked in the order: A>C>B. There are three types of adsorbed nicotine molecules desorb from the surface of silica gel, which can well explain the DTG curve of nicotine with a broad mass loss process in Fig. 2. When the temperature is raised up to 270℃, there will be a negative peak of molecular water present. This phenomenon can also be explained by our adsorption modes of nicotine on the surface of silica gel. When all the adsorbed nicotine desorbs from the surface of silica gel at 270℃, the occupied isolated hydroxyl groups will be released. Two isolated hydroxyl groups will lose a water molecule (Fig. 5 lower part) [28]. The moisture, coming from the occupied isolated hydroxyl groups on the surface of silica gel, plays a significant role in enhancing the interaction between nicotine/nicotine salts and silica gel. The nicotine precursor in our RTPs sample is mainly composed of nonprotonated nicotine, nicotine malate and nicotine citrate. We have carefully studied the desorption 13
and pyrolysis behaviors of nonprotonated nicotine, nicotine malate and nicotine citrate adsorbed on silica gel with TG and DRIFTS above. Since RTPs is a complex system, it is hard to study the interactions between nicotine precursor and RTPs by DRIFTS. As shown in Fig. S4 (Supplementary data), we can only catch the information including the reduction of adsorbed water (IR band at 3411 cm-1) and the generation of carbonyl (IR band at 1716 cm-1) with the temperature rising [26,29-31]. TG and DTG curves corresponding to the pyrolysis of RTPs in N2 are shown in Fig. 6. Within the studied temperature range, we can see that the remaining mass of RTPs is 25% at 500℃. From 200℃to 350℃, the mass loss rate is fastest. Here, we fit the DTG curve from 200℃to 350℃(Fig. 6), and two peaks of pyrolysis processes can be got. From the fitting results, the two steps of pyrolysis processes begin at 200℃and 250℃, respectively. In order to distinguish the two steps, we have studied the colour variety of RTPs pyrolysized at 210℃and 260℃, respectively [27,32]. The colours of RTPs pyrolized at 210℃ and 260℃ are shown in Fig S4 (Supplementary data), respectively. The RTPs at RT is bright yellow. When the RTPs are pyrolized at 210℃ for 5 minutes, the colour of them becomes dark yellow. The colour of the RTPs becomes black when they are pyrolized at 260℃for 5 minutes. According to the above study, the first pyrolysis process is a dehydration process beginning at 200℃, and the fastest mass loss rate is at 266℃. The second pyrolysis process is the carbonization of RTPs beginning at 250℃, and the fastest mass loss rate is at 305℃. In the DRIFTS results of nicotine and its salts, nicotine releases completely until the temperature was raised to 270℃. According to the results, we draw a conclusion that nicotine releases 14
at the first pyrolysis process. Next, we used a smoking machine to study release behavior of nicotine from the RTPs at selected temperatures. Cambridge filter pads (CFPs) were used to trap nicotine in the smoking process. The nicotine release of unit dry reconstituted tobacco particles (RTPs without moisture) at selected temperature is shown in Fig. 7. From 180℃to 270℃, nicotine release is gradually increasing. After 270℃, the nicotine release is constant, which is stable at around 6 mg/g. Nicotine releases before the second pyrolysis process, which means that nicotine releases before the carbonization of RTPs. By studying the simulated smoking state of RTPs, we find that the release behavior of nicotine corresponds well with our TG and DRIFTS results above.
4. Conclusions In conclusion, we employ nicotine, nicotine malate and nicotine citrate adsorbed on silica gel as the nicotine precursor to simulate the nicotine release process of RTPs. TG and DRIFTS are applied to study desorption and pyrolysis process of nicotine, nicotine malate and nicotine citrate, respectively. Finally, we used a smoking machine to study release behaviors of nicotine from the RTPs. Comparing the results between simulated study of nicotine and its salts adsorbed on silica gel and true nicotine release from RTPs, we can make the following conclusions: 1) Nicotine and its salts interacted with isolated hydroxyl groups of RTPs; 2) The main formof nicotine is nicotine salts in RTPs, and nicotine salts pyrolize from 200℃to 270℃. 15
We use a simple method to explain a complex question. The adsorption and thermal release of nicotine (its salts) from the RTPs have been indicated. Our results have a guiding significance for the development of heat-not-burn cigarettes.
ACKNOWLEDGEMENT The authors acknowledge the financial support from China National Tobacco Corporation, Shanghai Tobacco Group Co., LTD and Shanghai New Tobacco Research Institute.
.
REFERENCES [1] M. K. Akalin, S. Karagoz, Bioresources, 6 (2011) 1520-1531. [2] M. K. Akalin, S. Karagoz, Bioresources, 6 (2011) 1773-1805. [3] J. G. Lee, C. G. Lee, J. J. Kwag, M. S. Rhee, A. J. Buglass, G. H. Lee, J. Agric. Food. Chem., 55 (2007) 1097-1102. [4] C. L. Rainey, P. A. Conder, J. V. Goodpaster, J. Agric. Food. Chem., 59 (2011) 2745-2751.
16
[5] D. J. Eatough, C. L. Benner, H. Tang, V. Landon, G. Richards, F. M. Caka, J. Crawford, E. A. Lewis, L. D. Hansen, N. L. Eatough, Environ. Int., 15 (1989) 19-28. [6] M. S. Chen, L. Xu, H. L. Tian, C. Y. Yin, Z. L. Zhou, Y. Li, J. G. Ma, F. Zhong, J. Therm. Anal. Calorim., 118 (2014) 1747-1753. [7] M. Borgerding, H. Klus, Exp. Toxicol. Pathol., 57 (2005) 43-73. [8] C. F. Woodward, A. Eisner, P. G. Haines, J. Am. Chem. Soc., 66 (1944) 911-914. [9]
M. F. Borgerding, J. A. Bodnar, H. L. Chung, P. P. Mangan, C. C.
Morrison, C. H. Risner, J. C. Rogers, D. F. Simmons, M. S. Uhrig, F. N. Wendelboe, D. E. Wingate, L. S. Winkler, Food Chem. Toxicol., 36 (1998) 169-182. [10]
B. R. Bombick, H. Murli, J. T. Avalos, D. W. Bombick, W. T. Morgan,
K. P. Putnam, D. J. Doolittle, Food Chem. Toxicol., 36 (1998) 183-190. [11]
D. W. Bombick, P. H. Ayres, K. Putnam, B. R. Bombick, D. J. Doolittle,
Food Chem. Toxicol., 36 (1998) 191-197. [12]
S. C. Lee, K. L. Stokkeland, J. B. Cole, S. A. Bangh, Clin. Toxicol., 52
(2014) 777-777. [13]
D. J. Maron, S. P. Fortmann, Am. J. Public Health, 77 (1987) 546-549.
[14]
O. Rom, A. Pecorelli, G. Valacchi, A. Z. Reznick, Ann. Ny. Acad. Sci.,
1340 (2015) 65-74.
17
[15]
M. Forster, C. Liu, M. G. Duke, K. G. McAdam, C. J. Proctor, Chem.
Cent. J., 9 (2015). [16]
V. Valtchev, L. Tosheva, Chem. Rev. 113 (2013) 6734-6760.
[17]
W. C. Wilfong, B. W. Kail, C. W. Jones, C. Pacheco, M. L. Gray, ACS
Appl. Mater. Inter., 8 (2016) 12780-12791. [18]
J. Sirita, S. Phanichphant, F. C. Meunier, Anal. Chem., 79 (2007)
3912-3918. [19]
F. Rascon, R. Wischert, C. Coperet, Chem. Sci., 2 (2011) 1449-1456.
[20]
Y. Wang, B. A. Morrow, Langmuir, 12 (1996) 4153-4157.
[21]
N. Lazarevic, B. Adnadjevic, J. Jovanovic, Thermochim. Acta, 589 (2014)
100-106. [22]
Q. Hua, T. Cao, X.-K. Gu, J. Lu, Z. Jiang, X. Pan, L. Luo, W.-X. Li, W.
Huang, Angew. Chem. Int. Ed., 53 (2014) 4856-4861. [23]
U. Tumuluri, M. Isenberg, C. S. Tan, S. S. C. Chuang, Langmuir, 30
(2014) 7405-7413. [24]
L. Luo, Q. Hua, Z. Jiang, W. Huang, Rev. Sci. Instrum., 85 (2014)
064103. [25]
W. C. Wilfong, S. S. C. Chuang, Indeng. Chem. Res., 53 (2014)
4224-4231. [26]
A. Marcilla, M. I. Beltran, A. Gomez-Siurana, D. Berenguer, I.
Martinez-Castellanos, J. Anal. Appl. Pyrolysis, 119 (2016) 162-172.
18
[27]
A. Marcilla, M. I. Beltran, A. Gomez-Siurana, I. Martinez-Castellanos, D.
Berenguer, V. Pastor, A. N. Garcia, J. Anal. Appl. Pyrolysis, 112 (2015) 48-55. [28]
O. Sneh, M. A. Cameron, S. M. George, Surf. Sci., 364 (1996) 61-78.
[29]
C. R. Cardoso, C. H. Ataide, J. Anal. Appl. Pyrolysis, 99 (2013) 49-57.
[30]
O. Senneca, R. Chirone, P. Salatino, L. Nappi, J. Anal. Appl. Pyrolysis,
79 (2007) 227-233. [31]
O. Senneca, S. Ciaravolo, A. Nunziata, J. Anal. Appl. Pyrolysis, 79 (2007)
234-243. [32]
T. Ahamad, S. M. Alshehri, J. Hazard. Mater., 199 (2012) 200-208.
Figure captions: Fig. 1. A: TG (weight loss) and DTG (derivative of weight loss) curves corresponding to the thermal stability of silica gel in N2. B: DRIFTS spectra at selected temperature for the thermal stability of silica gel in He. Fig. 2. TG (A) and DTG (B) curves corresponding to the pyrolysis of nicotine and its salts adsorbed on silica gel in N2. Fig. 3. DRIFTS spectra at selected temperature for the desorption process of nicotine adsorbed on silica gel in He. 19
Fig. 4. DRIFTS spectra at selected temperature for the pyrolysis process of nicotine malate(A) and nicotine citrate(B) adsorbed on silica gel in He.
Fig. 5. Schematic diagram of the adsorption (upper) and desorption (lower) of nicotine on the surface of silica gel. Fig. 6. TG (weight loss) and DTG (derivative of weight loss) curves corresponding to the pyrolysis of RTPs in N2. Fig. 7. Nicotine release of unit dry RTPs at selected temperature.
Fig. 1. A: TG (weight loss) and DTG (derivative of weight loss) curves corresponding to the thermal stability of silica gel in N2. B: DRIFTS spectra at selected temperature for the thermal stability of silica gel in He.
20
Fig. 2. TG (A) and DTG (B) curves corresponding to the pyrolysis of nicotine and its salts adsorbed on silica gel in N2.
Fig. 3. DRIFTS spectra at selected temperature for the desorption process of nicotine adsorbed on silica gel in He. 21
Fig. 4. DRIFTS spectra at selected temperature for the pyrolysis process of nicotine malate(A) and nicotine citrate(B) adsorbed on silica gel in He.
Fig. 5. Schematic diagram of the adsorption (upper) and desorption (lower) of nicotine on the surface of silica gel.
22
Fig. 6. TG (weight loss) and DTG (derivative of weight loss) curves corresponding to the pyrolysis of RTPs in N2.
Fig. 7. Nicotine release of unit dry RTPs at selected temperature.
23