Size Control of Ti₄O₇ Nanoparticles by Carbothermal Reduction Using a Multimode Microwave Furnace

Abstract

Utilization of Ti₄O₇ in applications such as catalyst support calls for control over the size of the Ti₄O₇ nanoparticles. This can be achieved using a simple process such as carbothermal reduction. In this study, various sizes of Ti₄O₇ nanoparticles (25, 60, and 125 nm) were synthesized by carbothermal reduction using a multimode microwave apparatus. It was possible to produce Ti₄O₇ nanoparticles as small as 25 nm by precisely controlling the temperature, heating process, and holding time of the sample while taking advantage of the characteristics of microwave heating such as rapid and volumetric heating. The results show that microwave carbothermal reduction is advantageous in controlling the size of the Ti₄O₇ nanoparticles.

1. Introduction

Ti₄O₇ is a Magnéli phase material [1] that exhibits excellent electrical conductivity at room temperature, with a value of about 10³ S cm⁻¹, which is comparable to graphite [2,3,4]. In addition, Ti₄O₇ shows high corrosion resistance [5] and high stability in electrochemical environments [3] as well as acidic electrolytes [6]. From these features, it is possible to use Ti₄O₇ in different applications, such as unitized regenerative fuel cells [7,8], polymer electrolyte fuel cells [9,10,11], lithium-sulfur batteries [12,13,14], and water filtration systems [15,16].

For all these applications, it is necessary to synthesize single-phase Ti₄O₇ nanomaterials with a high specific surface area [4]. In previous works, Magnéli phase nanoparticles were synthesized using different methods, such as calcination of titanium ethoxide and polyethylene glycol solution [17], thermal plasma treatment of H₂TiO₃ under Ar/H₂ [18], sol-gel and calcination [19], sol–gel and vacuum-carbothermic processes [20], pulsed UV laser irradiation [21], and thermal-induced plasma processes [22]. In addition, Zhang et al. prepared fiber-like Ti₄O₇ by heating intermediate H₂Ti₃O₇ at 1050 °C under a hydrogen atmosphere [23], but the fibers sintered and became submicron chains. Size control of Ti₄O₇ nanoparticles is still a challenge because nanoparticles are prone to heavy sintering, e.g., sintering of TiO₂ nanoparticles begins at approximately 700 °C [24].

In addition to controlling the size of the nanoparticles, it is also necessary to employ simple, practical techniques for the synthesis. In our previous works, Ti₄O₇ nanoparticles were prepared using microwave irradiation by a carbothermal reduction process, where the device used was a single-mode furnace [25,26]. To scale up the microwave carbothermal process for industry, the process has to be adapted to multimode furnaces. It is more difficult to control the sample temperature in a multimode-type furnace than in a single-mode microwave furnace because the electromagnetic field distribution tends to be no-uniform in the former case. In order to prepare nanoparticles by the carbothermal reduction process, accurate temperature control is required, especially with rapid temperature increase.

Taking all these factors into account, we synthesized Ti₄O₇ nanoparticles in various sizes (25, 60, and 125 nm) using a carbothermal reduction method with a multimode-type microwave apparatus. Our experimental set-up included a self-made susceptor (using carbon powder with high microwave absorbing power) and proportional–integral–derivative (PID) temperature control.

2. Materials and Methods

Figure 1 shows the sample setting. TiO₂ samples with three particle sizes were used as pristine samples: Sample 1 (TTO-51(A), particle size about 25 nm, Ishihara Sangyo Kaisha, Ltd.), Sample 2 (particle size about 60 nm, 99.5%, IoLiTec-Ionic Liquids Technologies GmbH, Heilbronn, Deutschland), and Sample 3 (particle size about 110 nm, EM Japan Co., Ltd., Tokyo, Japan). Sample 1 was washed with 5M NaOH to eliminate Al(OH)₃, which is used as a protective agent. Pristine TiO₂ was dispersed in water with dissolved polyvinylpyrrolidone (PVP, molecular weight: 40,000, Wako Pure Chemical Industries, Ltd., Osaka, Japan) by sonication. The ratio of weight between TiO₂ and PVP is 1.21 g:5.59 g. The solution was dried to obtain PVP-coated TiO₂ nanoparticles. PVP decomposes and becomes a carbon source during microwave heating. The PVP-coated TiO₂ nanoparticles (0.15 g of Sample 1, 0.3 g of Samples 2 and 0.3 g of Sample 3) were filled in a quartz tube. After sealing with silica wool, 0.8 g of titanium sponge (Wako Pure Chemical Industries, Ltd., Osaka, Japan) was added as an oxygen absorber. This quartz tube was set in a quartz container filled with a carbon susceptor. The pressure inside the quartz tube was controlled with a rotary pump so that the base pressure of the system was about 10 Pa. A microwave irradiation furnace from μReactor Ex (Shikoku Instrumentation Co., Ltd., Kagawa, Japan) was used. The process temperature was measured with a thermocouple, which was placed between the susceptor and the quartz tube. Microwave-irradiated samples were analyzed by X-ray diffraction (XRD, RINT-2200/PC, Rigaku Co., Tokyo, Japan) and a field emission scanning electron microscope (FE-SEM, S4800, Hitachi High-Technologies Co., Tokyo, Japan). In the analysis of particle size distribution, we traced each surface of particles in FE-SEM images, and used ImageJ software to calculate the number average particle diameter from the area of traced particles.

3. Results and Discussion

To achieve a reduction of the pristine TiO₂ to Ti₄O₇ and retaining of the nanomorphology of pristineTiO₂ at the same time, various experiments were conducted in different heating regimes to decide the optimal experiment condition. The experimental results of various heating regime to synthesize Ti₄O₇ nanoparticles from Sample 1 was summarized as

Table 1. Experimental results of various heating regime to synthesize Ti₄O₇ nanoparticles from Sample 1.

No.	Rate of Heating/°C s−1	Holding Temperature/°C	Holding Time/min.	Synthesized Phase
1	12.0	900	10	TiO2
2	12.0	925	10	Ti7O13 + Ti4O7
3	12.0	925	11	Ti4O7 + Ti6O11
4	12.0	925	13	Ti4O7
5	12.0	925	15	Ti4O7 + Ti3O5
6	12.0	925	20	Ti3O5 + Ti4O7
7	10.0	950	30	Ti3O5 + Ti2O3

. In 25 nm pristine TiO₂ case, reduction reaction was too fast to obtain Ti₄O₇ when the heating regime was same to the previous paper (No. 7) [25]. To control the reduction-reaction speed in high temperature region, rate of heating, holding temperature and holding time was changed to fast, low and short, respectively. Below 900 °C, sample was not reduced. At 925 °C, Ti₄O₇ phase was obtained, and when the holding time was 13 min, single phase Ti₄O₇ was obtained.

Figure 2 shows the profiles of microwave power and process temperature during microwave processing when a reduction of the pristine TiO₂ to Ti₄O₇ and retaining of the nanomorphology of pristineTiO₂ was achieved at the same time.

Table 2. Experimental conditions in microwave processing.

No.	Rate of Heating /°C s−1	Holding Temperature/°C	Holding Time/min.	Microwave Power @ Holding/W
Sample 1	12.0	925	13	168.0
Sample 2	3.4	975	30	171.0
Sample 3	5.7	975	30	181.1

shows the optimal condition of the rate of heating, holding temperature, holding time, and average microwave power during temperature holding for each process. For Sample 1, the holding temperature was lower and the holding time was shorter than the other samples as a smaller particle size results in a more effective reduction reaction. In addition, to obtain Ti₄O₇ nanoparticles, it is necessary to increase the temperature rapidly and shorten the holding time at high temperatures (above 700 °C) in order to prevent grain growth. The process temperature was 975 °C for Sample 2. This temperature is different compared to a previous study [25] and is due to the temperature measurement method. In the previous study, the thermocouple was located in the sample powder. However, in this experiment, the thermocouple was placed between the susceptor and the quartz tube used to hold the sample powder. Thus, the process temperature measured in this experiment is different from the sample temperature: it is possible that the true sample temperature is lower. We can consider that the temperature of the susceptor is the main factor in maintaining the temperature of the system because the carbon volume of the susceptor is much larger than that of PVP. Regardless of the particle size, the average microwave power values during the holding process were almost the same.

Figure 3 shows XRD patterns of pristine TiO₂ and the samples after the microwave process. All pristine TiO₂ samples were in the rutile phase. The right-side figure shows XRD patterns of synthesized samples. In this figure, Ti₄O₇_1, Ti₄O₇_2 and Ti₄O₇_3 refer to synthesized samples after microwave processing from Sample 1, Sample 2 and Sample 3, respectively. The synthesized samples were single-phase Ti₄O₇ without any other Ti-O phase.

Crystallite diameters of the pristine TiO₂ and the synthesized Ti₄O₇ were analyzed from XRD patterns using Scherrer equation, and the results were summarized in

Table 3. Crystallite diameter of pristine TiO₂ and synthesized Ti₄O₇ by microwave; MW: microwave

	Before MW Process	After MW Process
Sample 1	9.2 (0.3) nm	9.2 nm
Sample 2	28.3 (0.4) nm	25.3 nm
Sample 3	31.0 (0.9) nm	29.5 nm

. All crystallite diameters were smaller than the primary particle diameter confirmed by FE-SEM: thus the pristine TiO₂ and the synthesized Ti₄O₇ was polycrystalline. The crystallite diameter of pristine TiO₂ was almost the same value as synthesized Ti₄O₇.

Figure 4 shows FE-SEM images of pristine TiO₂ and synthesized Ti₄O₇ nanoparticles acquired in SE mode. Although some coarse particles were observed in pristine TiO₂, the particle diameter was uniform. From SE-images of Ti₄O₇_1 and Ti₄O₇_2, there seems to be grain growth, but this is because of residual carbon around the Ti₄O₇ nanoparticles. Figure 5 shows FE-SEM images of the samples acquired in TE mode, where Ti₄O₇ particles can be clearly observed, since electrons go through the carbon layer. The TE-images for Ti₄O₇_1 and Ti₄O₇_2 also show the difference between particles and residual carbon when compared with SE-images. For Ti₄O₇_3 nanoparticles, the carbon coating was relatively thinner, making it easier to observe and determine the diameter of the nanoparticles.

Figure 6 shows histograms of particle sizes for pristine TiO₂ and synthesized Ti₄O₇.

Table 4. Average particle size (Ave. diameter), standard deviation (S.D.), maximum particle diameter (Max.), minimum particle diameter (Min.), sample number (n), and standard error (S.E.) of pristine TiO₂ and synthesized Ti₄O₇.

No.	Ave. Diameter/nm	S.D./nm	Max./nm	Min./nm	n/-	S.E./nm
Sample 1	25.8	7.9	69.5	13.9	100	0.8
Ti4O7_1	24.7	6.7	43.5	12.2	100	0.7
Sample 2	54.6	13.9	91.1	29.2	120	1.3
Ti4O7_2	60.4	12.8	94.6	30.7	120	1.2
Sample 3	108.0	38.0	246.2	41.8	50	5.4
Ti4O7_3	125.0	37.5	215.6	51.0	50	5.3

shows the average particle size (Ave. diameter), standard deviation (S.D.), maximum particle diameter (Max.), minimum particle diameter (Min.), sample number (n), and standard error (S.E.). The average nanoparticle diameter of Sample 1, Sample 2, and Sample 3 was 25.8, 54.6, and 108.0 nm, respectively. The average nanoparticle diameter of Ti₄O₇_1 was 24.7 nm, which is within the error margin of the average diameter of Sample 1. On the other hand, the average nanoparticle diameter of Ti₄O₇_2 was 60.4 nm; indicating a 3.3 nm grain growth. For Ti₄O₇_3, the average nanoparticle diameter was 6.3 nm larger compared to the average nanoparticle diameter of Sample 3. However, since the maximum particle size of Sample 3 was larger than the measured sizes of Ti₄O₇_3 particles, it can be deduced that there was no significant grain growth.

From these results, even when a simple apparatus such as a multimode microwave irradiation furnace is used, it is possible to produce Ti₄O₇ nanoparticles with an average size as low as 25 nm. This can be done using a carbon thermal reduction method by precise control of the heating rate, holding temperature, and holding time.

4. Conclusions

Ti₄O₇ nanoparticles of various sizes (25, 60, 125 nm) were prepared in a multimode microwave furnace. In general, the synthesized Ti₄O₇ nanoparticles retained the size of the pristine TiO₂ nanoparticles. The process time and temperature varied depending on the size of nanoparticles because the reduction reaction was slower when larger pristine nanoparticles were used. It is possible to precisely control the temperature, heating process, and holding time of the sample while taking advantage of the characteristics of microwave heating such as rapid and volume heating. This microwave carbothermal reduction method is thus highly effective in controlling the size of the synthesized Ti₄O₇ particles.