Abstract

Graphene has attracted enormous research interest due to its extraordinary material properties. Process control to achieve high-quality graphene is indispensable for graphene-based applications. This research investigates the effects of process parameters on graphene quality in a low-pressure chemical vapor deposition (LPCVD) graphene growth process. A fractional factorial design of experiment is conducted to provide understanding on not only the main effect of process parameters, but also the interaction effect among them. Graphene quality including the number of layers and grain size is analyzed. To achieve monolayer graphene with large grain size, a condition with low CH4–H2 ratio, short growth time, high growth pressure, high growth temperature, and slow cooling rate is recommended. This study considers a large set of process parameters with their interaction effects and provides guidelines to optimize graphene growth via LPCVD focusing on the number of graphene layers and the grain size.

1 Introduction

As a material with enormous potential, graphene has been fabricated through both physical and chemical processes, including mechanical exfoliation, liquid-phase exfoliation, graphene oxide reduction, epitaxial growth, and chemical vapor deposition (CVD). Among these methods, CVD on copper has become the prevalent method for large-scale monolayer and few-layer graphene growth because of its cost-effectiveness of both the growth substrate and carbon source material. Depending on the chamber pressure and the presence of plasma, chemical vapor deposition for graphene fabrication can be further divided into atmospheric CVD, plasma-enhanced CVD (PECVD), and low-pressure CVD (LPCVD). PECVD allows a lower growth temperature, because energetic electrons produced by the plasma help to enhance the ionization, excitation, and dissociation of hydrocarbon precursors at relatively low temperatures [1]. However, the graphene grown by atmospheric CVD and PECVD has relatively low quality compared with LPCVD, because the processes are easily affected by environmental conditions such as humidity and the ion bombardment effect. To date, LPCVD remains the most efficient and scalable method to produce high-quality and cost-effective graphene in large size.

Graphene growth in LPCVD is affected by both the growth substrate and process conditions. Different substrates yield different graphene quality and the number of graphene layers in the CVD process. Because of different catalytic activities, the growth mechanism of graphene on various substrates may be different. Many researchers have conducted graphene growth on different substrates, such as Cu [2], Ni [3], Pd [4], Ru [5], and Ir [6]. The graphene growth on these metal substrates largely follows two different routes. When the carbon solubility in the substrate is low, the graphene growth is basically through surface adsorption [7,8]. When the carbon solubility is high, the graphene growth goes through a series of activities, including dissolution, surface segregation, and precipitation. In an extremely well-controlled process, the graphene growth could be limited to one route; however, both routes could be in play in a CVD graphene growth process. By carefully controlling the process parameters, monolayer as well as few-layer graphene could be obtained depending on the growth time [9].

For a given substrate and carbon source, five key process parameters have been identified in the past in the LPCVD graphene growth process to achieve high graphene quality. These parameters are methane/hydrogen (CH4/H2) ratio, time, pressure, temperature, and cooling rate. It has been shown that these parameters could significantly influence the graphene nucleation density, graphene grain size, and graphene grain shape. Table 1 summarizes the effects of these parameters from previous studies. Zhang et al. [10] has found that the shape of graphene grains changes with the increasing CH4/H2 ratio. For example, the graphene grains change from a hexagonal shape at CH4/H2 = 1:30 to mostly a four-lobe structure at CH4/H2 = 1:20. This study also showed that the morphology of graphene grains would change with increasing growth pressure. Graphene grains changed from small irregular flakes at 80 mTorr to mostly four-lobed grains at 100 mTorr, and eventually to almost continuous film when the pressure was increased to 400 mTorr. Regmi et al. [11] showed that the growth time had a relatively high impact on graphene grain size, but low impact on nucleation density and grain shape. After nucleation occurred, the graphene grains continued to grow laterally as growth time increased. The graphene grain size increased until the grain edges merged with each other. Monolayer graphene covering more than 95% of the surface was obtained. Double-layer (2 L) and multilayer graphene flakes appeared before the monolayer was completed. The 2 L and multilayer flakes seemed to appear at the initial nucleation sites. Vlassiouk et al. [12] found that growth temperature had high impact on nucleation density and graphene grain size, but relatively low influence on the grain shape. The graphene nucleation density decreased dramatically with increasing temperature, while the graphene grain size increased with increasing temperature. Choi et al. [13] showed that the cooling rate had a strong impact on nucleation density, but a weaker effect on graphene grain size and grain shape.

Table 1

Factors affecting LPCVD-grown graphene

FactorNucleation density effectGrain size effectGrain shape effectNotesReferences
CH4/H2 ratioMiddleHighHigh1:30–small and close to hexagonal shape[10]
1:20–four-lobe structures
1:15–1:12.5—six-lobe flowers
1:10–irregular flakes
1:5–1:2—quasi-continuous graphene film
Growth timeLowHighLowTime↑: graphene grain size↑[11]
Growth pressureMiddleHighHigh80 mTorr: irregular small flakes[10,11]
100 mTorr: four-lobe grains
125 mTorr: four-lobe and six-lobe flowers
150–300 mTorr: six-lobe flowers
400 mTorr: quasi-continuous graphene film
Growth temperatureHighHighLowTemp.↑: graphene nucleation density↓ and graphene grain size↑[12]
Cooling rateHighMiddleLowCooling rate↑: graphene nucleation density↓ and graphene grain size↑[13]
FactorNucleation density effectGrain size effectGrain shape effectNotesReferences
CH4/H2 ratioMiddleHighHigh1:30–small and close to hexagonal shape[10]
1:20–four-lobe structures
1:15–1:12.5—six-lobe flowers
1:10–irregular flakes
1:5–1:2—quasi-continuous graphene film
Growth timeLowHighLowTime↑: graphene grain size↑[11]
Growth pressureMiddleHighHigh80 mTorr: irregular small flakes[10,11]
100 mTorr: four-lobe grains
125 mTorr: four-lobe and six-lobe flowers
150–300 mTorr: six-lobe flowers
400 mTorr: quasi-continuous graphene film
Growth temperatureHighHighLowTemp.↑: graphene nucleation density↓ and graphene grain size↑[12]
Cooling rateHighMiddleLowCooling rate↑: graphene nucleation density↓ and graphene grain size↑[13]

Although important process parameters of LPCVD graphene have been identified, most of the previous studies have focused on individual and independent effects of these process parameters ranging from CH4/H2 ratio to cooling rate one at a time. This approach does not provide an estimation of interaction effects among the input factors. In order to produce graphene with high quality, a statistical analysis through a design of experiment approach is required [14,15]. Shanmugam et al. [16] conducted a design of experiment study on CVD graphene growth; however, only three process parameters, carbon source, growth temperature, and mass flow rate of the carbon source, were included in the study.

In this study, we conduct a design of experiment study on LPCVD graphene growth with all the five major process parameters that have been identified in literature. A fractional factorial design was selected to minimize the experimental resource and time required for the study. A statistical analysis on both the main effects and second-order interaction effects of experiment was conducted to reveal the interplay of process parameters. The study considers a larger set of process parameters including their interaction effects than existing studies. It provides guidelines to control the quality of graphene grown via LPCVD focusing on the number of graphene layers and the graphene grain size.

2 Experimental

2.1 Experimental Setup.

The experiment was conducted with a LPCVD system, as shown in Fig. 1. The LPCVD system consisted of flow controllers, a quartz tube of one-inch diameter, a high-temperature tube furnace, a moving platform, a pressure gage close to the gas outlet, a computer for gas and motor control, and a vacuum pump to draw vacuum in the quartz tube. The furnace was mounted on a moving platform, such that the heating zone of the furnace could be moved away rapidly from the sample to control the cooling rate of the growth process. The moving platform was assembled with four V-type rail rollers and a gear rack under an aluminum plate. Two V-type rails were located under the four rollers. A stepper motor with a spur gear was connected to a motor drive controlled by an Arduino microcontroller and a DC power supply. The amount of methane and hydrogen were controlled through the flow controllers connected to the computer.

Fig. 1
LPCVD experimental setup with a moving platform
Fig. 1
LPCVD experimental setup with a moving platform

Commercially available copper foils with 125 μm thickness were used as the substrate for graphene growth. The copper foil was cut into 1 × 2 in.2 samples and chemically treated in a 0.5 M ammonium persulfate solution for 20 min to remove the native oxide layer and other contaminations that existed on the copper foil. The copper foil was sequentially cleaned with acetone, isopropyl alcohol, and de-ionized water. The treated copper foil was placed in the middle of the one-inch quartz tube. Graphene growth consisted of the following steps: (1) The pressure in the quartz tube was pumped down to 5 mTorr; (2) the temperature of the furnace was ramped up to 1030 °C with H2 flow at 10 sccm in the quartz tube for 50 min in order to further remove the oxide layer of the copper and to recrystallize the copper grains; (3) the required growth temperature was set and maintained for 10 min to reach a temperature steady-state; (4) the growth pressures was set by controlling the down flow valve of the LPCVD system under a constant H2 flow (10 sccm) environment; (5) one minute after the growth pressure was achieved, the carbon precursor CH4 was introduced under a constant H2 flow (10 sccm) rate; (6) after a controlled graphene growth time, the sample is allowed to cool down to ambient temperature at a controlled cooling rate by moving the furnace away from the sample; and (7) the flow of the gases and the furnace were turned off at 300 °C of the copper foil temperature.

2.2 Design of Experiment.

The process parameters of the LPCVD process are illustrated with Fig. 2. The overall LPCVD process can be largely divided into four phases: ramping, annealing, growth, and cooling. In the ramping phase, the temperature sharply increases from room temperature to 1030 °C within 20 min at 140 mTorr under 10 sccm H2. In the annealing phase, the temperature is maintained at 1030 °C for 40 min at 140 mTorr under 10 sccm H2. In the growth phase, the temperature is set at 1030 °C for 10 min, but the pressure is varied from 300 mTorr to 1000 mTorr under a constant H2 flow (10 sccm) and different CH4 flows (1–10 sccm). In the cooling phase, the temperature decreases from 1030 °C to room temperature in 50 min with the corresponding pressure in the growth phase under the constant H2 flow (10 sccm) and different CH4 flows (1–10 sccm).

Fig. 2
LPCVD process parameters for graphene growth
Fig. 2
LPCVD process parameters for graphene growth

Table 2 shows the experimental factors and their ranges selected for this study. Growth time is defined as the time from the start of CH4 flow to the start of the cooling process. Growth pressure is defined as the pressure at the start of CH4 flow and can be manually controlled by the gas valve. Growth temperature is the highest temperature in the LPCVD growth period. Cooling rate is achieved by controlling the speed of the moving platform. The maximum flow rates of CH4 and H2 were limited to 10 sccm due to equipment capability. Other factors including ramp time, annealing time, annealing temperature, and the H2 flow rate of 10 sccm were all fixed.

Table 2

Factors and their ranges for the design of experiment

Range
FactorsLevelMinimumMaximumRemarks
A: CH4/H2 ratio21:1010:1010 sccm: max. value of valve
H2—10 sccm fixed
B: growth time25 min30 minBetween the start of methane (CH4) flow and cooling
C: growth pressure2300 mTorr1000 mTorrManual control at the start of methane (CH4) flow
E: growth temperature2950 °C1030 °CCu—low carbon solubility
CH4—high activation energy
F: cooling rate20.5 °C/s5.0 °C/sMovable furnace
Range
FactorsLevelMinimumMaximumRemarks
A: CH4/H2 ratio21:1010:1010 sccm: max. value of valve
H2—10 sccm fixed
B: growth time25 min30 minBetween the start of methane (CH4) flow and cooling
C: growth pressure2300 mTorr1000 mTorrManual control at the start of methane (CH4) flow
E: growth temperature2950 °C1030 °CCu—low carbon solubility
CH4—high activation energy
F: cooling rate20.5 °C/s5.0 °C/sMovable furnace

A two-level fractional factorial design of experiment is conducted in this study. The fractional factorial design has the tradeoff between the number of runs and the resolution of the design. A 25−1 resolution V matrix (16 run) was chosen with I = ABCDE. Each main effect is aliased with a four-factor interaction effect (i.e., A = BCDE, B = ACDE, C = ABDE, D = ABCE and E = ABCD), and each two-factor interaction effect is aliased with a three-factor interaction effect (i.e., AB = CDE, AC = BDE, AD = BCE, AE = BCD, BC = ADE, BD = ACE, BE = ACD, CD = ABE, CE = ABD, and DE = ACB).

2.3 Sample Characterization and Data Analysis.

The samples were characterized with Raman spectroscopy and scanning electron microscopy (SEM). Raman spectroscopy (WITec Alpha 300 micro-Raman confocal microscope, λ = 488 nm) is used as a common tool to determine graphene defects using the ID/IG ratio and the number of graphene layers with the I2D/IG ratio [1719], where I2D/IG is the ratio of the intensities of the 2D to G peak in the Raman spectrum of the graphene grown on copper foil. The relative intensity of the I2D/IG ratio increases with decreasing layer numbers. SEM (FEI Quanta 650, Hillsboro, OR) is used to characterize the number of graphene grains on the copper foil [20]. Statistical analysis software jmppro 14 was used to analyze the experimental data.

3 Results and Discussion

3.1 Graphene Growth Results.

Figure 3 shows the Raman spectroscopy indicating defect (D), graphene (G), and two-dimensional (2D) peaks, 30 × 30 μm2 SEM, and 10 × 10 μm2 scale I2D/IG Raman mapping images of a typical sample (DoE case 2). The Raman spectroscopy shows monolayer graphene with few defects based on the presence of G peak, high 2D peak, and very low D peak intensities. The SEM image shows several wrinkles and graphene grain boundaries. The white dots in the SEM image are due to silicon contaminations in the bulk copper foil when they segregated to the surface. In the I2D/IG Raman mapping image (Fig. 3(c)), the bright yellow regions indicate monolayer graphene and the dark brown regions indicate multilayer graphene, while the black regions indicate the presence of voids. An averaged I2D/IG ratio was calculated from 900 measurement points on the Raman mapping image. An I2D/IG ratio value over two was used to detect the presence of monolayer graphene [21]. For this particular sample, a value of 2.6 was obtained, indicating almost the entire copper surface was covered with monolayer graphene [22].

Fig. 3
Results from a typical LPCVD sample (DoE—case 2): (a) Raman spectroscopy indicating defect (D), graphene (G), and 2D peaks, (b) 30 × 30 μm2 SEM, (c) 10 × 10 μm2 scale I2D/IG Raman mapping images, and (d) I2D/IG scale bar
Fig. 3
Results from a typical LPCVD sample (DoE—case 2): (a) Raman spectroscopy indicating defect (D), graphene (G), and 2D peaks, (b) 30 × 30 μm2 SEM, (c) 10 × 10 μm2 scale I2D/IG Raman mapping images, and (d) I2D/IG scale bar

3.2 Statistical Analysis.

Table 3 shows the results from total 19 cases, including 16 cases from the fractional factorial design of experiment and three trial cases (cases 17–19). The ID/IG ratio, I2D/IG ratio, and the number of graphene grains within the same 30 × 30 μm2 area were determined as response variables to examine the graphene quality in terms of graphene defect, the number of graphene layers, and graphene grain size, respectively.

Table 3

Results from the experimental study

CaseCH4/H2 ratioGrowth time (min)Growth pressure (mTorr)Growth temperature (°C)Cooling rate (° C/s)ID/IG ratioI2D/IG ratio# of grains
10.153009505.00.041.32180
20.1530010300.50.042.63343
30.1510009500.50.032.43145
40.15100010305.00.042.42288
50.1303009500.50.041.18636
60.13030010305.00.041.93717
70.13010009505.00.031.78288
80.130100010300.50.042.01229
91.053009500.50.041.59439
101.0530010305.00.042.48279
111.0510009505.00.042.55370
121.05100010300.50.041.63454
131.0303009505.00.031.37555
141.03030010300.50.052.16238
151.03010009500.50.042.09328
161.030100010305.00.042.40395
170.1530010305.00.042.10525
180.153009505.00.061.52235
190.15100010305.00.032.11284
CaseCH4/H2 ratioGrowth time (min)Growth pressure (mTorr)Growth temperature (°C)Cooling rate (° C/s)ID/IG ratioI2D/IG ratio# of grains
10.153009505.00.041.32180
20.1530010300.50.042.63343
30.1510009500.50.032.43145
40.15100010305.00.042.42288
50.1303009500.50.041.18636
60.13030010305.00.041.93717
70.13010009505.00.031.78288
80.130100010300.50.042.01229
91.053009500.50.041.59439
101.0530010305.00.042.48279
111.0510009505.00.042.55370
121.05100010300.50.041.63454
131.0303009505.00.031.37555
141.03030010300.50.052.16238
151.03010009500.50.042.09328
161.030100010305.00.042.40395
170.1530010305.00.042.10525
180.153009505.00.061.52235
190.15100010305.00.032.11284

An analysis of variance study was conducted to detect the goodness of fit of a statistical model. The coefficient of determination (R2), adjusted coefficient of determination (R2adj), and probability (P) of the model are shown in Table 4. With low R2 and R2adj values and a probability over 0.05, it shows that the model does not fit the ID/IG data well, which indicates that the tested process conditions do not affect the amount of graphene defects. In fact, all the samples obtained in this experimental study had good coverage and little defects. On the other hand, the I2D/IG ratio and the number of grains had a good fit with the experimental data, with high R2 and R2adj values and a P-value under 0.05. This means that the number of graphene layers and number of graphene grains can be optimized through the adjustment of processing conditions. Therefore, further statistical analyses were performed on the number of graphene layers and the number of graphene grains.

Table 4

The goodness of fitness of the response models

ResponseR2R2adjP
ID/IG ratio0.170.120.0803
I2D/IG ratio0.820.750.0002
# of grains0.890.720.0189
ResponseR2R2adjP
ID/IG ratio0.170.120.0803
I2D/IG ratio0.820.750.0002
# of grains0.890.720.0189

3.2.1 Effects on Number of Graphene Layers.

Process parameter effects on graphene layers are shown in Fig. 4. As shown in Fig. 4(a), in the range studied, a higher growth time was favorable for the graphene growth with a higher number of layers (I2D/IG ratio value between 1 and 2). Similarly, a lower pressure and a lower temperature were favorable for the graphene growth with a higher number of layers. CH4/H2 ratio and cooling rate are not statistically significant for the I2D/IG ratio, as confirmed by the parameter estimates in Fig. 4(c). A higher growth time allows a longer carbon source reaction time, promoting the formation of more graphene layers on the copper surface.

Fig. 4
Statistical analysis results for I2D/IG ratio: (a) main effect plots, (b) interaction effect plots, and (c) parameter estimates. Colors and asterisks indicate statistically significant parameters, orange and double asterisks for 1% level and red and single asterisk for 5% level.
Fig. 4
Statistical analysis results for I2D/IG ratio: (a) main effect plots, (b) interaction effect plots, and (c) parameter estimates. Colors and asterisks indicate statistically significant parameters, orange and double asterisks for 1% level and red and single asterisk for 5% level.

The interaction plots shown in Fig. 4(b) indicate strong interaction effects between growth pressure and growth temperature and between the CH4/H2 ratio and cooling rate. When the growth temperature was low, multilayer graphene was obtained with a low growth pressure, whereas monolayer graphene was obtained at a high growth pressure. On the other hand, when the growth temperature was high, both low and high growth pressures resulted in monolayer graphene. This phenomenon is consistent with the general CVD kinetics, which divides the growth rate into two regions, mass transport limited and surface-reaction limited, based on the growth temperature and pressure. The CVD growth of graphene on copper in a tube furnace basically involves two processes: (1) the mass transport process where diffusion of the carbon species pass through a boundary layer to reach the substrate surface and (2) the surface reaction process of the active carbon species forming the graphene lattice [23]. At high temperatures, there is a stable growth region for both low and high pressures, because the surface reaction rate is fast and the carbon species can be converted to graphene efficiently. At low temperatures, there may be a stable graphene growth region with high growth pressure due to the limiting mass transport process; however, with low growth pressure where more carbon species could diffuse through the boundary layer, the growth is in an unstable region, resulting in the growth of additional graphene layers. This phenomenon was also observed in previous studies [12,23,24].

The interaction plot between CH4/H2 ratio and cooling rate for the I2D/IG ratio response shows that at high CH4 flow (10 sccm) with slow cooling rate (0.5 °C/s), the I2D/IG ratio decreases, indicating the formation of graphene with multiple layers. This is because high CH4 flow and slow cooling rate increase the amount of carbon source and the reaction time, respectively, both leading to the growth of additional graphene layers.

3.2.2 Effects on Number of Graphene Grains.

The effects of process parameters on the number of graphene grains are shown in Fig. 5. As shown in Fig. 5(a), the graphene growth with a higher number of graphene grains is resulted from a longer growth time and a lower growth pressure. CH4/H2 ratio, growth temperature, and cooling rate are not statistically significant for the number of graphene grains, as confirmed by the parameter estimates in Fig. 5(c). The physical phenomena regarding the major factors such as growth time and growth pressure can be explained as follows. First, the nucleation density of graphene is increased by increasing growth time. This is because a longer growth time leads to an extended reaction time in which CH4 as the carbon supply in the CVD chamber is catalytically decomposed on the copper surface to create new nucleation sites of graphene. Therefore, a higher nucleation density of graphene with a longer growth time is expected [11]. Second, at a high pressure copper evaporation is suppressed, and the activation energy is assigned to the desorption energy of carbon species [12]. Therefore, a lower nucleation density of graphene is expected with a high growth pressure. Conversely, a low growth pressure leads to an increase of graphene grain nucleation.

Fig. 5
Statistical results for the number of graphene grains: (a) main effect plot, (b) interaction plot, and (c) parameter estimates. Asterisks indicate statistically significant parameters, double asterisks for 1% level and single asterisk for 5% level.
Fig. 5
Statistical results for the number of graphene grains: (a) main effect plot, (b) interaction plot, and (c) parameter estimates. Asterisks indicate statistically significant parameters, double asterisks for 1% level and single asterisk for 5% level.

The interaction plots highlighted in a box in the first row in Fig. 5(b) indicate that there are strong interaction effects among CH4/H2 ratio, growth time, growth pressure, and growth temperature on the number of graphene grains. It shows that with low CH4 flow (1 sccm), a higher number of graphene grains is obtained with a longer growth time (30 min) and a lower growth pressure (300 mTorr). This is because at a lower CH4 flow, the amount of available carbon species is limited, making the effects of reaction time and growth pressure more pronounced. At a high CH4 flow (10 sccm), a higher number of graphene grains are obtained with a lower growth temperature (950 °C). This is because the high carbon supply (CH4) is catalytically decomposed on the copper surface to create more nucleation sites of graphene [10]. Moreover, at a low temperature, the desorption of carbon species adsorbed on copper surface is suppressed [12]. Therefore, a high CH4/H2 ratio and low temperature are expected to increase the number of graphene grains. The interaction plot between CH4/H2 ratio and growth temperature also shows that at a low CH4 flow (1 sccm), a higher number of graphene grains are obtained with a higher growth temperature (1030 °C), because more reaction will happen, resulting in more free carbon species to form nuclei. The interaction plot highlighted in the box in the second row between growth time and growth pressure for the number of graphene grains shows that a long growth time (30 min) and a low pressure (300 mTorr) lead to a higher number of graphene grains. The interaction plot between growth time and cooling rate (also shown in a red box) shows that a long growth time (30 min) and a fast cooling rate (5 °C/s) yield a higher number of graphene grains. This is because that copper would form small grains due to a fast cooling rate, leading to a large number of smaller graphene grains. The interaction plot highlighted in the fourth row between cooling rate and growth temperature shows that a low growth temperature (950 °C) and a slow cooling rate (0.5 °C/s) or a high growth temperature (1030 °C) and a fast cooling rate (5 °C/s) would lead to a high number of graphene grains.

Table 5 provides a summary of the effect of each factor on the above two response variables. To obtain monolayer graphene with large grain size and fewer graphene grain boundaries, it is desirable to have fewer graphene grains and a higher I2D/IG ratio (i.e., # of grains↓ and I2D/IG↑). Based on the significance of the major factors and interaction effects, a favorable condition with low CH4/H2 ratio, short growth time, high growth pressure, high growth temperature, and slow cooling rate can be suggested to obtain monolayer graphene with large grain size.

Table 5

The direction and effect level of each factor to obtain monolayer graphene with large grain size ( and : strong effects, and : intermediate effects)

FactorsMonolayer graphene (I2D/IG↑)Large grain size (# of grains↓)
CH4/H2 ratio
Growth time
Growth pressure
Growth temperature
Cooling rate
FactorsMonolayer graphene (I2D/IG↑)Large grain size (# of grains↓)
CH4/H2 ratio
Growth time
Growth pressure
Growth temperature
Cooling rate

4 Conclusion

Graphene quality in a LPCVD process is investigated in this study through a fractional factorial design of experiment method. Five key process parameters are identified as CH4/H2 ratio, growth time, growth pressure, growth temperature, and cooling rate. The graphene quality is examined as not only the number of graphene layers through the I2D/IG ratio, but also the number of graphene grains through SEM images. Statistical analysis reveals the interaction effect as well as the major and minor factors for each response. For the I2D/IG ratio response, the major factors are growth time, growth pressure, and growth temperature. There are strong interaction effects among growth pressure, growth temperature, CH4/H2 ratio, and cooling rate. For number of graphene grains, the major factors are growth time and growth pressure. The interaction plots show that there are interaction effects among CH4/H2 ratio, growth time, growth pressure, growth temperature, and cooling rate. For monolayer graphene growth with larger grain size, conditions with low CH4/H2 ratio, short growth time, high growth pressure, high growth temperature, and slow cooling rate are generally favored. This study examines a large set of process parameters of LPCVD graphene growth and reveals the interaction effects among the process parameters. It provides guidelines to control the quality of graphene for large-scale fabrication.

Acknowledgment

Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.

Funding Data

  • National Science Foundation under Cooperative (Agreement No. EEC-1160494; Funder ID: 10.13039/100000001).

References

1.
Li
,
M.
,
Liu
,
D.
,
Wei
,
D.
,
Song
,
X.
,
Wei
,
D.
, and
Wee
,
A. T. S.
,
2016
, “
Controllable Synthesis of Graphene by Plasma‐Enhanced Chemical Vapor Deposition and Its Related Applications
,”
Adv. Sci.
,
3
(
11
), p.
1600003
.10.1002/advs.201600003
2.
Li
,
X.
,
Cai
,
W.
,
An
,
J.
,
Kim
,
S.
,
Nah
,
J.
,
Yang
,
D.
,
Piner
,
R.
,
Velamakanni
,
A.
,
Jung
,
I.
,
Tutuc
,
E.
,
Banerjee
,
S. K.
,
Colombo
,
L.
, and
Ruoff
,
R. S.
,
2009
, “
Large-Area Synthesis of High-Quality and Uniform Graphene Films on Copper Foils
,”
Science
,
324
(
5932
), pp.
1312
1314
.10.1126/science.1171245
3.
Kim
,
K. S.
,
Zhao
,
Y.
,
Jang
,
H.
,
Lee
,
S. Y.
,
Kim
,
J. M.
,
Kim
,
K. S.
,
Ahn
,
J.-H.
,
Kim
,
P.
,
Choi
,
J.-Y.
, and
Hong
,
B. H.
,
2009
, “
Large-Scale Pattern Growth of Graphene Films for Stretchable Transparent Electrodes
,”
Nature
,
457
(
7230
), pp.
706
710
.10.1038/nature07719
4.
Kwon
,
S.-Y.
,
Ciobanu
,
C. V.
,
Petrova
,
V.
,
Shenoy
,
V. B.
,
Bareño
,
J.
,
Gambin
,
V.
,
Petrov
,
I.
, and
Kodambaka
,
S.
,
2009
, “
Growth of Semiconducting Graphene on Palladium
,”
Nano Lett.
,
9
(
12
), pp.
3985
3990
.10.1021/nl902140j
5.
Sutter
,
P. W.
,
Flege
,
J.-I.
, and
Sutter
,
E. A.
,
2008
, “
Epitaxial Graphene on Ruthenium
,”
Nat. Mater.
,
7
(
5
), pp.
406
411
.10.1038/nmat2166
6.
N'Diaye
,
A. T.
,
Coraux
,
J.
,
Plasa
,
T. N.
,
Busse
,
C.
, and
Michely
,
T.
,
2008
, “
Structure of Epitaxial Graphene on Ir (111)
,”
New J. Phys.
,
10
(
4
), p.
043033
.10.1088/1367-2630/10/4/043033
7.
Wang
,
C.
,
Vinodgopal
,
K.
, and
Dai
,
G.-P.
,
2018
, “
Large-Area Synthesis and Growth Mechanism of Graphene by Chemical Vapor Deposition
,”
Chemical Vapor Deposition for Nanotechnology
,
IntechOpen Limited
,
London, UK
, p.
18
.
8.
Li
,
X.
,
Cai
,
W.
,
Colombo
,
L.
, and
Ruoff
,
R. S.
,
2009
, “
Evolution of Graphene Growth on Ni and Cu by Carbon Isotope Labeling
,”
Nano Lett
,
9
(
12
), pp.
4268
4272
.10.1021/nl902515k
9.
Zhang
,
Y.
,
Gomez
,
L.
,
Ishikawa
,
F. N.
,
Madaria
,
A.
,
Ryu
,
K.
,
Wang
,
C.
,
Badmaev
,
A.
, and
Zhou
,
C.
,
2010
, “
Comparison of Graphene Growth on Single-Crystalline and Polycrystalline Ni by Chemical Vapor Deposition
,”
J. Phys. Chem. Lett.
,
1
(
20
), pp.
3101
3107
.10.1021/jz1011466
10.
Zhang
,
Y.
,
Zhang
,
L.
,
Kim
,
P.
,
Ge
,
M.
,
Li
,
Z.
, and
Zhou
,
C.
,
2012
, “
Vapor Trapping Growth of Single-Crystalline Graphene Flowers: Synthesis, Morphology, and Electronic Properties
,”
Nano Lett.
,
12
(
6
), pp.
2810
2816
.10.1021/nl300039a
11.
Regmi
,
M.
,
Chisholm
,
M. F.
, and
Eres
,
G.
,
2012
, “
The Effect of Growth Parameters on the Intrinsic Properties of Large-Area Single Layer Graphene Grown by Chemical Vapor Deposition on Cu
,”
Carbon
,
50
(
1
), pp.
134
141
.10.1016/j.carbon.2011.07.063
12.
Vlassiouk
,
I.
,
Smirnov
,
S.
,
Regmi
,
M.
,
Surwade
,
S. P.
,
Srivastava
,
N.
,
Feenstra
,
R.
,
Eres
,
G.
,
Parish
,
C.
,
Lavrik
,
N.
,
Datskos
,
P.
,
Dai
,
S.
, and
Fulvio
,
P.
,
2013
, “
Graphene Nucleation Density on Copper: Fundamental Role of Background Pressure
,”
J. Phys. Chem. C
,
117
(
37
), pp.
18919
18926
.10.1021/jp4047648
13.
Choi
,
D. S.
,
Kim
,
K. S.
,
Kim
,
H.
,
Kim
,
Y.
,
Kim
,
T.
,
Rhy
,
S-h.
,
Yang
,
C.-M.
,
Yoon
,
D. H.
, and
Yang
,
W. S.
,
2014
, “
Effect of Cooling Condition on Chemical Vapor Deposition Synthesis of Graphene on Copper Catalyst
,”
ACS Appl. Mater. Interfaces
,
6
(
22
), pp.
19574
19578
.10.1021/am503698h
14.
Wirtz
,
C.
,
Lee
,
K.
,
Hallam
,
T.
, and
Duesberg
,
G. S.
,
2014
, “
Growth Optimisation of High Quality Graphene From Ethene at Low Temperatures
,”
Chem. Phys. Lett.
,
595–596
, pp.
192
196
.10.1016/j.cplett.2014.02.003
15.
Narula
,
U.
, and
Tan
,
C. M.
,
2016
, “
Determining the Parameters of Importance of a Graphene Synthesis Process Using Design-of-Experiments Method
,”
Appl. Sci.
,
6
(
7
), p.
204
.10.3390/app6070204
16.
Shanmugam
,
R.
,
Rangarajan
,
M.
,
Devanathan
,
S.
,
Sathe
,
V. G.
,
Senthilkumar
,
R.
, and
Kothurkar
,
N. K.
,
2016
, “
A Design of Experiments Investigation of the Effects of Synthesis Conditions on the Quality of CVD Graphene
,”
Mater. Res. Exp.
,
3
(
12
), p.
125601
.10.1088/2053-1591/3/12/125601
17.
Ferrari
,
A. C.
,
2007
, “
Raman Spectroscopy of Graphene and Graphite: Disorder, Electron–Phonon Coupling, Doping and Nonadiabatic Effects
,”
Solid State Commun.
,
143
(
1–2
), pp.
47
57
.10.1016/j.ssc.2007.03.052
18.
Ferrari
,
A. C.
, and
Basko
,
D. M.
,
2013
, “
Raman Spectroscopy as a Versatile Tool for Studying the Properties of Graphene
,”
Nat. Nanotechnol.
,
8
(
4
), pp.
235
246
.10.1038/nnano.2013.46
19.
Ferrari
,
A. C.
,
Meyer
,
J. C.
,
Scardaci
,
V.
,
Casiraghi
,
C.
,
Lazzeri
,
M.
,
Mauri
,
F.
,
Piscanec
,
S.
,
Jiang
,
D.
,
Novoselov
,
K. S.
,
Roth
,
S.
, and
Geim
,
A. K.
,
2006
, “
Raman Spectrum of Graphene and Graphene Layers
,”
Phys. Rev. Lett.
,
97
(
18
), p.
187401
.10.1103/PhysRevLett.97.187401
20.
Larsen
,
M. B. B. S.
,
2015
, “
Chemical Vapour Deposition of Large Area Graphene
,” Ph.D. thesis, Technical University of Denmark, Lyngby, Denmark.
21.
Cho
,
J. H.
,
Gorman
,
J. J.
,
Na
,
S. R.
, and
Cullinan
,
M.
,
2017
, “
Growth of Monolayer Graphene on Nanoscale Copper-Nickel Alloy Thin Films
,”
Carbon
,
115
, pp.
441
448
.10.1016/j.carbon.2017.01.023
22.
Lee
,
B.
, and
Li
,
W.
,
2020
, “
Performance of Different Layers of Graphene as Protective Coating for Copper Wire
,”
Mater. Lett.
,
273
, p.
127875
.10.1016/j.matlet.2020.127875
23.
Bhaviripudi
,
S.
,
Jia
,
X.
,
Dresselhaus
,
M. S.
, and
Kong
,
J.
,
2010
, “
Role of Kinetic Factors in Chemical Vapor Deposition Synthesis of Uniform Large Area Graphene Using Copper Catalyst
,”
Nano Lett.
,
10
(
10
), pp.
4128
4133
.10.1021/nl102355e
24.
Cho
,
J. H.
,
Na
,
S. R.
,
Park
,
S.
,
Akinwande
,
D.
,
Liechti
,
K. M.
, and
Cullinan
,
M. A.
,
2019
, “
Controlling the Number of Layers in Graphene Using the Growth Pressure
,”
Nanotechnology
,
30
(
23
), p.
235602
.10.1088/1361-6528/ab0847