Origin of the Electric Property Change of a Single-Wall Carbon Nanotube Caused by Low-Energy Irradiation : Defects or Substrate Charging ?

Low-energy irradiation-induced conductivity decrease of a single-wall carbon nanotube has been well established experimentally. However, its origin is still controversial. Irradiation effects on suspended single-wall carbon nanotubes, which are much less affected by substrate charging effects, were studied to distinguish possible origins. The results indicate that the conductivity decrease is not caused by substrate charging, but by irradiation-induced defect formation. [DOI: 10.1380/ejssnt.2011.103]


I. INTRODUCTION
Single-wall carbon nanotubes (SWNTs) show excellent electric properties and therefore possess a powerful potential for future electronics.SWNTs are also chemically stable owing to the robustness of the graphene sheet.However, their characteristic physical properties disappear when they are irradiated by low-energy (where "lowenergy" means that the energy is much smaller than the knock-on threshold.)electrons or photons in a vacuum, as demonstrated by Raman [1,2] and photoluminescence (PL) [3] spectroscopy.The electric properties are also largely altered by the irradiation.A moderate irradiation can convert the room temperature electric properties of a metallic SWNT to semiconducting [4].An intensive irradiation finally makes a SWNT almost insulating [5][6][7][8].The decreased conductivity with the irradiation, as well as the degraded Raman and PL spectra, can recover reversibly by annealing [1,3,6] or applying a high bias to the SWNTs [7,8].
Although the irradiation-induced conductivity decreases in a SWNT and its recovery have been experimentally well established [5][6][7][8], the origin of the conductivity decrease is still controversial.We have reported that the irradiation-induced physical property changes are caused by defects created by the irradiation (low-energy irradiation damage) [6].The number of carbon atoms should be preserved during the defect formation in order to account for the fact that the defect recovers reversibly.Although such damage is not known for graphite, a similar reversible structural change is known to occur in C 60 with electron or photon irradiation [9].The electric property changes can also be explained by the defect formation, assuming a local band gap opening induced by the defect formation [4].On the other hand, there has been an attempt to explain the electric property changes by irradiationinduced charging of substrate (back-gate dielectric) just under the SWNT.In Refs.[7] and [8], the authors observed a large conductivity decrease by electron irradiation in a scanning electron microscope (SEM).They reversibly recovered the electric properties by applying a high bias voltage (∼10 V) to the SWNT.Although the phenomena they observed are very similar to ours, they ascribed the conductivity decrease to a local band gap opening caused by irradiation-induced charging of dielectric SiO 2 layer just under the SWNT.In fact, theoretical calculations predict that a uniform [10] or inhomogeneous [11] electric field can open a gap in a metallic SWNT of specific chiralities.In their model, the recovery is explained by a release of the trapped charges caused by the strong electric field in the vicinity of the SWNT [7,8].
The previous experiments were performed using conventional SWNT devices, in which the SWNTs were lying on a gate dielectric SiO 2 layer [5][6][7][8].In this geometry, the SWNT might be strongly affected by charging of the dielectric in the vicinity of it, considering that the electric field formed by a point charge is inversely proportional to the square of the distance.On the other hand, a suspended SWNT is expected to be much less affected by the substrate charging.In this work, in an attempt to distinguish the above two models, I studied the irradiation effects on suspended SWNTs.The results indicate that the conductivity decrease can not be due to substrate charging effects.

II. EXPERIMENTAL
Our device structure is similar to those described in Refs.[12] and [13].A heavily doped Si wafer with an oxide layer of 500 nm was used as a substrate.The source and drain electrodes were fabricated by electron beam lithography and deposition of a 300-nm-thick Pt/Ti thin film by radio frequency magnetron sputtering.The distance between the source and drain electrodes was 500 to 1000 nm.The substrate served as a back-gate electrode.The pad electrodes were fabricated by Pt/Ti deposition and photolithography.Then, a subnanometer-thick catalytic Co thin film was deposited on the top surface of the source electrode by using electron beam lithography and electron beam evaporation techniques.Finally, a suspended SWNT was grown between the source and drain electrodes by the chemical vapor deposition method with ethanol as a carbon source [Fig.1(a)].The height of the SWNT from the substrate surface determined by the height of the Pt electrodes was 300 nm.The device characteristics were first checked in a conventional probe station pumped by a turbo molecular pump.In-situ electric measurements during electron irradiation were performed in a Hitachi S-4300 cold cathode field emission SEM equipped with microprobes for electric measurements.All measurements were performed at room temperature.

III. RESULTS AND DISCUSSION
Gate characteristics of a semiconducting on-substrate SWNT and a suspended SWNT are shown in Figs.1(b) and (c).The SEM image shown in (a) was taken after electric measurements to avoid the irradiation-induced changes in the electric properties.As is well known, a large hysteresis is observed for the on-substrate device.The hysteresis is caused by charge traps on or in the dielectric [14].In good contrast to the on-substrate device, the suspended SWNT device shows almost no hysteresis, which is consistent with a previous report [14].The normally-on and nonmonotonic gate characteristics happened to be observed and are not common features.The former may suggest that the semiconducting transport characteristics are caused by natural defects in a metallic SWNT.Normally-on semiconducting properties have been observed for metallic SWNTs slightly damaged by Ar plasma treatment [15].The latter may be due to the formation of a tiny dot caused by defects in the SWNT channel [16].Anyway, forward and reverse sweeps are almost consistent with each other for suspended SWNTs.
In-situ electric measurements during electron irradiation were performed on metallic SWNT devices, which showed almost no gate voltage-dependence.Figure 2(a) shows a SEM image of a metallic SWNT device.The device consists of two SWNTs (A branch is seen between the electrodes).A result of in-situ electric measurement on the device during electron irradiation is shown in Fig. 2(b).The drain voltage was set to 0.1 V, and the substrate (back-gate) was grounded.The device was first irradiated by an electron beam of 1 keV with the beam current of 58 pA using the normal SEM observation mode.The irradiation dose rate was 1.2 × 10 13 cm −2 s −1 on average in the observation area.The SEM observation gradually decreased the conductivity.Then, at 42 s, the SWNTs were intensively irradiated by using the line scan mode.The local dose rate was estimated to be ∼ 6 × 10 16 cm −2 s −1 .This irradiation decreased the conductivity by two orders of magnitude in only a few seconds and, finally, almost insulating properties were observed.
The gate voltage characteristics of the device before and after the irradiation are shown in Fig. 2(c).The irradiation decreased the two-probe conductivity by four to five orders of magnitude in the whole gate voltage range.Considering that the initial resistivity of the device would be dominated by the contact resistance between the SWNTs and electrodes, the intrinsic conductivity decrease would be much larger.Results of in-situ electric measurement on another SWNT device [Fig.3(a)] are shown in Fig. 3(b).In this case, the electron kinetic energy and beam current were 20 keV and 202 pA, respectively.The irradiation dose rate during SEM observation was 4.1 × 10 13 to 3.7 × 10 14 cm −2 s −1 in the observation area.The rapid current changes during the SEM observation were due to specimen stage movements and SEM magnification changes.At 40 s, a line scan irradiation was started.The local dose rate was ∼ 6×10 17 cm −2 s −1 .The conductivity rapidly extinguished again.Almost insulating properties were finally observed in the whole gate voltage range, as shown in Fig. 3(c).
The results shown in Figs. 2 and 3 are consistent with each other.They are also consistent with previous studies using on-substrate SWNT devices [5][6][7][8].Theoretical calculations have shown that a high electric field of ∼1 V•nm −1 barely opens a band gap of several ten millielectron volts.It is very unlikely that such a high electric field is formed at SWNTs suspended 300 nm above the substrate.In fact, a simulation [11] has been performed under a condition where the gate voltage was applied from a metal tip located only 0.5 nm from a SWNT.Thus, the charging effect does not explain the irradiationinduced conductivity changes of suspended SWNTs.Similarly, it is also very difficult to explain the irradiationinduced degradation of Raman and PL spectra of suspended SWNTs [3] by the charging effect.The energy barrier height observed after irradiation is not consistent with the field-induced band gap opening, either.The calculations show that the maximum value of the fieldinduced band gap is at most ∼0.1 eV [10,11]  observed at room temperature.In fact, an energy barrier of ∼0.6 eV was observed for a SWNT whose room temperature electric properties were converted from metallic to semiconducting by irradiation [4].Considering that the irradiation-induced physical property changes can recover at a moderate temperature (∼300 • C [6]), the high biasinduced recovery observed in Refs.[7] and [8] seems to be due to current-induced heating.In fact, low-energy irradiation itself does not cut a SWNT or make it completely insulating.The drain current can still be observed even after intense irradiation, as shown in Figs.2(c) and 3(c).Thus, it is possible that damaged SWNTs are annealed by Joule heating.On the other hand, the defect formation hypothesis is consistent with the results of this study, as well as with those of previous Raman and PL studies [3] of suspended SWNTs.Moreover, Yamada et al. [17] recently reported that electron injection from a scanning electron microscope (STM) tip to a metallic SWNT caused defect formation and a local and large (0.7 eV) band gap opening.In this case, the SWNT was located on a metallic substrate having no charge traps.I can now unambiguously conclude that the irradiation-induced conductivity decrease experimentally observed is essentially due to defect formation in the SWNTs of a specific chirality.I do not deny an electric field-induced band gap opening [10,11] in a metallic SWNT.However, I do not think that such a high or inhomogeneous electric field is produced by simple SEM observation or line scans.

IV. CONCLUSIONS
Electron irradiation effects on the electric properties of suspended metallic SWNTs were measured.A drastic decrease of the conductivity was observed with the irradiation, indicating that the reduction is not due to the charging effect.I conclude that the decrease is caused by irradiation-induced defect formation.

FIG. 1 :
FIG. 1: (a) SEM image of a SWNT suspended between Pt electrodes.Scale bar: 500 nm.(b) Gate characteristics of an on-substrate SWNT device.(c) Gate characteristics of the suspendedSWNT shown in (a).

FIG. 2 :
FIG. 2: (a) SEM image of a suspended SWNT.Scale bar: 200 nm.(b) Drain current during SEM observation and line scans.(c) Gate characteristics of the device before and after the electron irradiation.

FIG. 3 :
FIG. 3: (a) SEM image of a suspended SWNT.Scale bar: 200 nm.(b) Drain current during SEM observation and line scans.(c) Gate characteristics of the device before and after the electron irradiation.