comment发出去没多久，就收到了主编转来的作者回复。当然作者一再强调他的文章是没有任何问题的。之所以文的中的暗电流高，是因为他用Keithley测的，Keithley测出来噪音高；真正他拿来算的是用Lock-in测的暗电流。原文是这么说的：

“The I-V curves displayed in Figure 3a was measured using Keithley DC source meter which has high external noise, thus the dark current value is high.We did not use this to calculate the detectivities。。。

The use of a lock-in amplifier is necessary as it reduces external noise and records accurate dark current
values. According to Merlin lock-in amplifier the dark current values are 1.64 x 10-10 A/cm2 and 1.63 X 10-9
A/cm2 which justifies the displayed detectivities data in Fig 3b i.e D* ~ 3 x 10 13 jones at 350nm and
D* ~5.7 x 10 12 jones at 1000nm, respectively.”

  其中还附了一张图。如下：

  我一看这张图就被雷翻了。暗电流居然画成了波长的函数，而且计算的时候分别取为1.64 x 10-10 A/cm2 和 1.63 X 10-9
A/cm2 去算detectivity。基本的常识啊，大哥。暗电流不应该是个常数吗，暗电流跟波长有毛关系！而且大家都用Keithley测暗电流，用Lock-in测光电流是为了减除暗电流噪音的影响。于是我再回了一个rebuttal如下：

Dear Editor:

 

Thanks for the prompt reply to our inquiry. The authors' response to our comments provides some more details about the experiments and calculations involved in the reported study, which were missing in the original paper. Unfortunately, we have found that the authors have made some very fundamental mistakes therein, which could would seriously undermine the validity of the experiments and the conclusions drawn thereof. As such, we believe it is worth writing to you again to pinpiont those mistakes.

 

In the response, the author showed a figure that plots the dark current as a function of wavelength, the dark current at long wavelength range is 10 times greater than at short wavelength range. This is fundamentally wrong! Because the definition of dark current is: ‘the residual electric current flowing in a photoelectric device when there is no incident illumination’. Therefore, dark current is definitely independent of the illumination, it has nothing to do with the wavelength, it should be constant if the device is stable. Moreover, the author pointed out in response that “According to Merlin lock-in amplifier the dark current values are 1.64×10^-10 A/cm2 and 1.63×10^-9A/cm2 which justifies the displayed detectivities data in Fig 3b i.e D* ~ 3×10¹³ jones at 350nm andD* 5.7×10¹² jones at 1000nm, respectively.” Once again, the author took different values of dark current at different wavelength, which means he simply does not know the concept of dark current whatsoever.

 

The author also claimed that “The I-V curves displayed in Figure 3a was measured using Keithley DC source meter which has high external noise, thus the dark current value is high.” This is, again, a wrong claim. To the best of knowledge, many relevant laboratories worldwide, including Heeger’s group [1], Sargent’s group [2], and So’s group [3], are all using Keithley DC source to characterize the dark current of photodetectors. None of them have ever complained about the high dark current-measurement issues with Keithley DC sources. In contrast, the authors have reported to use Merlin Lock-in amplifier for dark current characterization BY MISTAKE. According to the manual of a lock-in amplifier,  “Lock-in amplifiers are used to detect and measure very small AC signals even when the small signal is obscured by noise sources many thousands of times larger”. Prof. Maxson in Leigh University [4] has pointed out that “the desired signal (detected by lock-in amplifier) can be several orders of magnitude less than the signal from myriad noise sources, such as stray light, dark current, or inherent device noise.” Therefore, the real dark current had been ruled out as a noise when it was measured with lock-in amplifier in the authors' study, and no wonder the author got an incorrect 'wavelength-dependent dark current’.

 

In summary, the authors had shown little knowledge about the concept of dark current in their photodetector study, which made many key conclusions of their study, such as the those in detector sensitivity, sensor detectivity and detection noises, seriously wrong.  We believe it is necessary for the authors to acknowledge those fundamental mistakes in public for the sake of rigorousity and also for the great reputation of Advanced Materials. The prompt action upon this issue by the editor will be greatly appreciated. . Thanks.

References:

[1]. X. Gong, M. Tong, Y. Xia, W. Cai, J. Moon, Y. Cao, G. Yu, C. Shieh, B. Nilsson, A. Heeger, Science, (2009), 325, 1665.

[2]. J. Clifford, G. Konstantatos, K. Johnston, S. Hoogland, L. Levina, E.H. Sargent, Nature Nanotechnology, (2009), 4, 40.

[3]. G. Sarasqueta, K Choudhury, J. Subbiah, F. So, Adv. Funct. Mater. (2011), 21, 167.

[4]. J. Maxson, Using a Lock-in Amplifier, link: http://www.lehigh.edu/~jph7/website/Physics262/aMaxsonLockIn.pdf

我以为这事到这就算结束了。没想到昨天还收到作者的回复了，而且主编转回复的时候说的那一句话也挺有意思的："Enclosed please find the response from the authors which I hope could get the issue closed up."。

未完待续。。。


转载本文请联系原作者获取授权，同时请注明本文来自蒋振宇科学网博客。
链接地址：https://m.sciencenet.cn/blog-250704-612877.html

上一篇：读文献发现其中的问题怎么办？(一)