Dear Arjun.

Considering the tip-sample system as a capacitor, the  contact potential difference (CPD) is the difference in work function Φ  between sample and tip:

VCPD = (Φsample-Φtip)/e  (eq. 1), where e is the elementary charge.

While the KPFM measurement results in the determination of the CPD, which

is the work function of the sample relative to that of the tip,the above equation  can be used to deduce the sample’s work function on an absolute scale. Using a calibrated tip with a known work function, the work function of the sample can be calculated from

the CPD measurement according to (eq. 1). However, for absolute work function measurements, operation under ultrahigh vacuum (UHV) conditions is mandatory, as it is well known that the work function is highly sensitive to the surface

cleanness.

But how do you calibrate the tip? The tip often has a  different work function with the reference samples.

Hi, Yuanjun.

I think, you could measure CPD for any well known material, for example, for a thick gold film, like a referee. An amorphous film would be better to avoid crystal orientation dependence.

Regards

Thank you so much. The thick and amorphous gold films is a good reference sample.

Hi,

Standard reference samples for calibration of the AFM tip in air is HOPG. The hydrophobic and inert surface of freshly cleaved HOPG is not covered by the adsorbed layer.

The equation should be VCPD = (Φsample-Φtip)/e? Or VCPD = (Φtip-Φsample)/e ?

I did KPFM of HOPG and in response I got a potential map with -6 to +6 mV scale.

How to calculate work function of tip from this map?

Robin, probably you have done something wrong, or it is error signal against bias voltage, you need a bias voltage of this map, and if you're measured it right, it will be equal to difference of work functions between sample and tip. If you know one of them, you can calculate other one. Polarity of voltage depends on what kind of AFM setup is used, bias can be applied either as voltage on sample, or voltage on tip with second electrode grounded and taken as reference.

KPFM require conductive tip, metal coatings can work up to 1V of bias, more strong coatings like TiN, W2C or DLC can work up to 10 V, it require some sort of reference electrode on sample, connected to right place of the setup, it require right height of second pass in two-pass methods and so on.

Hi Daniil,

I had used Ti/Pt tip with a tip biased of 1V and I am attaching potential image and corresponding height image, please look into this. I tell me if I had done something wrong

First of all, you have a bias value subtracted from both images (both has a mean-centered Z-scale). You need to disable all this postprocessing for Kelvin probe images.

You need a raw Z scale (disable all background subtraction and line processing) like at image below, and you need a bias voltage during the scan (see metadata image, Bias Voltage is what you need). Sum of voltage at image and bias voltage is contact potential difference between tip and sample point. At this example it is -0.43 V for background (gold surface), about -0.3 V on dark graphene flakes (single layer) and about -0.22 V for more charged (light) graphene flakes (two or three layers).

-0.43 V is than contact potential difference (or difference of two work functions) between gold surface on sample and tip material.

Example file taken from NTMDT scan gallery (link below) and processed with Gwyddion software.

http://www.ntmdt.com/scan-gallery/scan/graphene-flakes-on-gold-substrate

On HOPG I am getting avg potential of  +480 mV without flattening.

What is workfunction of my AFM tip ?

4.6 - 0.48

or

4.6 + 0.48

What AFM setup and what kind of AFM tips did you used for this experiment?

DC applied to tip and sample was grounded.

Tip is Ti coated on Ir.

Hi Robin,

The work function of your AFM tip is 5.08 eV.

Are your AFM and tip from Asylum Research? If the tip model is ASYELEC-01, we have a work function for the tip 4.9 eV.

Yes Tip is from Asylum and it is ASYELEC-02.

I have a doubt

because I am getting positive CPD so workfunction of sample should be more then the workfunction of Tip?

Seems that plus sign is more probable, but I'll rather refer to device documentation about DC voltage polarity.

Hi,

Positive CPD when applying bias to the probe means the work function of sample is less the work function of tip (see figure).

Thanks

Does CPD changes if go to lower scan area, like from 20 um to 1 um.

It should not, the voltage distribution on smaller scan should be the same as in chunk of larger scan. However, you can measure the Kelvin probe with something like 35 nm resolution at its best (it defined by tip sharpness, its form and distance between tip and sample while measuring), so 1 mkm scan will have to much sample points to be measured properly, and 20 mkm scan will have gaps between points if you use 512 data points per scan dimension. Also measurement noise should be affected by scan sampling.

i am also using the ambient kelvin proble to measure the work function of VO2 thin films. The CPD goes from positive to negative values with respect to the points. is this normal to have varying CPD (negative)?

Itani: Are values similar if you scan the same area twice?

Hi every one! 

When we use KPFM signal to calculate the work function of tip through HOPG, it is a map (distribution of surface potential). So, how can we calculate the workfunction of Tip? Is it calculated through the average value of surface potential? 

Dear Hoa Dao,

When we use KPFM to scan HOPG, we usually peel a new layer of HOPG for the calibration of the tip work function. This is because the HOPG layer might be contaminated, thus it will show a distribution of surface potential. However, if the surface potential measurement is not done under ultrahigh vacuum (UHV) condition, the quantitative measurement of the surface potential will be inaccurate, thus we will normally analyze the contact potential difference (CPD) of the measurement.

Best Regards

Fang Sheng 

Concerning the use of HOPG as 'calibration' sample: it is true, as Nadim wrote, thatHOPG is hydrophobic but I wouldn't be sure that no water layer is in between the tip and the sample, in ambient air (RH as high as 50 or 60%), also dut to the tip side (durface) of the interface (ie water squeezed in between anyway, at the meniscus).

Marco Salerno Nadim Davletkildeev Freshly cleaved HOPG is actually hydrophilic. Check : Effect of airborne contaminants on the wettability of supported graphene and graphite Article Effect of airborne contaminants on the wettability of suppor...

While calculating W.F. of the sample, should we take W.F. of the tip as -ve value? For e.g. if the work function of tip is 4.7 eV and CPD is 0.1 V, then W.F. of sample will be 4.7-0.1= 4.6 eV or -4.7-0.1=-4.8 eV?

Thanks

Urvashi Bothra It depends on how the UDC is connected. If DC voltage is applied to the sample, then W.F.sample=W.F.sample-CPD*e；if DC voltage is applied to the probe, then W.F.sample=W.F.sample+CPD*e，where e is the electronic charge.

Hao Zhang Hi, do you mean, if DC voltage is applied to probe then W.F. probe=W.F. sample+VCPD*e? Such that we take W.F. value positive.

Thanks

Urvashi Bothra Yes, I think so. You can read this article: Kelvin probe force microscopy and its application.

Thanks Hao Zhang

Urvashi Bothra sorry, I made an editing mistake. If DC voltage is applied to the probe, then W.F.sample=W.F.tip-CPD*e；if DC voltage is applied to the sample, then W.F.sample=W.F.tip+CPD*e. I hope you can refer to the following ref:

1) Two-dimensional surface band structure of operating light emitting devices (

J. Appl. Phys. 86, 107 (1999); doi: 10.1063/1.370706)

2) High-resolution work function imaging of single grains of semiconductor surfaces (Appl. Phys. Lett. 80, 2979 (2002); doi: 10.1063/1.1471375)

3) Kelvin probe force microscopy in ultra high vacuum using amplitude modulation detection of the electrostatic forces ( Applied Surface Science 157 2000 263–268)

4) High-sensitivity quantitative Kelvin probe microscopy by noncontact ultrahigh-vacuum atomic force microscopy (Appl. Phys. Lett. 75, 286 (1999); doi: 10.1063/1.124357)

Hao Zhang Thanks for sharing the relevant references.

Two questions related to determining absolute WF of sample:

1. How do you know how much voltage is applied to the probe or sample

2. what if the applied voltage is zero, only bias voltage (BV) to U-tip, is this the same probe voltage you are referring to.

The applied voltage is equal to the contact potential difference i.e. the difference between W.F. of tip and sample.

Thanks, I got it. However, I still have one more question which is about the sign:

is it: -eV_CPD = WF_tip - WF_s

or eV_CPD = WF_tip - WF_s ?

This negative sign is intriguing as I see both in the literature. I want to ensure which one is correct.

Sanju Gupta This is just the definition of V_CPD. I think it is all right, but we can only carry out the subsequent calculation in one form.

Yes, I got that also, BUT the negative sign on the right of the CPD can result in two numbers, a lower or higher. How do we know which is the correct to report as It is making me a question do I have lower CPD meaning n-type doping or if I calculate higher CPD because of the sign, is it because of MoO3, in MoS2, I see both in my data, but they are opposite to each other. How we interpret WF in such a scenario.

Sanju Gupta To solve your problem, we must know the relationship between Udc and Ucpd. Udc is obtained in the experiment, not Ucpd. We only know that |Udc|=|Ucpd|, then Udc=Ucpd or Udc=-Ucpd? This is a key. It depends on the following two questions:

1) The definition of Ucpd: -eV_CPD = WF_tip - WF_s or eV_CPD = WF_tip - WF_s ?

2) The DC voltage is applied to the sample or the tip?

① If -eV_CPD = WF_tip - WF_s and the DC voltage is applied to the sample, Udc=Ucpd. So, W.F.sample=W.F.tip+Udc*e.

② If -eV_CPD = WF_tip - WF_s and the DC voltage is applied to the tip, Udc=-Ucpd. So, W.F.sample=W.F.tip-Udc*e.

how mV measured in AFM converted to eV when using HOPG sample. What is the conversion method for other samples?

1 volt is actually equal to 1 eV based on the definition of the eV unit and is independent of the sample.

thank you sir...

Nadim Davletkildeev Hao Zhang Arjun Singh Sanju Gupta

Is it possible to calculate the work functions of nanoparticles such as HgTe, HgSe QDs, or nanolayers of Bi2Te3 film with respect to references like HOPG or Au?

I started to scan them at room temperature and got foe them some different values of CPD, now the question is will it be a more sensitive measurement to scan them in dry air like purging by Ar or N2 gas flow. Is it necessary to do the experiment in the UHV condition to calculate the WF of nanoparticles as QD & 2D layers?

Thanks for feed-back