What can Raman images reveal

Raman imaging can display the distribution of different chemical and structural substances in a sample. Understand how to collect and analyze Raman images.
Raman imaging displays the spatial distribution of spectral information in the sample. We use a micro Raman spectrometer to collect spectral information from a point array on or within the sample. Raman imaging technology can easily reveal changes in the chemical and structural properties of substances in one-dimensional line contours, two-dimensional surface areas, or three-dimensional volumes.
What can Raman imaging tell you
By utilizing the spectral information of each pixel in the Raman image, we can determine:
Is there a substance or type of substance present
Is there an unknown substance present
Distribution of substances or types of substances
The size of any particle or region
Relative content of substance or substance type
Structural changes in substances, such as crystallinity or stress state
The thickness and composition of layered materials, such as polymer laminates (thickness ranging from micrometers to millimeters)
White light and Raman images of laundry detergent. Raman images can display regions with different chemical properties, which are not visible in white light images.
Qualitative and quantitative analysis of Raman imaging
Pseudo color Raman images can effectively highlight the distribution of chemical and structural properties within the sample. The brightness, contrast, and color of an image help to display the material composition. By stacking multiple Raman images, the distribution of multiple substance types or properties can be displayed simultaneously.
How do we collect Raman images?
We use a micro Raman spectrometer to collect Raman spectra from each position of the sample. Then, save all Raman spectra in a data file called a spectral hypercube. Finally, analyze the spectral hypercube to generate Raman images.
Here are several Raman imaging methods, such as:
Point focu

The micro Raman spectrometer focuses the laser onto a point on the sample. We place the sample on the automatic sample stage, which moves the sample under the laser. The spectrometer collects spectra from a point array on the sample.
The scanning mode that is faster than point focusing imaging is the imaging technique. Using StreamHR Rapide technology for Raman imaging, over 1000 spectral points can be collected per second.
Optimize sampling and oversampling
Line focus
Laser irradiates a line on the sample, not a point. By using this method, you can simultaneously collect spectra from multiple locations on the sample, saving time while also using higher total laser power without damaging the sample. Imaging technology has achieved this vision in a precise and innovative way. Through line focusing imaging, laser irradiates a vertical line on the sample instead of a point.
It is crucial to consider the potential adverse effects of oversampling during imaging. If the distance between the laser points or laser lines is smaller than the distance between the acquisition points, oversampling will occur. By using StreamLine technology and Slalom pattern, this problem has been solved.
StreamLine technology can achieve fast and smooth Raman imaging
The system scans the laser line on the sample along the y-axis. The CCD detector simultaneously collects data from multiple points on the sample.
The step size of the x-axis is equal to the width of the laser line. This method can completely cover the sampling area, but the speed is not the fastest.
Over sparse sampling without Slalom mode
The step size of the x-axis is greater than the width of the laser line. The laser line does not scan between pixels, so the spectrometer will not analyze certain areas of the sample. In this example, the system only collects data from approximately 20% of the sample area.
The combination of Streamline technology and Slalom mode can achieve fast and complete coverage of the sampling area
The laser line moves in a zigzag pattern and can scan the sample area between pixels. This makes the step size of the x-axis greater than the width of the laser line.
Using this method, the spectrometer collects data from 100% of the sample area at the fastest speed.
How to analyze Raman images?
We can analyze the Raman spectra obtained in imaging experiments to generate one-dimensional line contours, two-dimensional surface images, or three-dimensional rendering images. Raman images can display simple univariate band parameters, such as the intensity of Raman bands. You can also perform a complete multivariate analysis of the entire Raman spectrum in the spectral hypercube.
The software includes many data analysis options suitable for Raman imaging:
The intensity of a certain frequency in the spectrum
These images are generated quickly, but may be misleading as it is impossible to distinguish the signal intensity generated by the Raman spectral band of interest from the signal intensity associated with the broad fluorescence background.
Curve fitting parameters
For each spectrum in the Raman image, you can fit the theoretical curve for each Raman band. You can calculate Raman band parameters, such as Raman frequency shift, band width, or relative intensity. Raman images typically display changes in Raman frequency shift within the spectral band, which can indicate local stress. It can also display changes in Raman band width, which can indicate different degrees of crystallinity.
Multivariate parameters
Multivariate analysis is powerful because it utilizes information from the entire spectrum, rather than just a single Raman band parameter (such as the intensity of a certain frequency or a curve fitting band). This method typically generates higher quality Raman images with higher chemical specificity.
If you have reference spectra of the chemical components in the sample, you can easily create Raman images to display the distribution of each component. In this case, you can use compositional analysis methods such as direct classical least squares (DCLS) or non negative least squares (NNLS). You can use these component analysis methods to obtain concentration estimates.
If there is no reference spectrum, you can use unsupervised chemometric methods to create Raman images. This includes functions such as cluster analysis, principal component analysis (PCA), or software. We can use these stoichiometric methods without knowing the composition of the sample. These methods can analyze the systematic variance between Raman spectra to predict spectral components. Then Raman images can display the distribution of spectral components in the sample.
We usually use cluster analysis and PCA to analyze Raman images of biological tissues and cells. These powerful stoichiometric methods can detect structures in biological samples that typically do not contain pure biochemical substances.
The EmptyModeling function is a simple and easy-to-use multivariate curve resolution alternating least squares (MCR-ALS) analysis tool. This method can create Raman images of samples to identify unknown regions in pure components. This is very important for samples such as semiconductor wafers or drug formulations.
These advanced data analysis functions are included in the chemical stoichiometry package of WiRE software. You can use these methods directly without any programming required.