Physical characteristics of inks under the electron microscope

[shadowblade]

Having had some spare time and access to a scanning electron microscope over the weekend, I thought it might be interesting to have a look at some ink samples under the microscope, to see if structural differences in the pigment could account for differences in ink characteristics between inks that otherwise use the same, or similar, colourants.

The inks I looked at were:

- Epson Ultrachrome HDR photo black, matte black, yellow and cyan
- Canon Lucia EX photo black, matte black, yellow and cyan
- HP Vivera photo black, matte black, yellow and cyan
- Piezography Carbon (black - the darkest one)
- MIS Eboni carbon

To do this, I placed a small droplet of each ink onto a slide (really a piece of aluminium) using a micropipette, then placed the sample into the microscope. No need to wait for the sample to dry, or to actually print something - I was interested in the structure of the pigment particles themselves, and the vacuum inside the microscope meant that the inks dried instantly (due to the near-zero vapour pressure in the vacuum). I didn't coat it in gold/platinum/palladium for added contrast, as is usual with SEM samples, as the 20-30nm thick metal layer would have made a significant difference in the visualised size of the pigment particles, but, seeing as I was mainly looking at particle size, the contrast was sufficient for comparison even without it.

Here's what I saw:
- Epson and Canon colour pigment particles are around the same size. HP's pigment particles are noticeably larger. Piezography particles seem to be about the same size as Epson/Canon, and significantly smaller than MIS Eboni. Matte black particles are noticeably larger than photo black particles (in all cases, they're reasonably huge).
- Although this wasn't really a test for pigment concentration, HP inks appeared to have more pigment load than the others. All the others appeared fairly similar, apart from the Epson yellow. I'm not sure if anyone's put the inks in a centrifuge to confirm this, though. Piezography Black appears to have a lot of pigment.

Based on what we already know the effect of particle size in pigments, these would certainly explain a few things that have already been observed.

- All else being equal (chemical composition, etc.) larger particles have greater longevity than smaller particles, given the greater number of redundant molecules per particle and the smaller surface area-to-mass ratio of the pigment particles. The larger particles in HP's Vivera ink would certainly help explain their greater longevity. Canon and Epson inks having similar particle sizes would also help explain their similar longevity, once the less-stable Epson yellow pigment is removed (e.g. via a RIP). As for Piezography, carbon particles are essentially inert anyway (doubly so after encapsulation) so they could afford to go with the smaller particles.

- It is also known that larger particles also result in less brilliance and gamut than smaller particles. This is because larger particles increase scatter and are also more opaque. This would certainly explain why Vivera inks seem to have slightly less gamut than Epson/Canon inks; I think HP may have helped compensate for this by using a greater pigment load in their ink. The one exception is black ink, where opacity increases Dmax, since, unlike with colours, you *don't* want light to be reflected off the substrate. The larger particle size probably helps explain HP's greater Dmax on matte papers.

- HP's larger particles may help explain why the surface of HP prints seem more fragile than Epson and Canon inks, prior to spraying with a protective coat. The larger pigment particles are probably easier to brush off.

- Given that the Epson yellow appears to have the same pigment load as the other Epson inks, I suspect that the rumour that Epson K3 and HDR yellow contains some dye (increasing saturation while reducing lightfastness) is just a myth. The test certainly doesn't exclude the presence of yellow dye, but, given that the pigment load appears to be similar to the other inks, I doubt it. The reduced longevity is probably just due to Epson having chosen a brighter, but more fugitive, yellow pigment for their ink.

- The smaller Piezography particles compared to the MIS Eboni particles probably explain why Piezography has a warmer tone than MIS carbon (which is more neutral), and also why MIS carbon inks can't print on glossy paper.

Not that it actually changes what we do in day-to-day printing, but I just thought it was interesting...

[Ernst Dinkla]

Having had some spare time and access to a scanning electron microscope over the weekend, I thought it might be interesting to have a look at some ink samples under the microscope, to see if structural differences in the pigment could account for differences in ink characteristics between inks that otherwise use the same, or similar, colourants.

The inks I looked at were:

- Epson Ultrachrome HDR photo black, matte black, yellow and cyan
- Canon Lucia EX photo black, matte black, yellow and cyan
- HP Vivera photo black, matte black, yellow and cyan
- Piezography Carbon (black - the darkest one)
- MIS Eboni carbon

To do this, I placed a small droplet of each ink onto a slide (really a piece of aluminium) using a micropipette, then placed the sample into the microscope. No need to wait for the sample to dry, or to actually print something - I was interested in the structure of the pigment particles themselves, and the vacuum inside the microscope meant that the inks dried instantly (due to the near-zero vapour pressure in the vacuum). I didn't coat it in gold/platinum/palladium for added contrast, as is usual with SEM samples, as the 20-30nm thick metal layer would have made a significant difference in the visualised size of the pigment particles, but, seeing as I was mainly looking at particle size, the contrast was sufficient for comparison even without it.

Here's what I saw:
- Epson and Canon colour pigment particles are around the same size. HP's pigment particles are noticeably larger. Piezography particles seem to be about the same size as Epson/Canon, and significantly smaller than MIS Eboni. Matte black particles are noticeably larger than photo black particles (in all cases, they're reasonably huge).
- Although this wasn't really a test for pigment concentration, HP inks appeared to have more pigment load than the others. All the others appeared fairly similar, apart from the Epson yellow. I'm not sure if anyone's put the inks in a centrifuge to confirm this, though. Piezography Black appears to have a lot of pigment.

Based on what we already know the effect of particle size in pigments, these would certainly explain a few things that have already been observed.

- All else being equal (chemical composition, etc.) larger particles have greater longevity than smaller particles, given the greater number of redundant molecules per particle and the smaller surface area-to-mass ratio of the pigment particles. The larger particles in HP's Vivera ink would certainly help explain their greater longevity. Canon and Epson inks having similar particle sizes would also help explain their similar longevity, once the less-stable Epson yellow pigment is removed (e.g. via a RIP). As for Piezography, carbon particles are essentially inert anyway (doubly so after encapsulation) so they could afford to go with the smaller particles.

- It is also known that larger particles also result in less brilliance and gamut than smaller particles. This is because larger particles increase scatter and are also more opaque. This would certainly explain why Vivera inks seem to have slightly less gamut than Epson/Canon inks; I think HP may have helped compensate for this by using a greater pigment load in their ink. The one exception is black ink, where opacity increases Dmax, since, unlike with colours, you *don't* want light to be reflected off the substrate. The larger particle size probably helps explain HP's greater Dmax on matte papers.

- HP's larger particles may help explain why the surface of HP prints seem more fragile than Epson and Canon inks, prior to spraying with a protective coat. The larger pigment particles are probably easier to brush off.

- Given that the Epson yellow appears to have the same pigment load as the other Epson inks, I suspect that the rumour that Epson K3 and HDR yellow contains some dye (increasing saturation while reducing lightfastness) is just a myth. The test certainly doesn't exclude the presence of yellow dye, but, given that the pigment load appears to be similar to the other inks, I doubt it. The reduced longevity is probably just due to Epson having chosen a brighter, but more fugitive, yellow pigment for their ink.

- The smaller Piezography particles compared to the MIS Eboni particles probably explain why Piezography has a warmer tone than MIS carbon (which is more neutral), and also why MIS carbon inks can't print on glossy paper.

Not that it actually changes what we do in day-to-day printing, but I just thought it was interesting...

Interesting and your experiment results fit the practical observations and theories we already had. Any chance we may see some images of that experiment?
No agglomeration of pigment particles observed in any of the ink samples? I understand that is harder to see in dried samples, maybe irregular stacking might show it.


Met vriendelijke groet, Ernst

http://www.pigment-print.com/spectralplots/spectrumviz_1.htm
January 2016 update, 700+ inkjet media white spectral plots

[shadowblade]

Interesting and your experiment results fit the practical observations and theories we already had. Any chance we may see some images of that experiment?
No agglomeration of pigment particles observed in any of the ink samples? I understand that is harder to see in dried samples, maybe irregular stacking might show it.


Met vriendelijke groet, Ernst

http://www.pigment-print.com/spectralplots/spectrumviz_1.htm
January 2016 update, 700+ inkjet media white spectral plots

No photos, unfortunately - not my microscope, I just happened to have some time to play with it.

I didn't look for agglomeration because the dried samples are not the same as how they are applied in an inkjet printer. Even the smallest, picolitre-range droplets I could produce using the micropipette are many times larger than the droplets produced by an inkjet printer. Also, the drying method is different - on a print substrate, drying is achieved by separation of the coated pigment particles from the carrier fluid via adsorption/adhesion to the inkjet layer, whereas the microscope samples were essentially flash-dried, the carrier fluid evaporating instantly in the vacuum and dumping the fluid-free pigment onto the aluminium slide.

[Paul Roark]

...
 Matte black particles are noticeably larger than photo black particles (in all cases, they're reasonably huge).

...I'm not sure if anyone's put the inks in a centrifuge to confirm this, though. Piezography Black appears to have a lot of pigment.

Based on what we already know the effect of particle size in pigments, these would certainly explain a few things that have already been observed.

- All else being equal (chemical composition, etc.) larger particles have greater longevity than smaller particles,  ...

- ... I suspect that the rumour that Epson K3 and HDR yellow contains some dye (increasing saturation while reducing lightfastness) is just a myth.
...

- The smaller Piezography particles compared to the MIS Eboni particles probably explain why Piezography has a warmer tone than MIS carbon (which is more neutral), and also why MIS carbon inks can't print on glossy paper.
...

Regarding "Eboni" MK, its relative neutrality and the relatively more neutral printing of the older, large dot size printers appears to be a function of the edge vs. area; just consider the formulas for the circumference vs. the area.  As the size increases the area where light is absorbed increases much faster than the circumference.  (At least that model appears to explain a lot of what I see in that regard.)

The current generation of Eboni is not the original, and the original was more neutral than the current generation.  From what I can tell or been told, the carbon is the same.  The dispersant, however, is updated to reflect the state of the art.  A few years ago a researcher (in Australia, I think) used an electron microscope to look at why Eboni was more neutral.  It appeared from the photo I received that, in fact, many of the particles were flocculating (agglomerating), thus increasing the, de facto, size of the particles.  We know also that the original Eboni needed more agitation than most pigments, though there were almost never clogs in the dilute Eboni-6 inks, probably because no binder was in the dilution base I made.  While the particles were relatively large, they also simply bury themselves in matte papers.  Eboni on Arches watercolor paper can even be rinsed with very little effect.

For better or worse, the new Eboni (version 1.1) uses a modern dispersant.  As a result, it is not as neutral as the older version, but it also stays in suspension much better, while still being less warm than the alternatives I've found.  People will notice that the new B&W inkset designs by me now have a light blue "toner" to offset the carbon warmth as needed.  I've gone back to my original "variable tone" approach.  (The toner uses Canon Lucia cyan and blue pigments, diluted with a clear base.)

(A glossy version of the "Eboni Variable tone" inksets is in a printer now being tested.  I have not yet decided on which carbon to use.)

I do centrifuge testing, but I have not tested Jon's Piezo inks, nor do I let the centrifuge run long enough to settle out all the solids.  I'm interested in the relative rate of compression of the spacing of the particles, thus the relative need for agitation.  The tests also show the relative sedimentation that afflicts too many inks.  As to that latter issue, I would warn that age of the ink correlates with the sedimentation for many inks.  Pay attention to the expiration dates.  Some dispersants appear to oxidize and lose their effectiveness.

Regarding rumored dyes in the pigments, many if not most of our color pigs have been referred to as "dye stacks."  They are non-soluble forms of the color dyes.  A major factor in lightfastness is the size of the final particle on the paper.  The larger they are the greater the resistance to oxidation (surface area vs. volume), which is what most fading is.  The outstanding dyes like the Epson Claria (sold in large carts to the Noritsu dry lab market) appear to be almost like little mini pigments that are still water soluble.  If Epson (probably actually Fujifilm chemistry) could make a dye that approached pigment longevity, the color pigs would be quickly replaced.  Color dyes are wonderful.  I use the Noritsu dyes in my Epson 4000 -- no agitation needed and no clogging.  (The needed dilution base formula is in my PDFs and also sold pre-mixed by MIS/www.inksupply.com -- no kickbacks to me.)

Paul
www.PaulRoark.com

[shadowblade]

Regarding "Eboni" MK, its relative neutrality and the relatively more neutral printing of the older, large dot size printers appears to be a function of the edge vs. area; just consider the formulas for the circumference vs. the area.  As the size increases the area where light is absorbed increases much faster than the circumference.  (At least that model appears to explain a lot of what I see in that regard.)

The current generation of Eboni is not the original, and the original was more neutral than the current generation.  From what I can tell or been told, the carbon is the same.  The dispersant, however, is updated to reflect the state of the art.  A few years ago a researcher (in Australia, I think) used an electron microscope to look at why Eboni was more neutral.  It appeared from the photo I received that, in fact, many of the particles were flocculating (agglomerating), thus increasing the, de facto, size of the particles.  We know also that the original Eboni needed more agitation than most pigments, though there were almost never clogs in the dilute Eboni-6 inks, probably because no binder was in the dilution base I made.  While the particles were relatively large, they also simply bury themselves in matte papers.  Eboni on Arches watercolor paper can even be rinsed with very little effect.

For better or worse, the new Eboni (version 1.1) uses a modern dispersant.  As a result, it is not as neutral as the older version, but it also stays in suspension much better, while still being less warm than the alternatives I've found.  People will notice that the new B&W inkset designs by me now have a light blue "toner" to offset the carbon warmth as needed.  I've gone back to my original "variable tone" approach.  (The toner uses Canon Lucia cyan and blue pigments, diluted with a clear base.)

(A glossy version of the "Eboni Variable tone" inksets is in a printer now being tested.  I have not yet decided on which carbon to use.)

I do centrifuge testing, but I have not tested Jon's Piezo inks, nor do I let the centrifuge run long enough to settle out all the solids.  I'm interested in the relative rate of compression of the spacing of the particles, thus the relative need for agitation.  The tests also show the relative sedimentation that afflicts too many inks.  As to that latter issue, I would warn that age of the ink correlates with the sedimentation for many inks.  Pay attention to the expiration dates.  Some dispersants appear to oxidize and lose their effectiveness.

That would probably be consistent with the electron microscopy findings.

Greater agglomeration = cooler carbon. Less agglomeration in the new inks = separate particles, which gives a slightly warmer tone. But the particles still appear to be larger than those in Piezography, so the colour tone appears cooler.

Regarding rumored dyes in the pigments, many if not most of our color pigs have been referred to as "dye stacks."  They are non-soluble forms of the color dyes.  A major factor in lightfastness is the size of the final particle on the paper.  The larger they are the greater the resistance to oxidation (surface area vs. volume), which is what most fading is.

Exactly. Even most UV-related fading is a combination of UV and oxidation effects - UV-wavelength photons carry the right amount of energy to ionise an atom in the pigment molecule, but this only results in oxidation and pigment loss if there's something for the ion to then react with. Without something to react with, it simply captures an electron and returns to its base state. UV light still has an effect, but much more slowly, since pigment degradation then relies on photons breaking covalent bonds within the structure of organic pigment molecules. Hence the utility of encapsulated pigment particles and protective sprays - the polymer layers formed are too thin to block a significant proportion of UV-wavelength photons, but, by reducing the contact between pigment molecules and potential oxidants, the effect of UV radiation is greatly reduced.

On the flip side, larger particles scatter the light more, resulting in less colour saturation and a smaller gamut. Increasing the pigment load goes a long way towards compensating for this.

  The outstanding dyes like the Epson Claria (sold in large carts to the Noritsu dry lab market) appear to be almost like little mini pigments that are still water soluble.  If Epson (probably actually Fujifilm chemistry) could make a dye that approached pigment longevity, the color pigs would be quickly replaced.  Color dyes are wonderful.  I use the Noritsu dyes in my Epson 4000 -- no agitation needed and no clogging.  (The needed dilution base formula is in my PDFs and also sold pre-mixed by MIS/www.inksupply.com -- no kickbacks to me.)

Perhaps a return to inorganic metal ion pigments, as opposed to organic pigments, is in order. Many of these won't fade even in extremes of UV light; integrated into ceramic or glass pigment particles, they are also protected from chemical reactions that could render them colourless. No need to use the highly-toxic cadmium, arsenic and similar pigments - there are plenty of far less-toxic metal ion options available. In any case, no-one's going to be licking my photos (and, if you did, you probably deserve it...). Sure, they are less strongly coloured (due to their generally-larger particle size having a lower surface area for the same mass of pigment, although it is possible to grind them smaller) but this can be compensated for by increasing the pigment load.

[Paul Roark]

... Hence the utility of encapsulated pigment particles and protective sprays - ...

...

Perhaps a return to inorganic metal ion pigments, as opposed to organic pigments, is in order. ...

I looked for an oxygen barrier coating, but it appears O2 is really hard to stop -- thus the metallic coatings on those things that need isolation from oxygen to last.  That said, even glazing makes the images last longer.  A breeze is like blowing on embers in a fire. 

I asked some of the experts (a few years ago) about those metallic pigments.  The response was that they were not only too heavy to keep in suspension, but they also acted like a grinding powder that would chew up our inkjet heads.

Paul
www.PaulRoark.com

[shadowblade]

I looked for an oxygen barrier coating, but it appears O2 is really hard to stop -- thus the metallic coatings on those things that need isolation from oxygen to last.  That said, even glazing makes the images last longer.  A breeze is like blowing on embers in a fire. 

I asked some of the experts (a few years ago) about those metallic pigments.  The response was that they were not only too heavy to keep in suspension, but they also acted like a grinding powder that would chew up our inkjet heads.

Paul
www.PaulRoark.com

Yet some inkjet heads print ceramic pigments (often metal-based, due to their heat resistance) onto tiles and other ceramics, which are even more abrasive. The printed tiles are then heated to set the colour in the glazing. So it's probably not insurmountable. Not sure what the method is, though - perhaps it's an encapsulated pigment that both improves dispersion in the relevant solvent (whether hydrophilic or hydrophobic) and makes for a softer surface that isn't an abrasive (the particles being only as abrasive as their surface).

[Paul Roark]

I'm curious about that tile printing.  In my area the people with what I'd call standard inkjet printers are printing on tiles only via use of inkjet coated materials and/or transfers.  The heat resistance and longevity of the final product is minimal.  I shelved a tile project when I ran into the limitations of the resulting products.

Paul
www.PaulRoark.com

[shadowblade]

I'm curious about that tile printing.  In my area the people with what I'd call standard inkjet printers are printing on tiles only via use of inkjet coated materials and/or transfers.  The heat resistance and longevity of the final product is minimal.  I shelved a tile project when I ran into the limitations of the resulting products.

Paul
www.PaulRoark.com

Here's one using it to print to glass: http://www.glasscanadamag.com/innovations/glass-printing-comes-of-age-1489

They're not standard inkjet printers (they're a flatbed design, since they're printing rigid materials). But the print heads are inkjet heads.

[shadowblade]

Speaking of black inks, I wonder when this stuff will make it into inks - or onto the insides of lenses:

http://www.gizmag.com/vantablack-s-vis-spray/42298/

It would give ridiculous Dmax!

[Czornyj]

Could you observe polymer pigment particles encapsulation? Canon has variable size of pigment particles - did you notice it? How about Epson in this respect?

[shadowblade]

Could you observe polymer pigment particles encapsulation?

No. Electron microscopy is a bit different from the light microscopy - I could see the surface texture (i.e. size and shape) of the particles, but not their individual components.

I could probably have done it using some sort of contrast medium to highlight different parts of the particles, but couldn't do it on the day.

Canon has variable size of pigment particles - did you notice it? How about Epson in this respect?

I only noticed it in the matte black inks from Canon, Epson and HP. There were larger particles of what I assume were carbon black, and smaller particles of what I assume were coloured pigments to neutralise the colour. This size discrepancy was not noticeable in the PK inks - I assume the larger particles are unable to pass through the gloss polymer layer in front of the ink receptive layer, so they need to grind them to the same size as the other pigments.

Of course, that doesn't mean that all the coloured pigment particles are the same size - merely that, after pigment encapsulation, they're all around the same size. As I said, I was unable to distinguish between pigment and the resin it was embedded in. I couldn't tell if there were multiple, smaller particles of pigment embedded in a single piece of resin in one colour as opposed to a single large piece of pigment in another, or if some colours consisted of a small piece of pigment surrounded by a lot of resin and others consisted of a lot of pigment and not much resin.

Not that I tested them on the day, but this is what I'd expect from a few other inks and pigments:

White paint/enamel/other opaque white pigment (e.g. in the white plastic layer of RC papers, from UV-curing printers or in the white layer of dye-sub aluminium) - a heavy concentration of large pigment particles, maximising opacity, scattering and reflectance while minimising (and keeping spectrally even) absorption within the visible spectrum.

Translucent white ink (e.g. to print white on back-illuminated films) - large particles of translucent (not white) material, to maximise scattering while minimising opacity and absorption

UV-blocking pigments (e.g. in Timeless) - very small particles. Small enough to be transparent to visible light, but large enough to be opaque to UV light.