Hi Again!

One of the joys of alkanethiolate based self-assembly is that its pretty robust. If there are free sites on the gold then bonds fairly reliably form. Characterizing precisely the the SAM "quality" in an AuNp deposit is somewhat challenging, and I'm not sure the conditions or metrics used for a "high-quality" SAM on surfaces vs on NP arrays would line up. Looking at your values, your concentrations, times, and conditions look adequate.

A very important question is, what is the chemistry of the AuNp surface before you functionalize them? Citrate should rapidly exchange, but CTAB is tough to exchange. 

I wouldn't expect the aunp construct with PEG to resist protein adsorption as effectively as a bare gold surface, simply due to the substantially higher defect density-- the PEG monolayer will be similarly more chaotic. However, you might be able to pull it off nonetheless! Once you are onto functionalizing the monolayers, the reaction can disrupt the order of the thiolate film. On 2D surfaces, we use insertion to keep the density low, which doesn't blast the monolayer, but chemical reactions have been shown to induce a lot of disorder.

This article covers a technique for completely replacing SAMs by first doing a chemical reaction to the terminal groups.  10.1021/ja807648g 

I'm happy to answer any other SAM questions you might have as best as I'm able.

Cheers,

JNH

Thank you for answering all of my questions so well!

We indeed have citrate-capped sams and have no reason to think the reaction isn't going to completion; it's just any time my experiments show variability, I sort of panic and comb through ever possible explanation.  The temperature of our room tends to drop 10F overnight (probably like 74-64) because the ventiliation sucks, so this crossed my mind. Based on what I know about physics and nucleophilicity, it really didn't seem like such a small temperature perturbation could induce a first-order effect.

By the way, do you happen to know roughly the charge of citrate-stabilized gold paticles, vs say the charge of CTPEG-capped particles?  I was hoping to find a reliable way to estimate this, or even measure it.  Was going to open a new topic, but though you might have a guess?

I guess zeta potential is probably what I'd need to measure...

I don't have a lot of experience specifically with nanoparticles. I typically limited to surfaces or uncharged nanoparticle systems and thus never really had to deal with those specific challenges.

When I was doing STM work, I would try all sorts of tricks to make my SAMs come out better and more consistent-- humidity or temperature of the room, purity of the solvent, purity of the molecule, flame annealing of the gold-- I found that most of it was practically irrelevant to my specific application.

My target was the highest possible crystallinity of purely hydrophobic SAM systems, so once the substrate was well-annealed, I would cook the samples at 78 C for a period of many days. Any time after 24 h would yield films of exquisite crystallinity as long as the substrate was predominantly (111). The reason this works is that SAM formation is dynamic-- the initial product formed within microseconds of solution exposure is dominated by kinetics-- it is scrambled and messy-- and then crystalline domains grow as molecules pop on and off. The desorption rate for molecules in well ordered domains is slower, so in time the domains grow larger. My most recent article on exchange kinetics of selenium and thiolate systems should cover the relevant reaction conditions and considerations for those systems.

Because you have functional molecules, the SAM is able to interact with itself at the molecular scale-- nearby molecules are bound through the carboxyl group, hence my prior suggestion of using acetic acid in 10x abundance in the deposition solution. I don't think that will damage the ether linkages. It might give you some better consistency. C. Woll has a lot of work on this as well, if I recall correctly.

PEG are never going to be extremely well ordered, like those STM images of alkanethiolates, but the terminal chemistry will be present on the surface. Check the literature (i think to a previous answer i linked some PEG papers) and see if elevated temperatures were used; perhaps you could design a simple model system to help you become more comfortable with the chemistry. I might consider a glass or silicon slide coated first with gold and then exposed to citrate, and then monitor PEG deposition by infrared reflectance adsorption spectroscopy or ATR.

since you are doing nhs edc coupling, I'm thinking you might have too much carboxyl peg per unit area, so when the reaction happens it further disorders the film. This was something my colleagues worried about a lot.

Article Exchange Reactions between Alkanethiolates and Alkaneselenol...

Thanks!  I will try your acetic acid idea.  I read the paper but wasn't sure if it would apply to our PEG-COO-, as the paper uses alkanethiols.  I presume it shouldn't matter much since the headgroup is the main target for the deprotenation.  Thanks for this resource as well.

That's interesting about disordering the film. For us, the EDC/NHS step seems ok; we know immediately if the coating is off just based on the initial behavior of our sensor, before even the activation step.  

The coatings I did last night seemed to work nicely.  Was we did was make sure the CTPEG had warmed to room temp for an hour beforehand to ensure no residual interactions from the frozen sample.  We thaw these samples, so they're likely to take some time to get completely mobile in solution again.  This may have been the problem on my last batch; I added a very quickly thawed coating.  Additionally, instead of letting our CTPEG completely mix into the buffer before adding our samples, we left the samples in the buffer and added the CTPEG directly.  There is a two-minute period where the CTPEG is in high concentration while the DMSO that it's stock solution is in convectually moves through the solution.  Oddly, we've done this twice, and both times get the function behavior we want.  As you said, early in the formation, the layer forms in a rapid, disordered fashion.  It might be that having this large gradient present early on actually helps this step really pack the molecules on, then later the lower concentration helps them form.

At this point, whatever works we'll take.  I guess I'm probably just making up scenarios at this point to justify our luck with one protocol working and one not.

Thanks